FAULT TOLERANT CONTROL OF ROBOT MANIPULATORS WITH MANFIS AND GAIN SCHEDULING

Abstract

While robot manipulators cooperate with other manipulators or working with high accuracy in partnership with people, malfunctions or fault may occur. The robotic systems must be able to isolate and tolerate these faults with the designed controllers. In this study, a multiple-ANFIS-based fault diagnosis approach that converges to the failure function for manipulators is proposed and the coefficients of the Computed Torque PID (CT-PID) controller defined for manipulator control are varied using Gain Scheduling method as the fault tolerant control approach. Simulation results for a two DOF robot manipulator with dynamic uncertainties proves the robustness of the proposed system.