Qualification: 
Ph.D
ki_ram@cb.amrita.edu

Dr. K. I. Ramachandran currently serves as Professor at Department of Mechanical Engineering, School of Engineering and Professor at Center for Computational Engineering and Networking (CEN), Coimbatore Campus.

Publications

Publication Type: Journal Article

Year of Publication Title

2019

G. Rajendiran, Kumar, C., and Dr. K. I. Ramachandran, “Scalable fault models for diagnosis in a synchronous generator using feature mapping and transformation techniques”, International Journal of Prognostics and Health Management, vol. 9, p. 11, 2019.[Abstract]


Condition based maintenance (CBM) needs data acquired during healthy and faulty conditions to develop intelligent system for fault diagnosis. However, fault injection is not al-lowed/possible in a highly expensive components of complex/critical systems to collect fault condition data. Therefore , proto-type/small working models are used to conduct experiments for abnormal/fault conditions, to obtain and scale the intelligence of the system for effective health monitoring of complex system. This methodology is referred as scalable fault models. For proof of concept, in this work, we considered two different capacity synchronous generators with rating of 3 kVA and 5 kVA to emulate the behavior of proto-type/small working model and complex system respectively, for scalable fault models. We explored feature mapping and transformation techniques to achieve effective scalability. From the preliminary experiments, it is observed that the base-line system performance deteriorated due to the changes in the system (capacity) and its characteristics with load changes. We therefore, expressed the input features in terms of load and system independent manner, to make the features less dependent on load and system variations. We explored locality constrained linear coding (LLC) to express the features load/system independently. It is observed that experimenting LLC with the backend support vector machine (SVM) classifier gave the best fault classification performance for linear kernel, suggesting that the faults are linearly separable in the new feature space. Since the LLC mapped feature space is linearly separable, we then explored linear feature transformation technique, nuisance attribute projection (NAP) on the LLC mapped feature space to further minimize the load/system specific variations. We observed that LLC-NAP improved the overall accuracy and sensitivity of the classifier significantly. We also noted that the performance of NAP was limited in the original feature space since the feature space (NAP without LLC) is nonlinear with load/system variations.

More »»

2019

K. T. Sreekumar, George, K. K., C. Kumar, S., and Dr. K. I. Ramachandran, “Performance enhancement of the machine-fault diagnosis system using feature mapping, normalisation and decision fusion”, IET Science, Measurement Technology, vol. 13, pp. 1287-1298, 2019.[Abstract]


A unified fault modelling approach for machines with varying operating speeds is of interest in automating production facilities. Building a unified fault model is challenging due to the presence of speed-specific attributes in the features derived. In this work, it is shown how feature selection, mapping, normalisation and fusion of heterogeneous systems can help enhance the performance of speed-independent (SI) machine-fault diagnosis systems. Statistical features, obtained after applying feature selection, are used as input to a support vector machine (SVM) back-end classifier as the baseline system. Entropy-based feature selection algorithm is proposed to improve the performance of the fault diagnosis system. Furthermore, to make the fault diagnosis system independent of speed, locality constrained linear coding (LLC), Fisher vector encoding (FVE) and mean and variance normalisation (MVN) are used. The LLC-MVN system and FVE-MVN system map the input features in terms of SI basis vectors to make the features robust to speed-specific variations. Finally, the decision scores of the time-domain LLC-MVN-SVM, frequency-domain LLC-MVN-SVM systems and variational mode decomposition-based FVE-MVN-SVM system were fused with appropriate weighting factors. The detection error trade-off curve is also used as a performance measure for intelligent fault diagnosis systems.

More »»

2019

A. Lekshmi and Dr. K. I. Ramachandran, “Parkinson's Tremor Suppression Using Active Vibration Control Method”, IOP Conference Series: Materials Science and Engineering, vol. 577, p. 012056, 2019.[Abstract]


Parkinson’s disease is the second most common neuro-degenerative disorder. The hallmark symptom of Parkinson’s disease known as tremor, affects the patients in carrying out most of their daily life activities. In this paper, a biodynamic model of human arm is used to imitate the behavior of Parkinson’s tremor and suppression of this tremor is performed using a Proportional-Integral-Derivative (PID) controller which is optimized using Genetic Algorithm (GA) to obtain better performance. Active vibration controlling technique which incorporates sensors and actuators to detect and counteract the tremor is implemented with the help of PID controller.

More »»

2019

P. T. Kumar, Dr. Saimurugan M., Haran, R. B. Hari, Siddharth, S., and Dr. K. I. Ramachandran, “A multi-sensor information fusion for fault diagnosis of a gearbox utilizing discrete wavelet features”, Measurement Science and Technology, vol. 30, 2019.[Abstract]


A gear box is widely employed in automobiles and industrial machines for transmitting power and torque. It operates under various working conditions for prolonged hours increasing the chance of gearbox failure. Major faults in gear systems are caused due to wear, scoring, pitting, tooth fracture, etc. Gear box failure leads to increases in machine downtime and maintenance costs. The nature and location of such failures can be identified with precision using condition monitoring techniques. In this study, machine condition data are acquired from the gear box using a vibration accelerometer, microphone and acoustic emission sensors under different operating conditions, such as three loading conditions (0 N, 5 N, 10 N) and three rpm variations (500, 750, 1000). The wavelet features are extracted from the acquired vibration, sound and acoustic emission signal, and prominent features are identified. To automate the process of fault diagnosis, machine learning algorithms (artificial neural network, support vector machine, and proximal support vector machine) are utilized. Dual and multi-sensor fusion is implemented with the help of prominent features, to intensify the classification accuracy. The performance of the individual signals, and dual and multi-sensor fused models in gearbox fault diagnosis are compared and discussed in detail. © 2019 IOP Publishing Ltd.

More »»

2019

R. Gini John and Dr. K. I. Ramachandran, “Extraction of foetal ECG from abdominal ECG by nonlinear transformation and estimations.”, Comput Methods Programs Biomed, vol. 175, pp. 193-204, 2019.[Abstract]


BACKGROUND AND OBJECTIVE: This paper proposes a simple yet effective method for the extraction of foetal ECG from abdominal ECG which is necessary due to similar spatial and temporal content of mother and foetal ECG.

METHODS: The proposed algorithm for extraction of foetal ECG (fECG) from abdominal signal uses single channel. Pre-processing of abdominal ECG (abdECG) has been done to eliminate noise and condition the signal. The maternal ECG R-peaks have been detected based on thresholding, first order Gaussian differentiation and zero cross detection on pre-processed signal. Having identified R-peaks and pre-processed signal as base, using Maximum Likelihood Estimation, one beat including QRS complex morphology of maternal ECG (mECG) has been constructed. Extraction of maternal ECG from abdECG is done based on the constructed beat, R-peak locations and its corresponding QRS complex of abdECG. Extracted mECG has been cancelled from abdECG. This results in foetal ECG with residual noise. The noise has been reduced by Polynomial Approximation and Total Variation (PATV) to improve SNR. This approach ensures no loss of partially or completely overlapped fECG signals due to mECG removal. The algorithm is tested on three database namely daISy (DB), Physiobank challenge 2013 (DB) and abdominal and direct foetal ECG database (adfecgdb) of Physiobank (DB).

RESULTS: The algorithm detected no false positives or false negatives with certain channel for DB, DB and DB which shows that the proposed algorithm can achieve good performance. Overall accuracy and sensitivity of the system is 98.53% and 100% for DB. Best accuracy and sensitivity of 97.77% and 98.63% are obtained for DB. Best accuracy of 92.41% and sensitivity of 93.8% are obtained for DB. Correlation coefficient between actual foetal heart rate (fHR) and estimated fHR of 0.66 for DB and 0.59 for DB is obtained. The method has obtained overall F1 score of 99.25% for DB, 96.04% for DB and 94.25% for DB. It has obtained a best MSE of fHR and overall MSE of R-R interval which is 10.8bpm and 2.2 ms for DB, 12bpm and 2.14 ms for DB.

CONCLUSION: The results for different public databases show that the proposed method is capable of providing good results. The foetal QRS, R-peaks and R-R intervals have also been obtained in this method. Thus, it gives a significant contribution in the required area of research.

More »»

2018

V. D, Dr. K. I. Ramachandran, and S, A., “Pedestrian Detection using Data Fusion of Leddar Sensor and Visual Camera with the help of Machine Learning”, International Journal of Pure and Applied Mathematics, vol. 118, 20 vol., 2018.[Abstract]


In the concept of autonomous cars, pedestrian safety is one of the important factors. So there is a need for the improvement of the safety of pedestrians with growing number of automobiles and automobile technologies. There have been many research and development works going on in the area of pedestrian detection and distance estimation. And Data fusion has become one of the most talked about topics in the area of Advanced Driver Assistance System (ADAS) in the recent years. This paper attempts to use a data fusion technique to detect pedestrian as well as to estimate the distance of the pedestrian using Light Emitting Diode Detection and Ranging (Leddar) and Images sensor. The main problem in this data fusion is combining the output of the Leddar sensor with that of the image sensor. A unique method of correlating the Leddar output with image sensor output is identified and an algorithm is developed using this method which can identify Region of Interests (ROIs) in the image sensor output with respect to the Leddar sensor output. The ROIs will then be processed using machine learning algorithms to detect pedestrians. It is found that this data fusion is able to identify whether there is a pedestrian along with the distance of the pedestrian.

More »»

2018

H. K, Dr. K. I. Ramachandran, and S, A., “Driver Fatigue detection using Infrared sensor and Adaptive Neuro-Fuzzy Inference System”, International Journal of Pure and Applied Mathematics, vol. 118, 20 vol., pp. 4209-4214, 2018.[Abstract]


This paper discusses a non-invasive and non-intrusive
approach for driver fatigue detection.The fatigue condition of the
driver is identified with the help of a trained neuro- fuzzy system
mitigating the risk of a road accident. In this work, we have
analyzed the facial temperature variations of the driver for
estimating the level offatigue.Infrared sensor is used for
acquisition of facial temperature of subjects and adjacent
environment temperature changes. The temperature information
of the subjects is segregated into facial temperature and
environment temperature and is used for training the adaptive
neuro fuzzy system. Therefore, the fuzzy system will provide a
numerical output which helps in determiningthe fatigue
condition of the subject under consideration. Hence the fuzzy
system is well trained with all facial temperatures of the subject,
fatigue is determined accurately.

More »»

2018

S. Ravikumar and Dr. K. I. Ramachandran, “Tool Wear Monitoring of Multipoint Cutting Tool using Sound Signal Features Signals with Machine Learning Techniques”, Materials Today: Proceedings, vol. 5, pp. 25720 - 25729, 2018.[Abstract]


This paper proposes a tool wear monitoring system using sound signals acquired during milling of aluminium alloys. Tool wear monitoring is important for achieving surface finish and real time control of dimensional accuracy. Experiment was performed in a CNC machining centre with recommended cutting conditions. Tungsten carbide inserts in a face milling cutter was used and the wear conditions were simulated. Statistical features of the signals were fed to random forest tree algorithm. The wavelet features of the signals were also extracted and a decision tree classification model was built. A feature subset selection was performed by feature evaluators with search algorithms. Observations were made on the performance of classifier model using statistical features and with full set of features over subset of wavelet.

More »»

2018

P. Krishnakumar, K. Ramesh Kumar, and Dr. K. I. Ramachandran, “Machine learning based tool condition classification using acoustic emission and vibration data in high speed milling process using wavelet features”, Intelligent Decision Technologies, vol. 12, pp. 1-18, 2018.[Abstract]


It is important to develop an intelligent tool condition monitoring system to increase productivity and promoting automation in metal cutting process. Many attempts have been made in the past to develop such systems using signals from various sensors such as dynamometer, current, accelerometer, acoustic emission, current and voltage, etc. But the successes of different sensor based systems are limited due to the complexity of tool wear process. The research is still ongoing for improved tool condition monitoring system with applications of advance signal processing techniques and artificial intelligent models. In this study, tool conditions are monitored using the vibration and acoustic emission signatures during high speed machining of titanium alloy (Ti-6Al-4V). Using discrete wavelet transforms wavelets coefficients of vibration and acoustic emission signals are extracted using haar, daubechies, biorthogonal and reverse biorthogonal wavelets. Machine learning algorithms such as decision tree, naive bayes, support vector machine and artificial neural networks are used to predict the tool condition. Results indicate the effectiveness of acoustic emission and vibration data using wavelets for classifying the tool conditions with the aid of machine learning algorithms. A correlation is established between the tool conditions and sensor data. Support vector machine trained by vibration data appears to be predicting the tool conditions with good accuracy compared to decision trees, naive bayes and artificial neural network. Results obtained in this study will be useful to develop an intelligent on-line tool condition monitoring system.

More »»

2018

S. Adarsh and Dr. K. I. Ramachandran, “Design of Sensor Data Fusion Algorithm for Mobile Robot Navigation Using ANFIS and Its Analysis Across the Membership Functions”, Automatic Control and Computer Sciences, vol. 52, pp. 382-391, 2018.[Abstract]


Abstract: Design and development of autonomous mobile robots attracts more attention in the era of autonomous navigation. There are various algorithms used in practice for solving research problems related to the robot model and its operating environment. This paper presents the design of data fusion algorithm using Adaptive Neuro Fuzzy Interface (ANFIS) for the navigation of mobile robots. Detailed analysis of various membership functions (MFs) provided in this paper helps to select the most appropriate MF for the design of similar navigation systems. The combined use of fuzzy and neural networks in ANFIS makes the measured distance value of the residual covariance consistent with its actual value. The data fusion algorithm within the controller of the mobile robot fuses the input from ultrasonic and infrared sensors for better environment perception. The results indicate that the data fusion algorithm provides minimal root mean square error (RMSE) and mean absolute percentage error (MAPE) when compared with that of the individual sensors. © 2018, Allerton Press, Inc.

More »»

2018

T. Praveenkumar, Sabhrish, B., Dr. Saimurugan M., and Dr. K. I. Ramachandran, “Pattern recognition based on-line vibration monitoring system for fault diagnosis of automobile gearbox”, Measurement, vol. 114, pp. 233 - 242, 2018.[Abstract]


Gearbox is an important equipment in an automobile to transfer power from the engine to the wheels with various speed ratios. The maintenance of the gearbox is a top criterion as it is prone to a number of failures like tooth breakage and bearing cracks. Techniques like vibration monitoring have been implemented for the fault diagnosis of the gearbox over the years. But, the experiments are usually conducted in lab environment where the actual conditions are simulated using setup consisting of an electric motor, dynamometer, etc. This work reports the feasibility of performing vibrational monitoring in real world conditions, i.e. by running the vehicle on road and performing the analysis. The data was acquired for the various conditions of the gearbox and features were extracted from the time-domain data and a decision tree was trained for the time-domain analysis. Fast Fourier Transform was performed to obtain the frequency domain which was divided into segments of equal size and the area covered by the data in each segment was calculated for every segment to train decision trees. The classification efficiencies of the decision trees were obtained and in an attempt to improve the classification efficiencies, the time-domain and frequency-domain analysis was also performed on the normalised time-domain data. From, the results obtained, it was found that performing time-domain analysis on normalised data had a higher efficiency when compared with the other methods. Instantaneous processing of the acquired data from the accelerometer enables faster diagnosis. Hence, online condition monitoring has gained importance with the advent of powerful microprocessors. A windows application that has been developed to automate the process was found to be essential and accurate.

