September 3, 2012
School of Engineering, Bengaluru
Did you know that a flying aircraft undergoes a change in its total mass due to fuel consumption? As the airplane flies, its mass slowly reduces, making necessary attendant changes in several control systems.
“Some dynamic systems to be controlled have slowly-varying or uncertain parameters. For instance, robotic manipulators may carry large objects with unknown inertial parameters. Power systems may be subjected to large fluctuations in loads,” explained Mr. R. Karthikeyan, Assistant Professor, Amrita Department of Electronics and Communication Engineering, Bengaluru.
Adaptive control systems are necessary not only for aircraft and rocket control, robotic manipulation and in power systems but also for ship steering and in bio- and chemical engineering.
A traditional method for constructing adaptive controllers is referred to as the Model Reference Adaptive Control (MRAC) system.
“MRACs can help sustain unfailing system performance in presence of uncertainty or unknown variations in plant parameters. The basic idea in adaptive control is to estimate the uncertain parameters based on measured system signals, and use these estimations in control input computation,” Mr. Karthikeyan elaborated.
Recently Mr. Karthikeyan presented a paper titled Analyzing Large Dynamic Set-Point Change Tracking of MRAC by Exploiting Fuzzy Logic Based Automatic Gain Tuning at the IEEE Control and System Graduate Research Colloquium in Shah Alam, Malaysia.
The paper was co-authored by Mr. Rahul Kumar Yadav, Mr. Hemanth Kumar G. and Dr. Shikha Tripathi of the same department.
“It was an enriching experience being in the company of eminent scholars working in related research areas who all provided good feedback and made many useful suggestions for my future research work,” shared Mr. Karthikeyan.
The presented paper proposed a method to overcome the pitfalls encountered in conventional MRACs, without the need for any human interference.
“MRACs belong to that class of adaptive servo systems in which the desired performance is expressed with the help of a reference model. There are adjustable parameters and a mechanism to adjust the parameters to maintain consistent system performance. We found that there is a tolerance band for the set point change which defines the effectiveness of a particular adaptive gain (γ). Any change in the set point which is beyond this band, calls for a γ-readjustment. Accordingly, we proposed a method to overcome this pitfall in conventional MRAC by fusing fuzzy logic to dynamically vary γ,” Mr. Karthikeyan explained.
“It was worthwhile for us to note that, while MRAC drives the plant’s response in a manner that makes it mimic the output of the model, the incorporated fuzzy logic control claims the responsibility of providing the needed gain, allowing the plant to follow the specified model. The simulation results show considerable improvement in performance over conventional MRAC systems. Future work includes testing the same on a three tank system and validating the results obtained through simulation,” he added.