Fuzzy gain scheduling PID control of a hybrid robot based on dynamic characteristics

Abstract

In this paper, a fuzzy gain scheduling proportional-integral-differential (PID) controller based on dynamic characteristics of a hybrid robot named TriMule is presented. Since the equivalent inertia and gravity imposed on motor shafts vary with the robot configuration, it is difficult to obtain satisfactory control performance with a linear, fixed gain PID controller. After analyzing the effect mechanism of dynamic characteristics on control performance, the clustering analysis algorithm and fuzzy logic are combined to deduce the control parameters online according to the command acceleration, equivalent inertia and gravity. The experimental results on a prototype machine show that, compared with the conventional PID and fuzzy gain scheduling PID controller, the proposed method can significantly reduce the effect of dynamic characteristics on control system, thus ensuring satisfactory control performance under any configuration in the task workspace.