More »»

2018

K. K. George, Kumar, C. S., Sivadas, S., Dr. K. I. Ramachandran, and Panda, A., “Analysis of cosine distance features for speaker verification”, Pattern Recognition Letters, vol. 112, pp. 285-289, 2018.[Abstract]


In this paper, we describe a method for representing the acoustic similarity of a target speaker with respect to a set of known speakers as a feature for speaker verification. We propose a novel distance based representation by encoding the cosine distance between i-vectors of the utterances belonging to target speaker and reference speakers. The new feature is referred to as cosine distance feature (CDF) and is used with a support vector machine (SVM) classifier (CDF-SVM). We show that reference speakers who rank high in acoustic similarity to the target speaker are more important for better speaker discrimination. A sparse representation of the CDF, that retains only a few of the largest values which correspond to the most similar reference speakers in the CDF vector is found to perform better than the baseline CDF system. We also explore speaker specific CDF where each target speaker has specific subset of most acoustically similar reference speakers. We show that the acoustic similarities between the target and reference speakers are best captured using an intersection kernel SVM. Experimental results on the core short2-short3 condition of NIST 2008 SRE, for both female and male trials, show that the speaker specific CDF outperforms the i-vector and speaker independent CDF based state-of-the-art speaker verification systems. © 2018 Elsevier B.V.

More »»

2018

Krishna Kumar P., K. Ramesh Kumar, and Dr. K. I. Ramachandran, “Acoustic emission based tool condition classification in a precision high speed machining of titanium alloy (Ti-6Al-4V): A machine learning approach”, International Journal of Computational Intelligence and Applications, vol. 17, 2018.[Abstract]


Mechanical and chemical properties of titanium alloy have led to its wide range of applications in aerospace and biomedical industries. The heat generation and its transfer from the cutting zone are critical in machining of titanium alloys. The process of transferring heat from the primary cutting zone is difficult due to poor thermal conductivity of titanium alloy, and it will lead to rapid tool wear and poor surface finish. An effective tool monitoring system is essential to predict such variations during machining process. In this study, using a high-speed precision mill, experiments are conducted under optimum cutting conditions with an objective of maximizing the life of tungsten carbide tool. Tool wear profile is established and tool conditions are arrived on the basis of the surface roughness. Acoustic emission (AE) signals are captured using an AE sensor during machining of titanium alloy. Statistical features are extracted in time and frequency domain. Features that contain rich information about the tool conditions are selected using J48 decision tree (DT) algorithm. Tool condition classification abilities of DT and support vector machines are studied in time and frequency domains. © 2018 World Scientific Publishing Europe Ltd.

More »»

2018

Krishna Kumar P., K. Ramesh Kumar, and Dr. K. I. Ramachandran, “Feature level fusion of vibration and acoustic emission signals in tool condition monitoring using machine learning classifiers”, International Journal of Prognostics and Health Management, vol. 9, no. 8, pp. 2153-2648, 2018.[Abstract]


To implement the tool condition monitoring system in a metal cutting process, it is necessary to have sensors which will be able to detect the tool conditions to initiate remedial action. There are different signals for monitoring the cutting process which may require different sensors and signal processing techniques. Each of these signals is capable of providing information about the process at different reliability level. To arrive a good, reliable and robust decision, it is necessary to integrate the features of the different signals captured by the sensors. In this paper, an attempt is made to fuse the features of acoustic emission and vibration signals captured in a precision high speed machining center for monitoring the tool conditions. Tool conditions are classified using machine learning classifiers. The classification efficiency of machine learning algorithms are studied in time-domain, frequencydomain and time-frequency domain by feature level fusion of features extracted from vibration and acoustic emission signature. © 2018, Prognostics and Health Management Society. All rights reserved.

More »»

2018

N. V. Varghees, Dr. K. I. Ramachandran, and Dr. Soman K. P., “Wavelet-based Fundamental Heart Sound Recognition Method using Morphological and Interval Features”, Healthcare Technology Letters, vol. 5, pp. 81-87, 2018.[Abstract]


Accurate and reliable recognition of fundamental heart sounds (FHSs) plays a significant role in automated analysis of heart sound (HS) patterns. This Letter presents an automated wavelet-based FHS recognition (WFHSR) method using morphological and interval features. The proposed method first performs the decomposition of phonocardiogram (PCG) signal using a synchrosqueezing wavelet transform to extract the HSs and suppresses the murmurs, low-frequency and high-frequency noises. The HS delineation (HSD) is presented using Shannnon energy envelope and amplitude-dependent thresholding rule. The FHS recognition (FHSR) is presented using interval, HS duration and envelope area features with a decision-rule algorithm. The performance of the method is evaluated on PASCAL HSs Challenge, PhysioNet/CinC HS Challenge, eGeneralMedical databases and real-time recorded PCG signals. Results show that the HSD approach achieves an average sensitivity (Se) of 98.87%, positive predictivity (Pp) of 97.50% with detection error rate of 3.67% for PCG signals with signal-to-noise ratio of 10 dB, and outperforms the existing HSD methods. The proposed FHSR method achieves a Se of 99.00%, Sp of 99.08% and overall accuracy of 99.04% on both normal and abnormal PCG signals. Evaluation results show that the proposed WFHSR method is able to accurately recognise the S1/S2 HSs in noisy real-world PCG recordings with murmurs and other abnormal sounds. © 2018 Institution of Engineering and Technology. All rights reserved.

More »»

2017

T. Praveenkumar, Dr. Saimurugan M., and Dr. K. I. Ramachandran, “Comparision of Sound, Vibration and motor current signature analysis for detection of gearbox faults”, International Journal of Prognostics and Health Management, vol. 8, no. 2, pp. 1-10, 2017.[Abstract]


Gear box is used in automobiles and industries for power transmission under different working conditions and applications. Failure in a gear box at unexpected time leads to increase in machine downtime and maintenance cost. In order to overcome these losses, the most effective condition monitoring technique has to be used for early detection of faults. Vibration and sound signal analysis have been used for monitoring the condition of rotating machineries. Motor Current Signature Analysis (MCSA) has rarely been used in gearbox condition monitoring. This work presents a methodology based on vibration, sound and motor current signal analysis for diagnosis of gearbox faults under various simulated gear and bearing fault conditions. Statistical features were extracted from the raw data of these three transducer signals and the best features were selected from the extracted features. Then the selected features were given as an input to Artificial Neural Network (ANN) and Support Vector Machine (SVM) classifiers and their performances were compared. In recent years, Hybrid Electric Vehicles (HEV) are gaining more interest for their advances and this work had a scope in monitoring the power loss in hybrid electric vehicle gearbox using MCSA.

More »»

2017

T. Praveenkumar, Saimurugan, M., and Dr. K. I. Ramachandran, “Comparison of vibration, sound and motor current signature analysis for detection of gear box faults”, International Journal of Prognostics and Health Management, vol. 8, 2017.[Abstract]


Gear box is used in automobiles and industries for power transmission under different working conditions and applications. Failure in a gear box at unexpected time leads to increase in machine downtime and maintenance cost. In order to overcome these losses, the most effective condition monitoring technique has to be used for early detection of faults. Vibration and sound signal analysis have been used for monitoring the condition of rotating machineries. Motor Current Signature Analysis (MCSA) has rarely been used in gearbox condition monitoring. This work presents a methodology based on vibration, sound and motor current signal analysis for diagnosis of gearbox faults under various simulated gear and bearing fault conditions. Statistical features were extracted from the raw data of these three transducer signals and the best features were selected from the extracted features. Then the selected features were given as an input to Artificial Neural Network (ANN) and Support Vector Machine (SVM) classifiers and their performances were compared. In recent years, Hybrid Electric Vehicles (HEV) are gaining more interest for their advances and this work had a scope in monitoring the power loss in hybrid electric vehicle gearbox using MCSA. © 2017, Prognostics and Health Management Society. All rights reserved.

More »»

2017

R. Gopinath, Kumar, C. S., and Dr. K. I. Ramachandran, “Fisher vector encoding for improving the performance of fault diagnosis in a synchronous generator”, Measurement: Journal of the International Measurement Confederation, vol. 111, pp. 264-270, 2017.[Abstract]


In this work, we experiment with Fisher vector encoding to map the feature vectors into a higher dimensional space and use linear support vector machine (SVM) for improving the performance of inter turn fault diagnosis in a 3 kVA synchronous generator. Fisher vector encoding computes the first and second order differences between the feature vectors and Gaussians. We compare the performance of Fisher vector encoding with sparse coding and locality constrained linear coding (LLC). From the experiments and results, we observed that Fisher vector encoding is the most computationally efficient algorithm when compared to feature mapping using sparse coding and locality constrained linear coding (LLC). © 2017 Elsevier Ltd

More »»

2017

J. Selvaraj, P. Marimuthu, Dr. Sriram Devanathan, and Dr. K. I. Ramachandran, “Mathematical modelling of raw material preheating by energy recycling method in metal casting process”, Pollution Research, vol. 36, pp. 217-228, 2017.[Abstract]


Modern economic development programs critically depend on reliable supply of energy. Energy conservation has become the need of the hour. Metal casting industry in one among the many industries which uses a lot of energy for its production. This paper presents a novel method of energy recycling in the sand casting process, which readily translates into substantial energy conservation in foundries. The heat that is being wasted into sand during the solidification process is used to preheat the raw material that is melted, for the subsequent pouring. The influence of the casting parameters such as, offset distance, moisture content in the molding sand, and the insulator thickness on the temperature gain by the raw material, have been analyzed. A mathematical model was developed via statistical analysis of the experimental data, to predict the amount of heat recovered from the solidifying molten metal, for any specific combination of values for the experimental parameters. The predicted values are in good agreement with the experimental values.

More »»

2017

N. V. Varghees and Dr. K. I. Ramachandran, “Effective heart sound segmentation and murmur classification using empirical wavelet transform and instantaneous phase for electronic stethoscope”, IEEE Sensors Journal, vol. 17, pp. 3861-3872, 2017.[Abstract]


Accurate measurement of heart sound and murmur parameters is of great importance in the automated analysis of phonocardiogram (PCG) signals. In this paper, we propose a novel unified PCG signal delineation and murmur classification method without the use of reference signal for automatic detection and classification of heart sounds and murmurs. The major components of the proposed method are the empirical wavelet transform-based PCG signal decomposition for discriminating heart sounds from heart murmurs and suppressing background noises, the Shannon entropy envelope extraction, the instantaneous phase-based boundary determination, heart sound and murmur parameter extraction, the systole/diastole discrimination and the decision rules-based murmur classification. The accuracy and robustness of the proposed method is evaluated using a wide variety of normal and abnormal PCG signals taken from the standard PCG databases, including PASCAL heart sounds challenge database, PhysioNet/CinC challenge heart sound database, and real-time PCG signals. Evaluation results show that the proposed method achieves an average sensitivity (Se) of 94.38%, positive predictivity (Pp) of 97.25%, and overall accuracy (OA) of 91.92% for heart sound segmentation and Se of 97.58%, Pp of 96.46%, and OA of 94.21% in detecting the presence of heart murmurs for SNR of 10 dB. The method yields an average classification accuracy of 95.5% for the PCG signals with SNR of 20 dB. Results show that the proposed method outperforms other existing heart sound segmentation and murmur classification methods. © 2017 IEEE.

More »»

2016

V. Abhijit, Sugumaran, V., and Dr. K. I. Ramachandran, “Fault Diagnosis of Bearings using Vibration Signals and Wavelets”, Indian Journal of Science and Technology, vol. 9, 2016.[Abstract]


Objectives: Being widely used in most of the industrial machineries, bearings are subjected to wear and tear. Failure of bearings can incur heavy losses in the industries. In order to prevent such mishaps during operation, it is necessary to subject the bearings to a suitable fault diagnosis technique. Methods/Statistical Analysis: Vibration analysis is performed to detect the fault in bearings. For the fault analysis, vibration signals were taken for good, inner race defect, outer race defect and combination of these defects. Since vibration signals are complex and the defect related signature is buried deep within the noise and high frequency resonance, simple signal processing cannot be used for effectively detecting bearing fault. In this paper, discrete wavelets transform were used to detect bearing faults. For wavelet and feature selection, J48 decision tree algorithm was used. For feature classification, Best First Tree (BFT) algorithm was used. Findings: The experimental results indicate biorthogonal wavelets show maximum successful bearing fault detection rate. The classification accuracy was calculated and found to be 96.25%. This result is further refined to get better classification accuracy and the final result was found to be 98%. Application/Improvements: This can be considered to be a part of a preventive maintenance method in order to avoid mishaps in industries. The classification accuracy can be further improved using different algorithms.

More »»

2016

C. Sukumar and Dr. K. I. Ramachandran, “Three dimensional design, simulation and optimization of a novel, universal diabetic foot offloading orthosis”, IOP Conference Series: Materials Science and Engineering, vol. 149, p. 012140, 2016.[Abstract]


Leg amputation is a major consequence of aggregated foot ulceration in diabetic patients. A common sense based treatment approach for diabetic foot ulceration is foot offloading where the patient is required to wear a foot offloading orthosis during the entire treatment course. Removable walker is an excellent foot offloading modality compared to the golden standard solution - total contact cast and felt padding. Commercially available foot offloaders are generally customized with huge cost and less patient compliance. This work suggests an optimized 3D model of a new type light weight removable foot offloading orthosis for diabetic patients. The device has simple adjustable features which make this suitable for wide range of patients with weight of 35 to 74 kg and height of 137 to 180 cm. Foot plate of this orthosis is unisexual, with a size adjustability of (US size) 6 to 10. Materials like Aluminum alloy 6061-T6, Acrylonitrile Butadiene Styrene (ABS) and Polyurethane acted as the key player in reducing weight of the device to 0.804 kg. Static analysis of this device indicated that maximum stress developed in this device under a load of 1000 N is only 37.8 MPa, with a small deflection of 0.150 cm and factor of safety of 3.28, keeping the safety limits, whereas dynamic analysis results assures the load bearing capacity of this device. Thus, the proposed device can be safely used as an orthosis for offloading diabetic ulcerated foot.

