Artificial Intelligence Based First Order Adaptive Sliding Mode Controller for Position Control of a DC Motor Actuator

Abstract

This paper presents an Artificial Intelligence (AI) based approach uniquely applied to permanent magnet DC motor actuator for position control. The AI method employed in this work is fuzzy logic. A first order lag sliding mode controller is tuned and combined with an adaptive Fuzzy-PI controller architecture which operates in parallel. The controller architecture proposed in this study is aimed at improving the disturbance rejection capability, steady state as well as transient performance of the conventional adaptive Fuzzy-PI controller and sliding mode controller. Hence, the robust control law of the proposed controller (SM+FZ-PI) consists of a discontinuous sliding mode output added to a continuous adaptive Fuzzy-PI controller output. The sliding mode controller switches on only when disturbance in the system is detected. The performance of the proposed controller architecture has been compared with a conventional PID and adaptive Fuzzy-PI controllers for performance evaluation with respect to several operating conditions such as load torque disturbance injection, noise injection in feedback loop, motor non-linearity exhibited by parameters variation, and a step change in reference input demand. The proposed controller (SM+FZ-PI), had the best disturbance rejection and steady state error elimination.