Publication

Abstract : Lithium-ion battery packs are widely used for automotive applications. An exact state of charge (SOC) estimation is essential to obstruct the battery from overcharge and discharge, increase the battery lifespan, and know the driving range. Even though a number of SOC estimation strategies have been presented by researchers, further exploration is necessary to identify an appropriate technique that can accommodate a variety of lithium-ion battery chemistries. In recent studies, it has been demonstrated that deep learning (DL), a well-known machine learning algorithm, performs better for SOC estimates than many other strategies. To get the most out of DL models, it is crucial to choose the proper hyperparameters and make effective use of relevant input characteristics. This work proposes a novel cloud-integrated battery management system (CIBMS) based on quantifiable features, including voltage, current, and temperature. The training and testing data is collected from the real-time electric vehicle drive set up using an Internet of Things (IoT) sensor and transferred to the cloud using an internet-connected Raspberry Pi 4B+ processor. The data is analyzed using various DL algorithms such as Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network with LSTM (CNN-LSTM), Convolutional Neural Network with Bi-directional LSTM (CNN BI-LSTM), and Convolutional Neural Network-Gated Recurrent Unit-Long Short-Term Memory (CNN-GRU-LSTM) for accurate battery status estimation. The proposed algorithm is incorporated to mitigate issues like pre- processing, overfitting challenges in real-time applications rather than conventional methods. The accomplishment of the algorithm has been evaluated in terms of mean square error (MSE) and root mean square error (RMSE) of the SOC. The obtained results clearly show that the proposed CNN-GRU-LSTM algorithm accurately estimated the battery status which is crucial for real-time decision-making in electric vehicles (EVs).