More »»

2016

A. Sasidharan, Salahuddin, M. Kaleemuddi, Bose, D., and Dr. K. I. Ramachandran, “Performance comparison of Infrared and Ultrasonic sensors for obstacles of different materials in vehicle/ robot navigation applications”, IOP Conference Series: Materials Science and Engineering, vol. 149, p. 012141, 2016.[Abstract]


In robotics, Ultrasonic sensors and Infrared sensors are commonly used for distance measurement. These low-cost sensors fundamentally address majority of problems related to the obstacle detection and obstacle avoidance. In this paper, the performance comparison of ultrasonic and infrared measurement techniques across obstacles of different types of materials presented. The Vehicle model integrated with the sensors, moving with constant velocity towards different types of obstacles for capturing the distance parameter. Based on the data acquired from the sensors, correlation analysis of the measured distance with actual distance performed. This analysis will be very much useful, to select the right sensor - Ultrasonic sensor / Infrared sensor or a combination of both sensors, while developing the algorithm for addressing obstacle detection problems. The detection range and inherent properties of sensors (reflection/ absorption etc.) also were tested in this experiment.

More »»

2016

Krishna Kumar P., Sripathi, J., Vijay, P., and Dr. K. I. Ramachandran, “Finite Element Modelling and Residual Stress Prediction in End Milling of Ti6Al4Valloy”, IOP Conference Series: Materials Science and Engineering, vol. 149, p. 012154, 2016.[Abstract]


Titanium and its alloys are materials that exhibit unique combination of mechanical and physical properties that enable their usage in various fields. In spite of having a lot of advantages, their usage is limited because they are difficult to machine due to their inherent properties of high specific heat capacity, reactivity with tool and low thermal conductivity thereby causing excessive tool wear. To facilitate the process of machining, it becomes necessary to find out and relieve the residual stress caused during machining. Since experiments cannot be performed for each instance, creation of an FE model is desirable. In this paper a finite element analysis (FEA) of the machining of Ti6Al4V for different cutting speeds is presented. A 3D finite element model is developed with the Titanium alloy (Ti6Al4V) as the workpiece and a four flute carbide tip end mill cutter as the tool to predict the residual stress developed within the titanium alloy after machining. The finite element model utilises the Johnson-Cook model to depict the plasticity and the damage criteria and implements the Arbitrary Lagrangian Eulerian (ALE) formulation to increase the accuracy of the model. The FE model has been developed and the findings are presented. The results indicate that residual stresses are maximum at the surface and decrease linearly along the depth and increase as the cutting speed and depth of cut are increased.

More »»

2016

R. Gopinath, Dr. Santhosh Kumar C., and Dr. K. I. Ramachandran, “Scalable Fault Models for Diagnosis of Synchronous Generators”, International Journal of Intelligent Systems Technologies and Applications, 2016.

2016

R. Gopinath, Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, Upendranath, V., and Kiran, P. V. R. Sai, “Intelligent fault diagnosis of synchronous generators”, Expert Systems with Applications, vol. 45, pp. 142 - 149, 2016.[Abstract]


Condition based maintenance (CBM) requires continuous monitoring of mechanical/electrical signals and various operating conditions of the machine to provide maintenance decisions. However, for expensive complex systems (e.g. aerospace), inducing faults and capturing the intelligence about the system is not possible. This necessitates to have a small working model (SWM) to learn about faults and capture the intelligence about the system, and then scale up the fault models to monitor the condition of the complex/prototype system, without ever injecting faults in the prototype system. We refer to this approach as scalable fault models. We check the effectiveness of the proposed approach using a 3 kVA synchronous generator as \{SWM\} and a 5 kVA synchronous generator as the prototype system. In this work, we identify and remove the system-dependent features using a nuisance attribute projection (NAP) algorithm to model a system-independent feature space to make the features robust across the two different capacity synchronous generators. The frequency domain statistical features are extracted from the current signals of the synchronous generators. Classification and regression tree (CART) is used as a back-end classifier. \{NAP\} improves the performance of the baseline system by 2.05%, 5.94%, and 9.55% for the R, Y, and B phase faults respectively.

More »»

2016

A. K. Sandeep, Nithin, S., and Dr. K. I. Ramachandran, “An image processing based pedestrian detection system for driver assistance”, International Journal of Control Theory and Applications, vol. 9, pp. 7369-7375, 2016.[Abstract]


Advanced Driver Assistance System (ADAS) has developed further from just improving traffic safety to an area to help the driver to anticipate accidents. The Pedestrian Protection System (PPS) is a part of ADAS that alerts the driver once the pedestrian is detected. This paper suggests an approach for detecting pedestrians from a real time video and performing braking action once detected. Real time video is captured and the moving background is first modeled. Foreground segmentation is performed on a modeled background. Blob analysis is performed on the foreground pixels to detect the presence of pedestrian. As a control action an alert is given acoustically and braking is performed by actuating the brake pedals using a controller and actuator as a further measure. © International Science Press.

More »»

2016

A. Sharma, Subramaniam, S. Devi, Dr. K. I. Ramachandran, Lakshmikanthan C., Krishna, S., and Sundaramoorthy, S. K., “Smartphone-based Fundus Camera Device (MII Ret Cam) and Technique with Ability to Image Peripheral Retina.”, Eur J Ophthalmol, vol. 26, no. 2, pp. 142-4, 2016.[Abstract]


<p><b>PURPOSE: </b>To demonstrate an inexpensive smartphone-based fundus camera device (MII Ret Cam) and technique with ability to capture peripheral retinal pictures.</p><p><b>METHODS: </b>A fundus camera was designed in the form of a device that has slots to fit a smartphone (built-in camera and flash) and 20-D lens. With the help of the device and an innovative imaging technique, high-quality fundus videos were taken with easy extraction of images.</p><p><b>RESULTS: </b>The MII Ret Cam and innovative imaging technique was able to capture high-quality images of peripheral retina such as ora serrata and pars plana apart from central fundus pictures.</p><p><b>CONCLUSIONS: </b>Our smartphone-based fundus camera can help clinicians to monitor diseases affecting both central and peripheral retina. It can help patients understand their disease and clinicians convincing their patients regarding need of treatment especially in cases of peripheral lesions. Imaging peripheral retina has not been demonstrated in existing smartphone-based fundus imaging techniques. The device can also be an inexpensive tool for mass screening.</p>

More »»

2015

N. V Varghees and Dr. K. I. Ramachandran, “Multistage decision-based heart sound delineation method for automated analysis of heart sounds and murmurs.”, Healthc Technol Lett, vol. 2, no. 6, pp. 156-63, 2015.[Abstract]


A robust multistage decision-based heart sound delineation (MDHSD) method is presented for automatically determining the boundaries and peaks of heart sounds (S1, S2, S3, and S4), systolic, and diastolic murmurs (early, mid, and late) and high-pitched sounds (HPSs) of the phonocardiogram (PCG) signal. The proposed MDHSD method consists of the Gaussian kernels based signal decomposition (GSDs) and multistage decision-based delineation (MDBD). The GSD algorithm first removes the low-frequency (LF) artefacts and then decomposes the filtered signal into two subsignals: the LF sound part (S1, S2, S3, and S4) and the high-frequency sound part (murmurs and HPSs). The MDBD algorithm consists of absolute envelope extraction, adaptive thresholding, and fiducial point determination. The accuracy and robustness of the proposed method is evaluated using various types of normal and pathological PCG signals. Results show that the method achieves an average sensitivity of 98.22%, positive predictivity of 97.46%, and overall accuracy of 95.78%. The method yields maximum average delineation errors of 4.52 and 4.14 ms for determining the start-point and end-point of sounds. The proposed multistage delineation algorithm is capable of improving the delineation accuracy under time-varying amplitudes of heart sounds and various types of murmurs. The proposed method has significant potential applications in heart sounds and murmurs classification systems.

More »»

2015

P. Krishnakumar, Rameshkumar, K., and Dr. K. I. Ramachandran, “Tool Wear Condition Prediction Using Vibration Signals in High Speed Machining (HSM) of Titanium (Ti-6Al-4V) Alloy”, Procedia Computer Science, vol. 50, pp. 270 - 275, 2015.[Abstract]


Ti-6Al-4V is extensively used in aerospace and bio-medical applications. In an automated machining environment monitoring of tool conditions is imperative. In this study, Experiments were conducted to classify the tool conditions during High Speed Machining of Titanium alloy. During the machining process, vibration signals were monitored continuously using accelerometer. The features from the signal are extracted and a set of prominent features are selected using Dimensionality Reduction Technique. The selected features are given as an input to the classification algorithm to decide about the condition of the tool. Feature selection has been carried out using J48 Decision Tree Algorithm. Classifications of tool conditions were carried out using Machine Learning Algorithms namely J48 Decision Tree algorithm and Artificial Neural Network (ANN). From the analysis, it is found that ANN is producing comparatively better results. The methodology adopted in this study will be useful for online tool condition monitoring.

More »»

2015

K. K. George, Dr. Santhosh Kumar C., Sreekumar, K. T., K Das, A., Thottupattu, A. J., Kumar, M. S., and Dr. K. I. Ramachandran, “Amrita SRE Database: A Database for Evaluating Speaker Recognition Systems with Mimicked Speech”, 2015.[Abstract]


Evaluating speaker recognition systems under challenging conditions with mimicked voices of target speakers as non-target trials is very important in making decisions of its effectiveness and deployability in critical applications. The lack of a standard database is one of the bottlenecks in pursuing research in this direction. In this paper, we present the details of a new multilingual speaker recognition evaluation database, Amrita SRE database, in which the impersonators mimic the voices of target speakers. Our database consists of 115 target speakers and 76 impersonators speaking in six different languages: English, Hindi, Malayalam, Kannada, Tamil or Telugu, makes a total of 815 target speaker models, 6994 target trials and 3976 non-target trials. We evaluate the performance of different state-of-the-art speaker recognition systems using the normal target and nontarget trials and compared its performance with non-target trials replaced with mimicked voices. For all the systems, an average performance deterioration of 10% absolute in equal error rate (EER) was observed when tested with mimicked non-target utterances.

More »»

2015

R. Gopinath, Dr. Santhosh Kumar C., Vishnuprasad, K., and Dr. K. I. Ramachandran, “Feature Mapping Techniques for Improving the Performance of Fault Diagnosis of Synchronous Generator”, International Journal of Prognostics and Health Management, vol. 6, 2015.[Abstract]


Support vector machine (SVM) is a popular machine learning algorithm used extensively in machine fault diagnosis. In this paper, linear, radial basis function (RBF), polynomial, and sigmoid kernels are experimented to diagnose inter-turn faults in a 3kVA synchronous generator. From the preliminary results, it is observed that the performance of the baseline system is not satisfactory since the statistical features are nonlinear and does not match to the kernels used. In this work, the features are linearized to a higher dimensional space to improve the performance of fault diagnosis system for a synchronous generator using feature mapping techniques, sparse coding and locality constrained linear coding (LLC). Experiments and results show that LLC is superior to sparse coding for improving the performance of fault diagnosis of a synchronous generator. For the balanced data set, LLC improves the overall fault identification accuracy of the baseline RBF system by 22.56%, 18.43% and 17.05% for the R, Y and Bphase faults respectively.

More »»

2015

Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, and A.A., K., “Vital sign normalisation for improving performance of multi-parameter patient monitors”, Electronics Letters, vol. 51, pp. 2089-2090, 2015.[Abstract]


Using covariance normalisation (CVN) of vital signs is explored to improve the performance of multi-parameter patient monitors with heart rate, arterial blood pressure, respiration rate, and oxygen saturation (SpO2) as its input. The baseline system for the experiments is a support vector machine classifier with a radial basis function kernel. Although an improvement in the overall classification accuracy with the use of CVN is obtained, there was a deterioration in sensitivity. Furthermore, it is noted that the estimate of the covariance is often noisy, and therefore the covariance estimates is smoothed to obtain a performance improvement of 0.23% absolute for sensitivity, 1.34% absolute for specificity, and 1.08% absolute for the overall classification accuracy. Multi-parameter intelligent monitoring in intensive care II database for all the experiments is used.

More »»

2015

K. Ka George, Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, and Panda, Ab, “Cosine distance features for improved speaker verification”, Electronics Letters, vol. 51, pp. 939-941, 2015.[Abstract]


Similarities are used with people known already as a means to enhance speaker verification accuracy. Motivated by this, experimental work has been conducted regarding the use of cosine distance (CD) similarity with respect to a set of reference speakers, CD features, with a back-end support vector machine (CDF-SVM) classifier for speaker verification. A state-of-the-art i-vector with CD scoring (i-CDS) is used as the baseline system for the experiments and for the computation of CD similarity. Experimental results on the telephone speech of the core short2-short3 conditions of NIST 2008 speaker recognition evaluation (SRE), for female, male and both-gender trials, show that the proposed CDF-SVM outperforms the baseline i-CDS system. The CDF-SVM achieved an absolute improvement of 1.16% in equal error rate (EER) and 0.38% in minimum DCF over the baseline i-CDS for female trials. Similar performance improvements were also obtained for the male and all-gender trials of the SRE. Finally, fusing the CDF-SVM with i-CDS gave the best overall performance, an absolute improvement of 4.19% in EER and 1.99% in minimum DCF, over the individual CDF-SVM system performance for the all-gender trials. Similar performance improvements were also achieved for male and female trials. © The Institution of Engineering and Technology 2015.

More »»

2014

V. Vaijeyanthi, Vishnuprasad, K., Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, Gopinath, R., A Kumar, A., and Yadav, P. Kumar, “Towards enhancing the performance of multi-parameter patient monitors”, Healthcare technology letters, vol. 1, p. 19, 2014.[Abstract]


Multi-parameter patient monitors (MPMs) have become increasingly important in providing quality healthcare to patients. It is well known in the medical community that there exists an intrinsic relationship between different vital parameters in a healthy person, these include heart rate, blood pressure, respiration rate and oxygen saturation. For example, an increase in blood pressure would lead to a decrease in the heart rate, and vice versa. Although it is likely to improve the performance of MPM systems, this fact is not explored in engineering research. In this work, experiments show that deriving additional features to capture the intrinsic relationship between the vital parameters, the alarm accuracy (sensitivity), no-alarm accuracy (specificity) and the overall performance of MPMs can be improved. The geometric mean of the product of all the vital parameters taken in pairs of two was used to capture the intrinsic relationship between the different parameters. An improvement of 10.55% for sensitivity, 0.32% for specificity and an overall performance improvement of 1.03% was obtained, compared to the baseline system using classification and regression tree with the four vital parameters.

More »»

2014

T. Praveenkumar, Dr. Saimurugan M., Krishna Kumar P., and Dr. K. I. Ramachandran, “Fault Diagnosis of Automobile Gearbox Based on Machine Learning Techniques”, Procedia Engineering, vol. 97, pp. 2092–2098, 2014.[Abstract]


Gearbox is an essential device employed in industries to vary speed and load conditions according to the requirements. More advancement in its design and operation leads to increase in industrial applications. The failure in any of the components of gearbox can lead to production loss and increase maintenance cost. The component failure has to be detected earlier to avoid unexpected breakdown. Vibration measurements are used to monitor the condition of the machine for predictive maintenance and to predict the gearbox faults successfully. This paper addresses the use of vibration signal for automated fault diagnosis of gearbox. In the experimental studies, good gears and face wear gears are used to collect vibration signals for good and faulty conditions of the gearbox. Each gear is tested with two different speeds and loading conditions. The statistical features are extracted from the acquired vibration signals. The extracted features are given as an input to the support vector machine (SVM) for fault identification. The Performance of the fault identification system using vibration signals are discussed and compared.

