Decentralized Direct Adaptive Fuzzy Control for Flexible-Joint Robots

Abstract

A flexible-joint robot is a complex system with nonlinearity, flexibility and uncertainty. The complexity of modeling and control by conventional methods leads us to apply the fuzzy control. However, stability analysis of the fuzzy control systems is a challenging problem. Based on the guaranteed stability method, this paper presents a novel decentralized direct adaptive fuzzy control approach for the flexible-joint robots. Compared with the majority of control schemes which ignore the actuator dynamics, this study considers the whole robotic system including its motors. Instead of the torque control strategy, the proposed approach is based on the voltage control strategy. The common control structure for flexible-joint robots employs two control loops whereas this control design has only one control loop. Compared with the torque-based control approaches, the proposed control is simpler, less computational and more effective. The proposed control design is verified by stability analysis. A comparison with a modified fuzzy proportional-integral controller is presented to show the effectiveness of the proposed fuzzy controller for tracking application.