Creating a sliding mode in a motion control system by adopting a dynamic defuzzification strategy in an adaptive neuro fuzzy inference system

Abstract

Adaptive neuro-fuzzy inference systems exhibit both the numeric power of neural networks and verbal power of fuzzy inference systems. Particularly in control applications, the prime difficulty in selecting the parameters of such systems stems from the unavailability of the desired control inputs. In this paper, a novel approach is presented for the establishment of a sliding mode in the plant under control. The approach presented adjusts solely the parameters of the defuzzifier. At the adjustment stage, a dynamic adaptation law is proposed and it is proved that the particularly chosen form of the adaptation strategy creates a sliding mode in the plant behavior while the parameters of the controller are also in a sliding mode. In the simulations, the dynamic model of a 2-DOF direct drive robotic manipulator is used. It is observed that the method discussed is highly robust against the disturbances like varying payload conditions and noisy observations.