More »»

2014

Dr. Saimurugan M. and Dr. K. I. Ramachandran, “A comparative study of sound and vibration signals in detection of rotating machine faults using support vector machine and independent component analysis”, International Journal of Data Analysis Techniques and Strategies, vol. 6, pp. 188–204, 2014.[Abstract]


M. Saimurugan obtained his BE in Mechanical Engineering at Kongu Engineering College, Erode under Bharathiar University, Coimbatore in 1998. He completed his ME in Computer Aided Design at Government College of Engineering, Salem under Periyar University, Salem in 2000. Then, he started his career as a Lecturer at Amrita Institute of Technology, Coimbatore. He has published one international journal paper and four international conference papers. Currently, he is working as an Assistant Professor at Amrita Vishwa Vidyapeetham, Coimbatore and is doing his PhD on vibration and sound-based fault diagnosis of rotating machines.

More »»

2014

V. N. Varghees and Dr. K. I. Ramachandran, “A novel heart sound activity detection framework for automated heart sound analysis”, Biomedical Signal Processing and Control, vol. 13, pp. 174-188, 2014.[Abstract]


In automated heart sound analysis and diagnosis, a set of clinically valued parameters including sound intensity, frequency content, timing, duration, shape, systolic and diastolic intervals, the ratio of the first heart sound amplitude to second heart sound amplitude (S1/S2), and the ratio of diastolic to systolic duration (D/S) is measured from the PCG signal. The quality of the clinical feature parameters highly rely on accurate determination of boundaries of the acoustic events (heart sounds S1, S2, S3, S4 and murmurs) and the systolic/diastolic pause period in the PCG signal. Therefore, in this paper, we propose a new automated robust heart sound activity detection (HSAD) method based on the total variation filtering, Shannon entropy envelope computation, instantaneous phase based boundary determination, and boundary location adjustment. The proposed HSAD method is validated using different clean and noisy pathological and non-pathological PCG signals. Experiments on a large PCG database show that the HSAD method achieves an average sensitivity (Se) of 99.43% and positive predictivity (+P) of 93.56%. The HSAD method accurately determines boundaries of major acoustic events of the PCG signal with signal-to-noise ratio of 5 dB. Unlike other existing methods, the proposed HSAD method does not use any search-back algorithms. The proposed HSAD method is a quite straightforward and thus it is suitable for real-time wireless cardiac health monitoring and electronic stethoscope devices. © 2014 Elsevier Ltd.

More »»

2013

J. Selvaraj, Dr. K. I. Ramachandran, and Keshore, D., “Novel approach for energy conservation by raw material preheating in green sand casting”, International Journal of ChemTech Research, vol. 5, pp. 871-879, 2013.[Abstract]


This paper discusses a novel method of mixing metallic raw materials (virgin/scrap) with green sand molds in such a way that the scrap is aligned close to the mold cavity and recovers the waste heat and gets preheated when the molten metal is poured into the mold cavity. When this preheated scrap is loaded into the furnace for the consecutive melting, it is found to take 21 % less energy than it would take to melt normal scrap. This principle has been improvised by insulators, resulting in better heat recovery. In the experiments, it is observed that apart from considerable energy savings, this method also enhances recyclability and conservation of molding sand, and reduced mold emissions, adding to the appropriateness of this method to address the current crisis in the fields of energy and environment. This method has been applied for patent.

More »»

2013

P. K Marimuthu, Krishna Kumar P., K. Ramesh Kumar, and Dr. K. I. Ramachandran, “Finite element simulation of effect of residual stresses during orthogonal machining using ALE approach”, International Journal of Machining and Machinability of Materials, vol. 14, pp. 213–229, 2013.[Abstract]


This paper presents a finite element model that has been developed to predict the effect of residual stress induced in the work material during multiple pass turning of AISI 4340 steel. Chip morphology and force variation during machining are also quantified using the FE model. Finite element model was developed using arbitrary Lagrangian-Eulerian formulation along with Johnson-Cook material model and Johnson-Cook damage model. The finite element model developed in this study was validated experimentally by studying the chip morphologogy and cutting force variation during the machining. Results indicate that there is good correlation existing between numerical results and experimental results.

More »»

2013

V. Vaijeyanthi, Kumar, C. S., Dr. K. I. Ramachandran, Joy, J. K., and Kumar, A. A., “Application-specific fine tuning of multi-parameter patient monitors”, Electronics Letters, vol. 49, pp. 1432-1433, 2013.[Abstract]


Multi-parameter patient monitors (MPMs) have become increasingly important in providing quality health care to patients. A high alarm accuracy (sensitivity) will need a lower threshold for alarm detection which will lead to lower no-alarm accuracy (specificity) and viceversa. MPMs when used in an intensive care unit (ICU) need to have high sensitivity. However they need to have high specificity when used in in-patient wards for regular health check-ups. Proposed is a novel algorithm to trade-off specificity for sensitivity and viceversa depending on the application. The proposed method is referred as detection error trade-off, trade-off specificity for better sensitivity and vice-versa. The algorithm will help to extend the application of MPMs from ICUs to in-patient wards and thus enhance the quality of health care. Experiments have been conducted with an MPM using the classification and regression tree algorithm. By using the proposed algorithm, an improvement of 10.18% in sensitivity was obtained by trading-off 0.40% in specificity. Furthermore, the overall performance of the refined system is 1.15% better than the baseline system.

More »»

2012

B. Devasenapati and Dr. K. I. Ramachandran, “Artificial Intelligence Based Green Technology Retrofit for Misfire Detection in Old Engines”, International Journal of Green Computing, vol. 3, pp. 43-55, 2012.[Abstract]


The core theme of the paper is misfire detection using random forest algorithm and decision tree based machine learning models for emission minimization in gasoline passenger vehicles. The engine block vibration signals are used for misfire detection. The signal is a combination of all vibration emissions of various engine components and also contains the vibration signature due to misfire. The quantum of information available at a given instant is enormous and hence suitable techniques are adopted to reduce the computational load due to redundant information. The random forest algorithm based model and the decision tree model are found to have a consistent high classification accuracy of around 89.7% and 89.3% respectively. From the results obtained the authors conclude that the combination of statistical features and random forest algorithm is suitable for detection of misfire in spark ignition engines and hence contributing to emission minimization in vehicles.

More »»

2012

J. Selvaraj and Dr. K. I. Ramachandran, “Energy conservation in aluminium foundries by waste heat recovery from solidifying molten metal”, International Journal of Energy Technology and Policy, vol. 8, pp. 32-49, 2012.[Abstract]


This paper presents a unique attempt to preheat the raw materials by recovering the waste heat from solidifying molten metal in shape castings of aluminium using sand moulds. Finite element analysis of the problem using ANSYS has also been done, results of which show satisfactory agreement with the experimental results. By using this preheating method, it is proved that around 16% of heat recovery could be realised, which is a substantial achievement in this field. Since energy generation has negative environmental impacts, this attempt is a definite contribution to the solution of global environmental problems such as climate change. Patent has been applied for this work and hence minimum details have been provided. Copyright © 2012 Inderscience Enterprises Ltd.

More »»

2011

C. Arunaa, Devasenapati, B., Dr. K. I. Ramachandran, Vishnuprasad, K., and Surendra, C., “Neural Predictive Controller Based Diesel Injection Management System for Emission Minimisation”, IJGC, vol. 2, pp. 63-82, 2011.[Abstract]


Rapid growth in production of automobiles has increased emissions. Automotive control engineers use innovative control techniques to meet the upcoming emission standards. This paper proposes a novel method of employing artificial neural network (ANN) based predictive controller design. The controller predicts the injection duration based on the inputs from various sensors. The results are then evaluated over a simulated engine model developed in AMESim (Advanced Modelling Environment for Simulation), which ensures a green design process with almost negligible carbon foot print. The results obtained are encouraging and promotes the use of neural predictive control. Implementation of the controller will lead to emission reduction.

More »»

2011

Va Sugumaran and Dr. K. I. Ramachandran, “Effect of number of features on classification of roller bearing faults using SVM and PSVM”, Expert Systems with Applications, vol. 38, pp. 4088-4096, 2011.[Abstract]


Bearings in the machines are the major components of interest for condition monitoring. Their failure causes increase in down time and maintenance cost. A possible solution to the problem is developing an on-line condition monitoring system. The vibration characteristics can be a determining factor that will reveal the condition of the bearing parts. Visual inspection of frequency-domain features of the vibration signals may be sufficient to identify the faults, but it requires large domain knowledge and it is a function of speed. Automatic diagnostic techniques allow relatively unskilled operators to make important decisions. In this context, machine learning algorithms have been successfully used to solve the problem with the help of vibration signals. The machine learning procedure has three important phases: feature extraction, feature selection and feature classification. Feature selection involves identifying the good features that contributes greatly for classification and determining the number of such features. Often researchers overlook the later issue and arbitrarily choose the number of features. As there is no science that will tell the right number of features, for a given problem, an extensive study is needed to find the optimum number of features and this paper presents the results of such a study using SVM and PSVM classifiers for statistical and histogram features of time domain signal. The findings are very interesting and challenging; some useful conclusions were drawn and presented. © 2010 Elsevier Ltd. All rights reserved.

More »»

2011

Dr. Elangovan M., S Devasenapati, B., Dr. Sakthivel N.R., and Dr. K. I. Ramachandran, “Evaluation of expert system for condition monitoring of a single point cutting tool using principle component analysis and decision tree algorithm”, Expert Systems with Applications, vol. 38, no. 4, pp. 4450–4459, 2011.[Abstract]


Tool wear and tool life are the principle areas are focus in any machining activity. The production rate, surface finish of machined component and the machine condition are directly related to the tool condition. This work on tool condition monitoring delves into data mining approach to discover the hidden information available in the tool vibration signals. The use of statistical features derived from the vibration data is used as the primary feature and Principle Component Analysis (PCA) transformed statistical features are evaluated as an alternative. In order to increase the robustness of the classifier and to reduce the data processing load, feature reduction is necessary. The feature reduction using (a) decision tree and (b) feature transformation and reduction using PCA are evaluated independently and the results are compared. The effective combination of feature reducer and classifier for designing the expert system is studied and reported.

More »»

2011

Sa Ravikumar, Dr. K. I. Ramachandran, and Sugumaran, Vb, “Machine learning approach for automated visual inspection of machine components”, Expert Systems with Applications, vol. 38, pp. 3260-3266, 2011.[Abstract]


Visual inspection on the surface of components is a main application of machine vision. Visual inspection finds its application in identifying defects such as scratches, cracks bubbles and measurement of cutting tool wear and welding quality. Machine learning approach to machine vision helps in automating the design process of machine vision systems. This approach involves image acquisition, preprocessing, feature extraction and classification. Study shows a library of features, and classifiers are available to classify the data. However, only the best combination of them can yield the highest classification accuracy. In this study, images with different known conditions were acquired, preprocessed, and histogram features were extracted. The classification accuracies of C4.5 classifier algorithm and Naïve Bayes algorithm were compared, and results are reported. The study shows that C4.5 algorithm performs better. © 2010 Elsevier Ltd. All rights reserved.

More »»

2011

Dr. Saimurugan M., Dr. K. I. Ramachandran, Sugumaran, V., and Dr. Sakthivel N.R., “Multi component fault diagnosis of rotational mechanical system based on decision tree and support vector machine”, Expert Systems with Applications, vol. 38, pp. 3819-3826, 2011.[Abstract]


The shaft and bearing are the most critical components in rotating machinery. Majority of problems arise from faulty bearings in turn affect the shaft. The vibration signals are widely used to determine the condition of machine elements. The vibration signals are used to extract the features to identify the status of a machine. This paper presents the use of c-SVC and nu-SVC models of support vector machine (SVM) with four kernel functions for classification of faults using statistical features extracted from vibration signals under good and faulty conditions of rotational mechanical system. Decision tree algorithm was used to select the prominent features. These features were given as inputs for training and testing the c-SVC and nu-SVC model of SVM and their fault classification accuracies were compared. © 2010 Elsevier Ltd. All rights reserved.

More »»

2011

Va Sugumaran and Dr. K. I. Ramachandran, “Fault diagnosis of roller bearing using fuzzy classifier and histogram features with focus on automatic rule learning”, Expert Systems with Applications, vol. 38, pp. 4901-4907, 2011.[Abstract]


Roller bearing is one of the most widely used elements in rotary machines. Condition monitoring of such elements is conceived as pattern recognition problem. Pattern recognition has three main phases: feature extraction, feature selection and feature classification. Histogram features can be used for fault diagnosis of roller bearing. This paper presents the use of decision tree for selecting best few histogram features (bin ranges) that will discriminate the fault conditions of the bearing from given train samples. These features are extracted from vibration signals. A rule set is formed from the extracted features and fed to a fuzzy classifier. The rule set necessary for building the fuzzy classifier is obtained largely by intuition and domain knowledge. This paper also presents the usage of decision tree to generate the rules automatically from the feature set. The vibration signal from a piezoelectric transducer is captured for the following conditions - good bearing, bearing with inner race fault, bearing with outer race fault, and inner and outer race fault. The histogram features were extracted and good features that discriminate the different fault conditions of the bearing were selected using decision tree. The rule set for fuzzy classifier is obtained by once using the decision tree again. A fuzzy classifier is built and tested with representative data. The results are found to be encouraging. © 2010 Elsevier Ltd. All rights reserved.

More »»

2011

Dr. Elangovan M., Sugumaran, V., Dr. K. I. Ramachandran, and Ravikumar, S., “Effect of SVM kernel functions on classification of vibration signals of a single point cutting tool”, Expert Systems with Applications, vol. 38, no. 12, pp. 15202–15207, 2011.[Abstract]


The studies on tool condition monitoring along with digital signal processing can be used to prevent damages on cutting tools and workpieces when the tool conditions become faulty. These studies have become more relevant in today’s context where the order realization dates are crunched and deadlines are to be met in order to catch up with the competition. Based on a continuous acquisition of signals with sensor systems it is possible to classify certain wear parameters by the extraction of features. Data mining approach is extensively used to probe into structural health of the tool and the process. This paper discusses condition monitoring of carbide tipped tool using Support Vector Machine and compares the classification efficiency between C-SVC and ν-SVC. It further analyses the results with other classifiers like Decision Tree and Naïve Bayes and Bayes Net. The vibration signals are acquired for various tool conditions like tool-good condition, tip-breakage, etc. The effort is to bring out the better features-classifier combination.

