Learning Controller Design using Fuzzy Modeling, Sliding Mode Control, and PID Controller for Robots

Abstract

This paper presents a novel approach to combine three powerful methods of control theory, namely, chattering free adaptive sliding mode control, fuzzy logic-based modeling method, PID control and fuzzy c-means clustering algorithm to build a robust learning control for nonlinear uncertain systems such as advance robot manipulators. The combination not only encompasses the features and capabilities of its components but also the limitations attributed to these techniques may be remedied by each other. To date different combinations of the above mentioned methods presented in the literature each having its own merits and limitations. But, for advance robotics applications, there is a pressing need for the control systems that are able to learn systematically and efficiently during the course of operation, from its own experience, from the demonstration and through supervised fashion as well. The controller proposed in this paper aims at building such a control system. The global stability and robustness of the proposed controller are established using Lyapunov's approach and fundamentals of sliding mode control theory. Based on the simulations and experimental results, the proposed controller performs remarkably well in terms of the tracking error convergence and robustness against uncertainties.