More »»

2010

D. S. Babu, Dr. K. I. Ramachandran, and Sugumaran, V., “Misfire Detection in Spark Ignition Engine using Support Vector Machines”, International Journal of Computer Applications, vol. 5, 2010.[Abstract]


To maintain optimum performance throughout the service life of an engine and to exercise a tight control over emissions, misfire detection is a vital activity. The engine block vibration contains valuable hidden information regarding the operating condition of the engine. Misfire can be detected by processing the vibration signals acquired from the engine using an accelerometer. The hidden information in the acquired signal can be analysed using various features extracted from the signals. A comparative performance analysis on classification accuracy of SVM when using statistical and histogram features for misfire detection in a spark ignition engine is presented.

More »»

2010

N. Saravanan and Dr. K. I. Ramachandran, “Incipient gear box fault diagnosis using discrete wavelet transform (DWT) for feature extraction and classification using artificial neural network (ANN)”, Expert Systems with Applications, vol. 37, pp. 4168 - 4181, 2010.[Abstract]


An efficient predictive plan is needed for any industry because it can optimize the resources management and improve the economy plant, by reducing unnecessary costs and increasing the level of safety. A great percentage of breakdowns in the productive processes are caused for gear box, they began its deterioration from early stages, also called incipient level. The extracted features from the DWT are used as inputs in a neural network for classification purposes. The results show that the developed method can reliably diagnose different conditions of the gear box. The wavelet transform is used to represent all possible types of transients in vibration signals generated by faults in a gear box. It is shown that the transform provides a powerful tool for condition monitoring and fault diagnosis. The vibration signal of a spur bevel gear box in different conditions is used to demonstrate the application of various wavelets in feature extraction. In this paper fault diagnostics of spur bevel gear box is treated as a pattern classification problem. The major steps in pattern classification are feature extraction, and classification. This paper investigates the use of discrete wavelets for feature extraction and artificial neural network for classification.

More »»

2010

N. Saravanan, Siddabattuni, V. N. S., and Dr. K. I. Ramachandran, “Fault diagnosis of spur bevel gear box using artificial neural network (ANN), and proximal support vector machine (PSVM)”, Applied Soft Computing, vol. 10, pp. 344–360, 2010.[Abstract]


Vibration signals extracted from rotating parts of machineries carries lot many information with in them about the condition of the operating machine. Further processing of these raw vibration signatures measured at a convenient location of the machine unravels the condition of the component or assembly under study. This paper deals with the effectiveness of wavelet-based features for fault diagnosis of a gear box using artificial neural network (ANN) and proximal support vector machines (PSVM). The statistical feature vectors from Morlet wavelet coefficients are classified using J48 algorithm and the predominant features were fed as input for training and testing ANN and PSVM and their relative efficiency in classifying the faults in the bevel gear box was compared. More »»

2010

B. S Devasenapati, Sugumaran, V., and Dr. K. I. Ramachandran, “Misfire Identification in A Four-Stroke Four-Cylinder Petrol Engine Using Decision Tree”, Expert Systems with Applications, vol. 37, pp. 2150–2160, 2010.[Abstract]


Misfire detection in an internal combustion engine is very crucial to maintain optimum performance throughout its service life and to reduce emissions. The vibration of the engine block contains indirect information regarding the condition of the engine. Misfire detection can be achieved by processing the vibration signals acquired from the engine using a piezoelectric accelerometer. This hidden information can be decoded using statistical parameters like kurtosis, standard deviation, mean, median, etc. This paper illustrates the use of decision tree as a tool for feature selection and feature classification. The effect of dimension, minimum number of objects and confidence factor on classification accuracy are studied and reported in this work. More »»

2010

Dr. Elangovan M., Dr. K. I. Ramachandran, and Sugumaran, V., “Studies on Bayes classifier for condition monitoring of single point carbide tipped tool based on statistical and histogram features”, Expert Systems with Applications, vol. 37, no. 3, pp. 2059–2065, 2010.[Abstract]


Various methods of tool condition monitoring techniques are used to control the tool wear during machining in CNC machine tools. Based on a continuous acquisition of signals with sensor systems it is possible to classify certain wear parameters by the extraction of features. Data mining approach is used to probe into the structural information hidden in the signals acquired. This paper discusses machine tool condition monitoring of carbide tipped tool by using Naïve Bayes and Bayes Net classifiers and compares the results of histogram features with the statistical features to establish better classification among the two. The vibration signals are acquired for various tool conditions like tool-good condition, tip-breakage, etc. The effort is to bring out the better feature–classifier combine. The results are discussed.

More »»

2009

N. Saravanan and Dr. K. I. Ramachandran, “Fault diagnosis of spur bevel gear box using discrete wavelet features and Decision Tree classification”, Expert Systems with Applications, vol. 36, pp. 9564–9573, 2009.[Abstract]


The wavelet transform (WT) is used to represent all possible types of transients in vibration signals generated by faults in a gear box. It is shown that the transform provides a powerful tool for condition monitoring and fault diagnosis. The vibration signal of a spur bevel gear box in different conditions is used to demonstrate the application of various wavelets in feature extraction. In present work, a discrete wavelet, Daubechies wavelets (db1–db15) is used for feature extraction and their relative effectiveness in feature extraction is compared. The major steps in pattern classification are feature extraction and classification. This paper investigates the use of discrete wavelets for feature extraction and a Decision Tree for classification. J48 Decision Tree algorithm has been used for feature selection as well as for classification. This paper illustrates the powerfulness and flexibility of the discrete wavelet transform to decompose linear and non-linear processing of vibration signal. More »»

2009

N. Saravanan and Dr. K. I. Ramachandran, “A case study on classification of features by fast single-shot multiclass PSVM using Morlet wavelet for fault diagnosis of spur bevel gear box”, Expert Systems with Applications, vol. 36, pp. 10854–10862, 2009.[Abstract]


This paper deals with the application of fast single-shot multiclass proximal support vector machine for fault diagnosis of a gear box consisting of twenty four classes. The condition of an inaccessible gear in an operating machine can be monitored using the vibration signal of the machine measured at some convenient location and further processed to unravel the significance of these signals. The statistical feature vectors from Morlet wavelet coefficients are classified using J48 algorithm and the predominant features were fed as input for training and testing multiclass proximal support vector machine. The efficiency and time consumption in classifying the twenty four classes all-at-once is reported. More »»

2009

V. Sugumaran and Dr. K. I. Ramachandran, “Wavelet Selection Using Decision Tree for Fault Diagnosis of Roller Bearings”, International Journal of Applied Engineering Research, vol. 4, pp. 201-225, 2009.[Abstract]


Fault diagnosis of the roller bearings as pattern classification problem has three main steps: feature extraction, feature selection and classification. Wavelets have been widely used for feature extraction from vibration signals. Identifying a suitable wavelet for a given application is a challenging task in the whole process. This paper investigates the use of decision tree for selecting apt wavelet for fault diagnosis of roller bearings with discrete wavelet transform features. The study is done on vibration signals of roller bearings from different fault conditions. The faults considered in this study are bearings with inner race fault, bearings with outer race fault and bearings with both of them. The decision tree has been used for feature selection as well as for classification. Many commonly used wavelets families have been considered in this study and their classification accuracies were compared. More »»

2009

N. Saravanan, Siddabattuni, V. N. S., and Dr. K. I. Ramachandran, “Static and Dynamic Analysis of Asymmetric Bevel Gears using Finite Element Method.”, International Journal of Applied Engineering Research, vol. 4, pp. 645-664, 2009.[Abstract]


The aim of asymmetric tooth is to improve the performance of gears such as increasing the load capacity or reducing noise and vibration. Application of asymmetric tooth side surfaces is able to increase the load capacity and durability for the drive tooth side. The tooth form has left-right symmetry in the involute cylindrical gear, and the same performance can be obtained at forward and backward rotation. However, both the forward and backward rotations are not always expected in the practically used gear units for power transmission. Therefore, two sides of the gear tooth are functionally different for most gears. Even if one side (drive side) is significantly loaded for longer periods, the opposite side (coast side) is unloaded or slightly loaded for short duration only. In several papers , the higher pressure angle profile for the drive side and lower pressure angle profile for the coast side have been considered. This kind of application makes it possible for the gear to reduce the bending stress. The asymmetric involute tooth can be manufactured by the same process as in generating the symmetric involute tooth. Asymmetric profile is achievable by adopting the different pressure angle values of coast side and drive side of the bevel gear for the two sides of the rack. Depending on the special tooling, production cost of these gears increases. Therefore, the gears with asymmetric teeth should be considered for gear systems that require extreme performance like aerospace applications and for mass production, where the share of the tooling cost per one gear is insignificant. The most promising application of asymmetric profiles seems to be in molded gears and powder gears. In this study, asymmetric spur bevel gear with higher drive side pressure angle than coast side pressure angle has been considered. The purpose of this study is to determine bending load carrying capacity and the dynamic characteristics of asymmetric bevel gear. More »»

2009

N. Saravanan, Cholairajan, S., and Dr. K. I. Ramachandran, “Vibration-based fault diagnosis of spur bevel gear box using fuzzy technique”, Expert Systems with Applications, vol. 36, no. 2, pp. 3119–3135, 2009.[Abstract]


To determine the condition of an inaccessible gear in an operating machine the vibration signal of the machine can be continuously monitored by placing a sensor close to the source of the vibrations. These signals can be further processed to extract the features and identify the status of the machine. The vibration signal acquired from the operating machine has been used to effectively diagnose the condition of inaccessible moving components inside the machine. Suitable sensors are kept at various locations to pick up the signals produced by machinery and these signals are very meaningful in condition diagnosis surveillance. To determine the important characteristics and to unravel the significance of these signals, further analysis or processing is required. This paper presents the use of decision tree for selecting best statistical features that will discriminate the fault conditions of the gear box from the signals extracted. These features are extracted from vibration signals. A rule set is formed from the extracted features and fed to a fuzzy classifier. The rule set necessary for building the fuzzy classifier is obtained largely by intuition and domain knowledge. This paper also presents the usage of decision tree to generate the rules automatically from the feature set. The vibration signal from a piezo-electric transducer is captured for the following conditions – good bevel gear, bevel gear with tooth breakage (GTB), bevel gear with crack at root of the tooth (GTC), and bevel gear with face wear of the teeth (TFW) for various loading and lubrication conditions. The statistical features were extracted and good features that discriminate the different fault conditions of the gearbox were selected using decision tree. The rule set for fuzzy classifier is obtained by once using the decision tree again. A fuzzy classifier is built and tested with representative data. The results are found to be encouraging. More »»

2008

V. Sugumaran, Sabareesh, G. R., and Dr. K. I. Ramachandran, “Fault diagnostics of roller bearing using kernel based neighborhood score multi-class support vector machine”, Expert Systems with Applications, vol. 34, pp. 3090–3098, 2008.[Abstract]


Roller bearing is one of the most widely used rotary elements in a rotary machine. The roller bearing’s nature of vibration reveals its condition and the features that show the nature are to be extracted through some indirect means. Statistical parameters like kurtosis, standard deviation, maximum value, etc. form a set of features, which are widely used in fault diagnostics. Finding out good features that discriminate the different fault conditions of the bearing is often a problem. Selection of good features is an important phase in pattern recognition and requires detailed domain knowledge. This paper addresses the feature selection process using decision tree and uses kernel based neighborhood score multi-class support vector machine (MSVM) for classification. The vibration signal from a piezoelectric transducer is captured for the following conditions: good bearing, bearing with inner race fault, bearing with outer race fault, and inner and outer race faults. The statistical features are extracted therefrom and classified successfully using MSVM. The results of MSVM are compared with and binary support vector machine (SVM). More »»

2008

N. Saravanan, Siddabattuni, V. N. S. Kumar, and Dr. K. I. Ramachandran, “A comparative study on classification of features by SVM and PSVM extracted using Morlet wavelet for fault diagnosis of spur bevel gear box”, Expert systems with applications, vol. 35, pp. 1351–1366, 2008.[Abstract]


The condition of an inaccessible gear in an operating machine can be monitored using the vibration signal of the machine measured at some convenient location and further processed to unravel the significance of these signals. This paper deals with the effectiveness of wavelet-based features for fault diagnosis using support vector machines (SVM) and proximal support vector machines (PSVM). The statistical feature vectors from Morlet wavelet coefficients are classified using J48 algorithm and the predominant features were fed as input for training and testing SVM and PSVM and their relative efficiency in classifying the faults in the bevel gear box was compared. More »»

2007

V. Sugumaran and Dr. K. I. Ramachandran, “Automatic rule learning using decision tree for fuzzy classifier in fault diagnosis of roller bearing”, Mechanical Systems and Signal Processing, vol. 21, pp. 2237–2247, 2007.[Abstract]


Roller bearing is one of the most widely used elements in rotary machines. Condition monitoring of such elements is conceived as pattern recognition problem. Pattern recognition has two main phases: feature extraction and feature classification. Statistical features like minimum value, standard error and kurtosis, etc. are widely used as features in fault diagnostics. These features are extracted from vibration signals. A rule set is formed from the extracted features and input to a fuzzy classifier. The rule set necessary for building the fuzzy classifier is obtained largely by intuition and domain knowledge. This paper presents the use of decision tree to generate the rules automatically from the feature set. The vibration signal from a piezo-electric transducer is captured for the following conditions—good bearing, bearing with inner race fault, bearing with outer race fault, and inner and outer race fault. The statistical features are extracted and good features that discriminate the different fault conditions of the bearing are selected using decision tree. The rule set for fuzzy classifier is obtained once again by using the decision tree. A fuzzy classifier is built and tested with representative data. The results are found to be encouraging. More »»

2007

V. Sugumaran, Muralidharan, V., and Dr. K. I. Ramachandran, “Feature selection using decision tree and classification through proximal support vector machine for fault diagnostics of roller bearing”, Mechanical Systems and Signal Processing, vol. 21, pp. 930–942, 2007.[Abstract]


Roller bearing is one of the most widely used rotary elements in a rotary machine. The roller bearing's nature of vibration reveals its condition and the features that show the nature, are to be extracted through some indirect means. Statistical parameters like kurtosis, standard deviation, maximum value, etc. form a set of features, which are widely used in fault diagnostics. Often the problem is, finding out good features that discriminate the different fault conditions of the bearing. Selection of good features is an important phase in pattern recognition and requires detailed domain knowledge. This paper illustrates the use of a Decision Tree that identifies the best features from a given set of samples for the purpose of classification. It uses Proximal Support Vector Machine (PSVM), which has the capability to efficiently classify the faults using statistical features. The vibration signal from a piezoelectric transducer is captured for the following conditions: good bearing, bearing with inner race fault, bearing with outer race fault, and inner and outer race fault. The statistical features are extracted therefrom and classified successfully using PSVM and SVM. The results of PSVM and SVM are compared. More »»

Publication Type: Conference Paper

Year of Publication Title

2018

S. Anagha, Suyampulingam, A., and Dr. K. I. Ramachandran, “A Better Digital Filtering Technique for Estimation of SPO2 and Heart Rate from PPG Signals”, in 2018 International Conference on Inventive Research in Computing Applications (ICIRCA), 2018.

2018

R. Nair, Dr. K. I. Ramachandran, S., N. S. R. Basav, and S.R., N., “Design and optimization of automotive energy absorber structure with functionally graded material”, in Materials Today: Proceedings, 2018, vol. 5, pp. 25640-25648.[Abstract]


An automotive energy absorber is a major safety component of a vehicle during head - on and rear collisions. The beam-structure should withstand and absorb the impact energy; thus preventing energy transfer. Design and analysis are done to reduce the impact resistance in vehicles like tractors which has low speed. This study includes the selection of functionally graded polyurethane (FGPU) and optimizing design thicknesses. Two grades of polyurethane (PU) material are tested for uniaxial compression using a universal testing machine (UTM), and the stress-strain plots are obtained. A hyperelastic constitutive material model is applied to perform the explicit dynamic analysis on the beam-structure. The objective is to maximize the stresses in the FGPU, until fracture. Dynamic analysis is performed using ANSYS-WORKBENCH 15.0, while design optimization is carried out using MINITAB 17 and MATLAB 16. © 2018 Elsevier Ltd.

More »»

2018

V. S. K. P. Varma, Adarsh, S., Dr. K. I. Ramachandran, and Nair, B. B., “Real time detection of speed hump/bump and distance estimation with deep learning using GPU and ZED stereo camera”, in Procedia Computer Science, 2018, vol. 143, pp. 988-997.[Abstract]


Most of the humps in India are not being constructed and maintained according to the public safety guidelines of Indian Road Congress (IRC) i.e., IRC099, which is resulting in damage to the vehicles, severe discomfort to the driver and even causing loss of direction control which is leading to fatalities. Very few methods were discussed in literature for un-marked speed hump/bump detection. We propose a method that detects and informs the driver about the upcoming un-marked and marked speed hump/bump in real time using deep learning techniques and gives the distance the vehicle is away from it using stereo-vision approaches. We have achieved using NVIDIA GPU and Stereolabs ZED Stereo camera hardware. With this driver or autonomous mode of the vehicle can control the vehicle speeds to be at safer limits in order to not cause any kind of discomfort to the passengers as well as damage to the vehicle. © 2018 The Authors. Published by Elsevier B.V.

More »»

2018

S. M. Sunil and Dr. K. I. Ramachandran, “Myoelectric Control System Based on Wavelet Features”, in Proceedings of the 4th International Conference on Biosignals, Images and Instrumentation, ICBSII 2018, 2018.[Abstract]


Electromyography (EMG) finds enormous applications in clinical/biomedical, prosthesis and rehabilitation devices. The main objective of this paper is to develop a cost-effective implementation of a prosthetic control system based on EMG signals. Non-invasive surface electrodes are used to acquire the signal for various actions. Since the signal is highly contaminated with noise, they are not used in its raw form to handle any sort of device. Amplification and filtering are therefore inevitable and becomes the foremost task prior to further processing so as to obtain a high-quality signal. After the conditioning of the signal, multi-level decomposition based on wavelet transform is performed and features are extracted from all the levels. They are then reduced to find the optimal performance. Finally, the selected features are able to distinguish between various hand movements and therefore helps in the recognition of the intended motion of the amputee. © 2018 IEEE.

More »»

2018

J. Sankar, Adarsh, S., and Dr. K. I. Ramachandran, “Performance Evaluation of Ultrasonic and Infrared Waves on Human Body and Metal Surfaces for Mobile Robot Navigation”, in Materials Today: Proceedings, 2018, vol. 5, pp. 16516-16525.[Abstract]


Ultrasonic and Infrared sensors are widely used for environment perception in most of the robotic navigation systems. In this paper, the behaviour of Ultrasonic and Infrared waves on human body and metal surfaces is carried out. The sensors were integrated in a mobile robot platform for acquiring the distance data. The sensed data was analysed both statistically as well as computationally to provide meaningful interpretations of the sensors responses on those surfaces. This analysis helps to design suitable algorithms based on the performance ranges of individual sensors. Fuzzy logic based analysis (both Type-1 and Type-2) was performed over the sensor data for the possible sensor data fusion. © 2017 Elsevier Ltd.

More »»

2018

C. C. Mahesh and Dr. K. I. Ramachandran, “Finite element modelling of functionally graded elastomers for the application of diabetic footwear”, in Materials Today: Proceedings, 2018, vol. 5, pp. 16367-16377.[Abstract]


Diabetic foot ulcers are major components of diabetic foot. Uncontrolled sugar levels can lead to this complication. During this complication, skin begins to peel off thus exposing the layers underneath. Big toes, heels and pads of the feet are the common places this could occur and can affect deep to the bones. In this paper, the complexities due to foot ulcers are analyzed. The treatments include, removing pressure from the wound in order to prevent pain. The aim of this paper is to provide a footwear made of functionally graded polyurethane in order to impart comfort for the patient. The concept of functionally graded polyurethane(FGPU) is to make a composite material by varying the properties from one material to another with a specific gradient. Uniaxial compression test is performed on five different densities of polyurethane foam and respective stress strain curves are obtained. The polymers are then stacked one above the other based on hardness starting from high to low. The behavior of this material is studied by applying appropriate hyperfoam material model. 3D model of the sole of the footwear was created in ANSYS and proper material properties are assigned to respective layers. According to patient's regular activity, FEM analysis is performed. The distribution of stresses vs time due to weight of the person are observed and further optimized if needed. © 2017 Elsevier Ltd.

More »»

2018

V. Aparna, Sarath, T. V., and Dr. K. I. Ramachandran, “Simulation model for anemia detection using RBC counting algorithms and Watershed transform”, in 2017 International Conference on Intelligent Computing, Instrumentation and Control Technologies, ICICICT 2017, 2018, vol. 2018-January, pp. 284-291.[Abstract]


Digital image processing is nowadays widely used in biomedical applications. Counting of the red blood cells from the blood smear images can help in detecting anemia. Since the manual location, identification and counting of red blood cells is a tedious, error prone and time consuming there is a rising need for automating the entire process. A simulation model for anemia detection using RBC counting algorithm is presented in this paper. Both Circular Hough Transform and Connected Component Labelling are implemented for counting the number of RBCs and the results are compared. Watershed transform to separate overlapping blood cells is also presented. Error analysis before and after Watershed transform and parameters like segmentation accuracy, sensitivity and specificity are also included in this paper. © 2017 IEEE.

More »»

2018

R. K. Krishna and Dr. K. I. Ramachandran, “Machinery Bearing Fault Diagnosis Using Variational Mode Decomposition and Support Vector Machine as a Classifier”, in IOP Conference Series: Materials Science and Engineering, 2018, vol. 310.[Abstract]


Crack propagation is a major cause of failure in rotating machines. It adversely affects the productivity, safety, and the machining quality. Hence, detecting the crack's severity accurately is imperative for the predictive maintenance of such machines. Fault diagnosis is an established concept in identifying the faults, for observing the non-linear behaviour of the vibration signals at various operating conditions. In this work, we find the classification efficiencies for both original and the reconstructed vibrational signals. The reconstructed signals are obtained using Variational Mode Decomposition (VMD), by splitting the original signal into three intrinsic mode functional components and framing them accordingly. Feature extraction, feature selection and feature classification are the three phases in obtaining the classification efficiencies. All the statistical features from the original signals and reconstructed signals are found out in feature extraction process individually. A few statistical parameters are selected in feature selection process and are classified using the SVM classifier. The obtained results show the best parameters and appropriate kernel in SVM classifier for detecting the faults in bearings. Hence, we conclude that better results were obtained by VMD and SVM process over normal process using SVM. This is owing to denoising and filtering the raw vibrational signals. © Published under licence by IOP Publishing Ltd.

More »»

2017

A. Tamrakar, Adasrh, S., and Dr. K. I. Ramachandran, “Fuzzy logic based fault detection of controller area network using microautobox - II”, in 2017 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2017, 2017, vol. 2017-January, pp. 1290-1295.[Abstract]


In automotive industry the role of electronic system becomes phenomenal. The complexity of the system increases to a larger extend now a days, and hence the diagnostics is also very much essential. The diagnostics can be mainly done in sensors/actuators, processors and on communication medium. Here, we have presented a fault detection system using fuzzy logic for the CAN (Controller Area Network). Whenever a fault occurred in the CAN, a Diagnostic Trouble Code (DTC) corresponding to that fault is recorded in an ECU (Electronic Controlled Unit) which is capable of detecting distinct faults only. In this paper we have proposed a prototype of an ECU using MicroAutoBox ' II hardware, with a CAN interface to detect faults induced into the network. Fuzzy engine within the ECU is capable of identifying the faults based on the number of error frames (Ef), ranges from 0-80 as well as change in differential bus resistance (Rdiff(total)). The validation is done based on the rule created in the fuzzy system. © 2017 IEEE.

More »»

2017

K. V. Geedhu, Dr. K. I. Ramachandran, and Adarsh, S., “CANFIS based robotic navigation”, in 2017 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2017, 2017, vol. 2017-January, pp. 1660-1664.[Abstract]


This paper aims at developing a sensor-based Co-Active adaptive neuro-fuzzy inference system (CANFIS) for solving navigation problems of the mobile robot in an uncertain environment. The infrared sensor reads the distances of right, front and left obstacle. The collision-free path is accomplished by CANFIS controller which selects the desired steering angle by construing the obstacle distance information measured by the infrared sensor. The simulation of CANFIS based algorithm provides more precise steering angle, which implements the navigation task securely and efficiently in an environment populated with static obstacles. © 2017 IEEE.

More »»

2017

M. Bhadane and Dr. K. I. Ramachandran, “Bearing fault identification and classification with convolutional neural network”, in Proceedings of IEEE International Conference on Circuit, Power and Computing Technologies, ICCPCT 2017, 2017.[Abstract]


Condition-based monitoring (CBM) is widely used methodology for the fault diagnosis, which provides the analysis for the safe and proper operations of any device or element. Vibration analysis is most accurate and reliable technique of CBM to reveal the condition of device or element. In this technique, fault can be diagnosed by analysing the vibration data acquired from accelerometer. Convolutional Neural Network (CNN) has emerged as one of the most widely used methodology in application of pattern recognition and acoustic data analysis. In this paper, CNN is used as back-end classifier for bearing fault detection. Vibration data is collected for three different conditions of bearings i.e. normal condition, inner race fault and outer race fault. Statistical features are extracted from vibration data and used as input to CNN classifier. Convolution filters are learned by training CNN and are used to detect the unique features for each condition of bearing. The obtained accuracy shows that CNN is very reliable and effective technique for bearing fault diagnosis. It exhibits good performance compared to peer algorithms. © 2017 IEEE.

More »»

2017

P. Gandhi, Adarsh, S., and Dr. K. I. Ramachandran, “Performance Analysis of Half Car Suspension Model with 4 DOF using PID, LQR, FUZZY and ANFIS Controllers”, in Procedia Computer Science, 2017, vol. 115, pp. 2-13.[Abstract]


The suspension system helps to enhance the ride quality, steering stability, passenger comfort and NVH. In this paper, using a half car active suspension model with 4 Degrees Of Freedom (4 DOF) the controllers such as Proportional Integral Derivative, Linear Quadratic Regulator, Fuzzy and Adaptive Neuro Fuzzy Inference System (ANFIS) are designed using MATLAB-Simulink. The response of these controllers has been analysed using the random road profile (ISO 8608) against the conventional passive suspension system. The results indicate that ANFIS based controller performs better on the parameters 'Settling Time' and 'Amplitude' of the road disturbances, compared with other controllers. © 2017 The Author(s).

More »»

2017

K. Sharmila, Sarath, T. V., and Dr. K. I. Ramachandran, “EMG controlled low cost prosthetic arm”, in 2016 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics, DISCOVER 2016 - Proceedings, 2017, pp. 169-172.[Abstract]


Electromyography (EMG) signals have been extensively used as a control signal in robotics, rehabilitation and health care. In this paper, cost effective design of prosthetic hand using EMG control is presented. Signal amplification and filtering is the primary step in surface EMG signal processing and application systems. Quality of the acquired EMG signal depends on the amplifiers and filters employed. Single channel continuous EMG signal has been acquired from the users arm for various hand movements. The acquired signal is passed through various stages of filters and amplifiers for amplification and noise reduction. The conditioned analog signal is converted into digital samples. After the signal acquisition process, features are extracted from the acquired signal and the extracted features are reduced to minimize the number of computations. These reduced feature parameters are used to classify the signal for different hand movements. Once the classifier identifies the intended motion, the control signal will be generated and given to the motors in the prosthetic hand to perform the intended movements. Experiments were done to find the efficiency of the developed system and it is found that this system can give basic movements at a very low cost. © 2016 IEEE.

More »»

2017

R. J. Gini, Dr. K. I. Ramachandran, and Ceerthibala, U. K., “Approach to extract twin fECG for different cardiac conditions during prenatal”, in IFMBE Proceedings, 2017, vol. 61, pp. 104-108.[Abstract]


During multiple fetus pregnancy, degree of risk for distinguishing the information of mother and fetus health condition is high. A proper distinguishable ECG of each fetus and mother gives information about the health conditions of individuals. In case of multiple fetal conditions, the heartbeat of the fetuses will be almost at the same rate. This algorithm has been aimed to separate mECG and the fECGs of the individual fetus. First, the signal for different medical conditions like Fibrillation, Apnea, Ventricular Ectopy, Singleton and Normal has been considered. The synthetic abdECG signal for the above mentioned cases has been formulated by preprocessing and considered as the input signal. RPeak of mECG in the abdECG signal has been located using First Order Gaussian Differentiator and Zero Crossing Detector. QRS complex has been considered around the identified R-Peak of abdECG. Identified QRS has been removed from the abdECG signal to obtain fECG with residual noise. The QRS complexes of fECG are detected the same way as mECG QRS were detected, and is represented as binary signals. The separation of the fetal ECG is done based on the individual presence of the fetus in the signals using Inter-beat averaging and Inter-beat standard deviation of the binary signal. The algorithm has been tested for above mentioned cardiac conditions during prenatal. The algorithm has been able to achieve 99% accuracy for particular cardiac condition with overall system accuracy of 80.4%. The standard cardiac signals of different cases have been sourced from Physionet database to construct the abdECG. © Springer Nature Singapore Pte Ltd. 2017.

More »»

2017

V. N. Varghees and Dr. K. I. Ramachandran, “Two-channel heart sound segmentation framework using phonocardiogram and pulsatile signals”, in 2016 IEEE Students' Technology Symposium, TechSym 2016, 2017, pp. 305-310.[Abstract]


Phonocardiogram (PCG) segmentation is the crucial first step in automated heart sound analysis and diagnostic systems. Recently, the cardiac signals (including, electrocardiogram, phonocardiogram and photoplethysmogram) are simultaneously recorded for most cardiac signal processing applications such as cardiovascular diagnostic system, biometric authentication, and emotion/stress recognition. In this paper, we present an effective two-channel heart sound segmentation framework using PCG and pulse signals. The proposed framework comprises the steps of: heart sound signal decomposition using stationary wavelet transform, Shannon entropy envelope extraction, heart sound endpoint determination, systolic peak detection, and heart sound discrimination. The proposed framework is tested and validated using the simultaneously recorded heart sound and pulse signals. Performance evaluation results demonstrate that the proposed heart sound endpoint and systolic peak detection methods can achieves an average Se of 98.98%, +P of 96.80% and Se of 99.57%, +P of 99.37%, respectively. The proposed framework achieves an identification accuracy of 100% in distinguishing the first heart sound (S1) and second heart sound (S2) under clean and noisy signal conditions. © 2016 IEEE.

More »»

2016

R. Anusree, Ranjith, R., and Dr. K. I. Ramachandran, “A noncontact vital sign monitoring algorithm using a camera”, in 2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI), Jaipur, India, 2016.[Abstract]


Currently contact probes are widely used to determine the vital signs. As the medical technology develops, non-invasive method becomes more feasible. We are using a new non-contact method of determining pulse rate (PR) commonly known as video plethysmography technique. In this method, face video is captured using a web camera and from the captured video frames, photoplethysmogram (PPG) signals are obtained from different regions of the face based on the pixel intensity variation. This PPG signals are preprocessed by filtering and noise compensated using dynamic weight factoring based on power spectral density and from this finally obtained PPG signal, we are determining the pulse rate by converting it into Fast Fourier Transform (FFT). In this paper, we present an algorithm for pulse rate determination which can be easily implemented on portable device or on standard PC. This method of noncontact detection and monitoring of cardiac pulse is more useful in both hospitals and telemedicine.

More »»

2016

K. K. George, Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, and Panda, A., “Improving Robustness of Speaker Verification Against Mimicked Speech”, in Odyssey 2016, 2016.[Abstract]


Making speaker verification (SV) systems robust to spoofed/mimicked speech attacks is very important to make its use effective in security applications. In this work, we show that using a proximal support vector machine backend classifier with i-vectors as inputs (i-PSVM) can help improve the performance of SV systems for mimicked speech as non-target trials. We compared our results with the state-of-the-art baseline i-vector with cosine distance scoring (i-CDS), i-vector with a backend SVM classifier (i-SVM) and cosine distance features with an SVM backend classifier (CDF-SVM) systems. In iPSVM, proximity of the test utterance to the target and nontarget class is the criteria for decision making while in i-SVM, the distance from the separating hyperplane is the criteria for the decision. It was seen that the i-PSVM approach is advantageous when tested with mimicked speech as non-target trials. This highlights that proximity to the target speakers is a better criteria for speaker verification for mimicked speech. Further, we note that weighting the target and non-target class examples helps us further fine tune the performance of i-PSVM. We then devised a strategy for estimating the weights for every example based on its cosine distance similarity with respect to the centroid of target class examples. The final i-PSVM with example based weighting scheme achieved an improvement of 3.39% absolute in EER when compared to the best baseline system, iSVM. Subsequently, we fused the i-PSVM and i-SVM systems and results show that the performance of the combined system is better than the individual systems.

More »»

2016

A.S. Raghunath, Sreekumar, K. T., Kumar, C. S., and Dr. K. I. Ramachandran, “Improving speed independent performance of fault diagnosis systems through feature mapping and normalization”, in Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016, 2016, pp. 764-767.[Abstract]


High accuracy fault diagnosis systems are extremely important for effective condition based maintenance (CBM) of rotating machines. In this work, we develop a fault diagnosis system using time and frequency domain statistical features as input to a backend support vector machine (SVM) classifier. We evaluate the performance of the baseline system for speed dependent and speed independent performance. We show how feature mapping and feature normalization can help in enhancing the speed independent performance of machine fault diagnosis systems. We first perform feature mapping using locality constrained linear coding (LLC) which maps the input features to a higher dimensional feature space to be used as input to an SVM classifier (LLC-SVM). It is seen that there is a significant improvement in the speed independent performance of the fault identification system. We obtain an improvement of 11.81% absolute and 10.53% absolute respectively for time and frequency domain LLC-SVM systems compared to the respective baseline systems. We then explore variance normalization considering the speed specific variations as noise to further improve the performance of the fault diagnosis system. We obtain a performance improvement of 8.20% absolute and 6.71% absolute respectively over the time and frequency domain LLC-SVM systems. It may be noted that that the variance normalized LLC-SVM system outperforms. © 2016 IEEE.

More »»

2015

Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, Gopinath, R., and Vaijeyanthi, V., “Fine Tuning Machine Fault Diagnosis System Towards Mission Critical Applications”, in International Symposium on Intelligent Systems Technologies and Applications (ISTA-2015), Kochi; 2015, 2015, pp. 217–226.

2015

K. K. George, Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, and Panda, A., “Cosine Distance Features for Robust Speaker Verification”, in Interspeech 2015, Dresden, Germany, 2015.[Abstract]


We use similarities with people we know already as a means to enhance the speaker verification accuracy. Motivated by this, we use cosine distance similarities with a set of reference speakers, cosine distance features (CDF), to improve the performance of speaker verification systems for clean and additive noise test conditions. We used mel frequency cepstral coefficients, power normalized cepstral coefficients, or delta spectral cepstral coef- ficients for deriving CDF. We then input CDF to a support vector machine (SVM) backend classifier (CDF-SVM). The performance of CDF-SVM was then compared with an i-vector with cosine distance scoring (i-CDS), and an i-vector with a backend SVM classifier (i-SVM) for stationary and non-stationary noises at different signal to noise ratio (SNR) levels. The experimental results show that, the CDF-SVM outperforms all other systems at high SNR and clean environments. However, in certain low SNR cases, i-CDS was found to be better. Finally, we fused the CDF-SVM with i-CDS and results show that the noise robustness of the combined system is significantly better than the individual systems for both high and low SNR levels. Index Terms: speaker verification, i-vectors, support vector machines, cosine distance features, noise robustness

More »»

2015

Dr. Santhosh Kumar C., George, K. K., Dr. K. I. Ramachandran, and Panda, A., “Weighted cosine distance features for speaker verification”, in 12th IEEE International Conference Electronics, Energy, Environment, Communication, Computer, Control: (E3-C3), INDICON 2015, 2015.[Abstract]


Cosine distance similarities with a set of reference speakers, cosine distance features (CDF), with a backend support vector machine classifier (CDF-SVM) have been explored in our earlier studies for improving the performance of speaker verification systems. Subsequently, we also investigated on its effectiveness in improving the noise robustness of speaker verification systems. In this work, we study how the performance of CDF-SVM systems can be further improved by weighting the feature vectors using latent semantic information (LSI) technique. We use mel frequency cepstral coefficients (MFCC), power normalized cepstral coefficients (PNCC), or delta spectral cepstral coefficients (DSCC) for deriving CDF. Experimental results on the female part of short2-short3 trials of NIST speaker recognition evaluation dataset show that the proposed weighted CDF-SVM system outperforms the baseline i-vector with cosine distance scoring (i-CDS), i-vector with a backend SVM classifier (i-SVM) and CDF-SVM systems. Finally, we fused the weighted CDF-SVM with i-CDS and the performance of the combined system was evaluated under different stationary and non-stationary additive noise test conditions. It was seen that the noise robustness of the fused weighted CDF-SVM+i-CDS system is significantly better than the individual systems and the fused CDF-SVM+i-CDS of our earlier work in both clean and noisy test environments except for the zero SNR level condition of certain noises. © 2015 IEEE.

More »»

2015

N. V. Varghees and Dr. K. I. Ramachandran, “Heart murmur detection and classification using wavelet transform and Hilbert phase envelope”, in 2015 21st National Conference on Communications, NCC 2015, Indian Institute of Technology BombayMumbai; India, 2015.[Abstract]


Detection and classification of heart murmurs play an important role in accurate diagnosis of different types of heart dysfunctions. In this paper, we present a noise-robust method for detection and classification of heart murmurs using stationary wavelet transform (SWT) and Hilbert phase envelope. The proposed method consists of five major stages: SWT based PCG signal decomposition for identifying heart sound (HS) including S1, S2, S3 and S4, and heart murmur(HM) subbands, Hilbert phase envelope based boundary determination, temporal feature extraction, murmur detection and classification rule. The boundaries of local acoustic HS segments are determined using the positive slope of instantaneous phase waveform of the smooth absolute envelope. The temporal features such as amplitude, duration, zerocrossing rate, interval, onset and offset time-instants of the detected HS and HM segments are used at the classification stage. The performance of the proposed method is tested and validated using a wide variety of normal and pathological signals containing different patterns of heart sounds and murmurs. The method achieves a probability of correctly detecting HM segments Pms=100%, a probability of correctly detecting HS segments Phs=97.33% and probability of falsely detecting segments Pfs=1.33% for SNR value of 15 dB, and murmur classification accuracy ranging from 82.76% to 100%. © 2015 IEEE.

More »»

2015

R. H. Nair, J. Gini, R., and Dr. K. I. Ramachandran, “A simplified approach to identify the fetal ECG from abdECG and to measure the fHR”, in IFMBE Proceedings, 2015, vol. 52, pp. 23-26.[Abstract]


Fetal ECG (fECG) recording aids physicians to diagnose congenital disorders and other anomalies like asphyxia at the early stages of pregnancy. The fECG extraction has been an area of intensive research. Despite the existence of sophisticated and detailed algorithms – based on adaptive filters, independent component analysis (ICA), &c – filtering out the fECG, buried in the noise and mixed up with the maternal ECG (mECG) remains a challenging task. Some residues of mECG are always present in the fECG extracted with all such techniques. A simple algorithm has been developed here to identify the local maxima in the pre-processed abdominal ECG (abdECG) through thresholding; it locates the mECG peaks explicitly. At the outset, the abdECG has been refined by removing the baseline wander and power line interference at a pre-processing stage. With these as pivots the mECG component is eliminated and the fECG of good quality culled out. The fetal heart rate (fHR) and information required to know the condition of fetal heart can be extracted from this fECG effectively. Extraction of these information helps reducing the rate of fetal mortality, and improving the health condition of fetus as well as mother. Performance of the method is better than the conventional adaptive filtering method and the same is proven quantitatively. A processor based realization of the scheme adds to its credibility substantially to ensure its usability in practice. © Springer International Publishing Switzerland 2015.

More »»

2015

SaSaranya Devi, Dr. K. I. Ramachandran, and Sharma, Ab, “Retinal vasculature segmentation in smartphone ophthalmoscope images”, in IFMBE Proceedings, 2015, vol. 52, pp. 64-67.[Abstract]


Retinal imaging system assists ophthalmologists to diagnose the diseases and to monitor the treatment processes. Conventionally, fundus retinal images are obtained from expensive systems like fluorescein angiography and fundus photography but these systems are large tabletop units and can only be handled by trained technicians. Hence, this study reports a low cost, compact and user friendly smartphone ophthalmoscope to perform indirect ophthalmoscopy. By using this system, initial and periodic screening of retina (both center and periphery regions) becomes easier. Traditionally, retinal diseases are diagnosed by manual observations of fundus images and it is a time consuming process. So, automatic retinal disease diagnosing systems are introduced by extracting the essential features of the fundus retinal images. One of the most essential features of the retina is the blood vessels as its morphological changes helps in diagnosing the retinal diseases. Hence, in this study blood vessels are extracted from smartphone ophthalmoscope (SO) images using level set method to develop an automatic retinal disease diagnosing systems for ophthalmologists. The performance of the retinal vasculature segmentation algorithm is compared and analyzed on DRIVE database of retinal images and on smartphone ophthalmoscope images using the measures like sensitivity, specificity and accuracy level. © Springer International Publishing Switzerland 2015.

More »»

2014

K. K. George, Arunraj, K., Sreekumar, K. T., Dr. Santhosh Kumar C., and Dr. K. I. Ramachandran, “Towards improving the performance of text/language independent speaker recognition systems”, in 2014 International Conference on Power Signals Control and Computations (EPSCICON), 2014.[Abstract]


Speaker Recognition is an active area of research for the last few decades for its applications in several national security, and other forensic applications. In this work, we present the details of a speaker recognition system developed using universal background model and support vector machines(UBM-SVM). We explored several techniques to improve the performance of the baseline system developed using mel frequency cepstral coefficients(MFCC) as input features. We developed and tested the speaker recognition system for 200 speakers, using the data collected over 13 different channels, such as handset regular phone, speaker phone, regular phone headphone, regular phone, etc. We experimented with the use of RelAtive SpecTrA (RASTA) processing, and feature warping on the input MFCC features, and nuisance attribute projection (NAP) on the Gaussian mixture model supervectors derived in the system. It was seen that these techniques have helped improve the system performance significantly by minimizing the effect of different channels on the system performance. The details of the system implementation and results are presented in this paper. The complete system is developed in MATLAB and C/C++.

More »»

2014

P. T Kumar, Jasti, A., Dr. Saimurugan M., and Dr. K. I. Ramachandran, “Vibration Based Fault Diagnosis of Automobile Gearbox Using Soft Computing Techniques”, in Proceedings of the 2014 International Conference on Interdisciplinary Advances in Applied Computing, New York, NY, USA, 2014.[Abstract]


Gearbox is the core component in any automotive/industrial application and it consists of gears and gear trains to vary the speed and torque of the machine. In order to reduce the machine breakdown cost and to increase the service life it is vital to know its operating conditions frequently to find the point of defect. The vibration signals are used to extract statistical features for 3 different classes namely Gearbox with Good gear, Gear Tooth breakage and Gear Face wear. The features were collected according to the experimental conditions with 3 fault classes, 3 speeds and 1 load condition with total of 9 testing conditions. The prominent statistical features were selected using decision tree algorithm. The set of IF-Then rule was generated and coded in LabVIEW for automated machine fault diagnosis.

More »»

2014

J. Selvaraj, Dr. K. I. Ramachandran, Venkatesh, D., Devanathan, S., B., S., J., G. Murali, and P., L. Kumar, “Greening the foundry sector by an innovative method of energy conservation and emission reduction”, in 2014 IEEE 8th International Conference on Intelligent Systems and Control: Green Challenges and Smart Solutions, ISCO 2014 - Proceedings, 2014, pp. 60-63.[Abstract]


Foundries are known for their energy intensiveness and environmental pollution. More than 50 % of the energy consumed by the foundries is spent in melting the raw materials and this energy goes waste while molten metal solidifies in sand molds. This paper aims at harvesting that waste heat liberated from molten metal, using the harvested heat to preheat the scraps that are embedded into the sand molds. This preheated scraps when melted for the next batch, the energy consumption in the furnace is reduced by 12 %. This energy conservation is a novel approach and readily gives rise to environmental benefits by reducing energy-related emissions. This method also improves the recyclability of the foundry sand, since the peak sand temperature reduces by more than 100 K, making this method doubly environmental-friendly. © 2014 IEEE.

More »»

2014

V. N. Varghees and Dr. K. I. Ramachandran, “Automated PCG signal delineation method for heart sound analysis”, in Twentieth National Conference on Communications (NCC), 2014 , 2014.[Abstract]


This paper presents an automated PCG signal delineation method for real-time wireless cardiac health monitoring and audio-visual stethoscope applications. The proposed method comprises the steps of: preprocessing, total variation filtering, smooth energy envelope computation, peak-amplitude determination and endpoint determination. The total variation filter is used to smooth out the background noises and preserves the heart sound components. The endpoints of heart sounds (S1, S2, S3, S4 and murmurs) are determined by using smooth energy envelope and hard-thresholding rule. The peaks of the heart sounds are determined by using candidate waveform obtained from Hilbert transformation of smooth envelope. The proposed delineation method is validated by using different clean and noisy pathological and non-pathological PCG signals. Experiments on a large PCG database show that the proposed method achieves an average sensitivity (Se) of 99.56% and positive predictivity (+P) of 96.97%, with the maximum average delineation error of 9.08-msec, 0.483-msec and 7.834-msec for endpoints of heart sounds, peak-locations and durations, respectively.

More »»

2012

B. S. Devasenapati and Dr. K. I. Ramachandran, “Random forest based misfire detection using kononenko discretiser”, in SOCO 2012, 2012.[Abstract]


This paper evaluates the use of random forest (RF) as a tool for misfire detection using statistical features. The engine block vibration contains hidden information about the events occurring inside the engine. Misfire detection was achieved by processing the vibration signals acquired from the engine using a piezoelectric accelerometer. The hidden information regarding misfire was decoded using feature extraction techniques. The effect of Kononenko based discretiser as feature size reduction tool and Correlation-based Feature Selection (CFS) based feature subset selection is analysed for performance improvement in the RF model. The random forest based model is found to have a consistent high classification accuracy of around 90% when designed as a multi class ,ode and reaches 100% when the conditions are clubbed to simulate a two-class mode . From the results obtained the authors conclude that the combination of statistical features and RF algorithm is well suited for detection of misfire in spark ignition engines.

More »»

2006

V. Sugumaran, Sachin, K., and Dr. K. I. Ramachandran, “Fault Diagnostics of Roller Bearing using Single-shot Multi-class Support Vector Machine”, in International Conference on Information Technology (ICIT 2006), Indian Institute of Science, Bangalore, 2006.

2006

G. R. Sabareesh, Sugumaran, V., and Dr. K. I. Ramachandran, “Fault diagnosis of a taper roller bearing through histogram features and proximal support vector machines”, in IEEE international conference on signal and image processing, December, B V Bhoomaraddi college of Engineering and Technology, Hubli, 2006.

Publication Type: Conference Proceedings

Year of Publication Title

2018

Dr. Anju Pillai S., Motaghare, O., and Dr. K. I. Ramachandran, “Predictive Maintenance Architecture”, IEEE International Conference on Computational Intelligence and Computing Research (IEEE ICCIC). 2018.[Abstract]


In industrial plants or any critical utility plants, the ultimate goal is to maximize the production quantity and quality but at the same time keeping the production cost as low as possible. To achieve this, it is mandatory to keep plants in fully efficient condition so that the throughput of the system is maximum. In order to keep the system fully efficient it needs to be maintained properly. There are different maintenance strategies being used to maintain the efficiency of the plant. For any specific type of industry, maintenance affects the cost of goods produced. To avoid breakdown, the maintenance strategies should be planned in such a way that the maintenance tasks are executed at right time. Unnecessary maintenance tasks increase the maintenance costs and also the time required to execute them. Through this paper, the prospect of optimizing the plant operation i.e. to reduce the down time of the system using predictive maintenance (PdM) approach which will lead to reduced production cost has been explored.

More »»

2018

J. Joseph, J. Gini, R., and Dr. K. I. Ramachandran, “Removal of BW and Respiration Noise in abdECG for fECG Extraction”, Advances in Signal Processing and Intelligent Recognition Systems, vol. 678. Springer International Publishing, Cham, pp. 3-14, 2018.[Abstract]


Electrocardiogram (ECG) signals are one of the most important diagnostic tools for any doctor, especially a cardiologist. It is important that the fetus present inside the abdomen undergoes a fetal ECG recording to assess the health of the fetus. Complications like disturbance because of movement of abdominal muscles are usually present during the recording and leads to the wrong diagnosis of the fetus ECG. In this paper, the signal in dispute had been altered in the proposed method so as to eliminate the wandering of the baseline, respiration noise and also expel the noise from other sources. The acquired abdominal ECG signal in a noninvasive manner had been considered for extracting the fetal ECG after eliminating the noise. The windowed zero mean method is used where the first step is segmentation. In segmentation, the abdominal ECG signal is divided into set of samples based on window size. Zero mean is applied across each of the windowed abdominal ECG signals to address the issue of baseline wandering and respiration noise. This is followed by the application of a bandpass filter to cancel the high-frequency noise component. This process results in an ECG signal that almost has no complications as present before. The fetal ECG signal that is procured using such a method is now easier to diagnose as compared to the acquired signal which contains noise. Thus, for a fetus, this can help in proper diagnosis. It is further noted that this method is very reliant on using and is lucid. It can be used to augment and alter signals where such complications arise in the field of medicine and clinical diagnosis

More »»

2018

R. J. Gini, Chakravarthy, P. Deepan, Dr. K. I. Ramachandran, and Anand, P., “Modeling of a System for fECG Extraction from abdECG”, Advances in Intelligent and Soft Computing, ISDA, vol. 736. Springer International Publishing, pp. 568-579, 2018.[Abstract]


The objective of this paper is to move a step ahead in investigation and create a feasible, cost effective fetal ECG analysis tool for clinical practice which will be easy for usage by any non-skilled personal and provide actionable medical information such as the QRS complex of fetal ECG, fetal HR etc. In this method, a composite abdominal ECG is subjected to a pre-processing stage which involves filtering and normalization, then fed into the `thresholding and peak finding' stage to detect the maternal ECG peaks. The next stage involves construction of the MLE of maternal ECG embedded in the abdominal ECG. After this, the constructed MLE which represent the maternal ECG is subtracted from the abdominal ECG to obtain fetal ECG along with a smidgen of noise. This noise which adulterates the fetal ECG is removed by filtering, done at the post processing stage. Thresholding and peak finding is done at the post processed signal to calculate the fetal HR. This paper puts forth a promising possibility of implementing the proposed algorithm in any suitable hardware model, since an average Accuracy of 76.8% and average Sensitivity of 90.7% is attained.

More »»

2016

R. J. Gini, Dr. K. I. Ramachandran, Nair, R. H., and Anand, P., “Portable Fetal ECG Extractor from abdECG”, 2016 International Conference on Communication and Signal Processing (ICCSP). pp. 0698 – 0701, 2016.[Abstract]


This paper aims at creating an affordable fECG extractor by simplifying the process of fECG extraction from abdECG. Even though invasive fECG extraction is more accurate, noninvasive method of extraction has been preferred during prenatal considering the fetus's health. This makes the noninvasive fECG extraction an emerging and required field of research. This paper gives a fundamental idea to create a prototype for extracting the fetal ECG from abdominal ECG. The abdECG has been preprocessed by normalization and filtering. Based on thresholding and first order differentiation, the maternal peak has been identified from the preprocessed abdECG signal. Using the identified maternal peaks, QRS complex of mECG has been identified and the same has been cancelled out from abdECG to cull out the fECG. The resultant signal has been a combination of fECG and noise. The fetal peaks have been identified from the culled out signal. The identified fetal peaks provide information like the QRS complex of the fetus, fetus heart rate, diagnosis of any congenital disorder and other anomalies. This simplified algorithm has been implemented with high level language C and executed using Raspberry Pi. The execution results with a second delay and Raspberry Pi can create a standalone platform at any place and is handy. The system resulted in 100% accuracy when the selected channel happened to be near the fetus's heart. Even in other cases, it has proven to be good and effective. This shows that the system is affordable and practically useable.

More »»

2015

K. T. Sreekumar, Gopinath, R., Pushparajan M., A.S. Raghunath, Dr. Santhosh Kumar C., Dr. K. I. Ramachandran, and Dr. Saimurugan M., “Locality Constrained Linear Coding for Fault Diagnosis of Rotating Machines using Vibration Analysis”, 12th IEEE India International Conference on Electronics, Energy, Environment, Communication, Computer Science, Control (INDICON, 2015). Institute of Electrical and Electronics Engineers Inc., JamiaMilliaIslamia, NewDelhi, pp. 1-6, 2015.[Abstract]


Support Vector Machine (SVM) is an important machine learning algorithm widely used for the development of machine fault diagnosis systems. In this work, we use an SVM back-end classifier, with statistical features in time and frequency domain as its input, for the development of a fault diagnosis system for a rotating machine. Our baseline system is evaluated for its speed dependent and speed independent performances. In this paper, we use locality constrained linear coding (LLC) to map the input feature vectors to a higher dimensional linear space, and remove some of the speed specific dimensions to improve the speed independent performance of the fault diagnosis system. We use LLC to do the feature mapping to the higher dimensional space, and select only the k nearest neighbour basis vectors to represent the input feature vector and thus reduce/minimize the effect of speed specific factors from the input feature vector, and thus improve the speed independent performance of the fault diagnosis. We compare the performance of the LLC-SVM system for the time and frequency domain statistical features. The proposed approach has improved the overall classification accuracy by 11.81% absolute for time domain features and 10.53% absolute for frequency domain features compared to the baseline speed independent system. © 2015 IEEE.

More »»

2015

K. Ka George, Dr. Santhosh Kumar C., Panda, Ab, Raji Ramachandran, Das, K. Aa, S. Veni, and Dr. K. I. Ramachandran, “Minimizing the false alarm probability of speaker verification systems for mimicked speech”, 2015 International Conference on Computing and Network Communications, CoCoNet 2015. Institute of Electrical and Electronics Engineers Inc., pp. 703-709, 2015.[Abstract]


Speaker verification (SV) systems need to be robust to mimicked voices of target speakers as non-target trials to make them usable in critical applications. However, the performance of SV systems for mimicked voice test conditions has not been extensively explored.

More »»

2013

R. Gopinath, Nambiar, T. N. P., Abhishek S., Pramodh, S. M., Pushparajan M., Dr. K. I. Ramachandran, Dr. Santhosh Kumar C., and Thirugnanam, R., “Fault Injection Capable Synchronous Generator for Condition based Maintenance”, 7th International Conference on Intelligent Systems and Control, ISCO 2013. Coimbatore, Tamilnadu, pp. 60-64, 2013.[Abstract]


This paper presents the design specifications of an experimental setup capable of injecting faults in a synchronous generator to develop and test algorithms for condition based maintenance of aerospace applications. A 3 kVA alternator is designed to inject faults in the stator winding and field windings. The system is capable of injecting open and short circuit faults in the stator and rotor windings using a fault injection unit and record current, voltage and vibration signals from the respective sensors, with programming capability. The response of the synchronous generator during normal condition and faulty condition are discussed. The work reported in this paper will help other researchers develop low cost experimental facility to pursue research in machine condition monitoring. © 2013 IEEE.

More »»

2009

Dr. Elangovan M., S. Devasenapati, B., and Dr. K. I. Ramachandran, “Condition Monitoring of a Single Point Cutting Tool Using Support Vector Machine”, International Conference on Advances in Mechanical and Building Sciences (ICAMB-2009) . Vellore Institute of Technology, Vellore, 2009.

Publication Type: Book Chapter

Year of Publication Title

2017

A. R., T., V., and Dr. K. I. Ramachandran, “A Hybrid Classifier for the Detection of Microaneurysms in Diabetic Retinal Images”, in The 16th International Conference on Biomedical Engineering, Springer, 2017, pp. 97-103.[Abstract]


Diabetic Retinopathy (DR) is a chronic, progressive ocular disease in which the human retina is affected due to an increasing amount of insulin in blood. The prevalence and incidence of DR is associated with people having prolonged hyperglycaemia and other symptoms linked with diabetes mellitus. DR, if not detected and treated in time poses threat to the patient’s vision ultimately causing total blindness. Among the various clinical signs, microaneurysms (MAs) appear as the early and first sign of DR. The accurate and reliable detection of microaneurysms is a challenging problem owing to its tiny size and low contrast. Successful detection of microaneurysms would be more useful for a proper planning and appropriate treatment of the disease at the early stage. The work mainly envisages the improvement of the classification accuracy by employing a hybrid classifier which combines Support Vector Machine (SVM), Naïve Bayes Classifier and the decision tree. In contrast to many other classifiers the proposed classifier works efficiently, proves to be simple in terms of computational complexity and also gives good results. The performance is evaluated using publicly available retinal image database DIARETDB1.The hard decision fusion among the three classifiers carried out using the majority voting rule gives accuracy, sensitivity and specificity of 82.2916%, 82.692%, 81.818% respectively.

More »»

Publication Type: Patent

Year of Publication Title

2016

Lakshmikanthan C., Dr. K. I. Ramachandran, Sharma, A., and Subramaniam, S. Devi, “A Handheld device to capture Fundus images”, 2016.[Abstract]


The present invention relates to a hand held device used to capture fundus images. More particularly the intention provides a low cost hand held device to image entire fundus either it is central or peripheral. The invention is smart phone based hand held device capable of capturing regions of peripheral most retina such as ora serrata and further till pars plana apart from central fundus. Advantageously, the present invention is operated by single hand.
(FR) La présente invention concerne un dispositif portatif utilisé pour capturer des images du fond de l'œil. Plus particulièrement, l'invention concerne un dispositif portatif économique permettant de former l'image de l'intégralité du fond de l'œil, qu'il soit central ou périphérique. L'invention est un dispositif portatif basé sur un téléphone intelligent susceptible de capturer des régions de la rétine la plus périphérique telle que l'ora serrata ou encore pars plana en plus du centre du fond de l'œil. De manière avantageuse, la présente invention est actionnée d'une seule main.

More »»

Publication Type: Book

Year of Publication Title

2010

Dr. Soman K. P. and Dr. K. I. Ramachandran, Insight into wavelets: From theory to practice. PHI Learning Pvt. Ltd., 2010.[Abstract]


Wavelet theory has matured and has entered into its second phase of development and evolution in which practitioners are finding newer applications in ever-widening scientific domains such as bio-informatics, computational drug discovery and nano-material simulation. Parallelly, the theory of wavelets got more and more demystified and has become an everyday tool for signal and image processing. Postgraduate courses in mathematics and physics now include a subject on wavelet theory either as a separate... More »»

2008

Dr. K. I. Ramachandran, Deepa, G., and P. K. Krishnan Namboori, Computational Chemistry and Molecular Modeling. Springer, 2008.[Abstract]


Computational chemistry and molecular modeling is a fast emerging area which is used for the modeling and simulation of small chemical and biological systems in order to understand and predict their behavior at the molecular level. It has a wide range of applications in various disciplines of engineering sciences, such as materials science, chemical engineering, biomedical engineering, etc. Knowledge of computational chemistry is essential to understand the behavior of nanosystems; it is probably the easiest route or ...

More »»