Hybrid Fuzzy Proportional-Integral plus Conventional Derivative Control of Robotics Systems

Abstract

This chapter presents a new approach towards optimal design of a hybrid fuzzy proportional-integral-derivative (PID) controller for robotics systems using genetic algorithm (GA). The proposed hybrid fuzzy PID controller is derived by replacing the conventional PI controller by a two-input normalized linear fuzzy logic controller and implementing the conventional D controller in an incremental form. The salient features of the proposed controller are: (1) Gain scheduling method is incorporated in the design to linearize the robotics system for a given reference trajectory. (2) Only one well-defined linear fuzzy control space is required for multiple local linearized systems. (3) The linearly defined fuzzy logic controller can generate sector bounded nonlinear outputs so that the closed-loop system is stable and has better performance. (4) Optimal tuning of controller gains is carried out by using GA. (5) It is simple and easy to implement. Simulation studies on a pole balancing robot and a multi-link robot manipulator demonstrate the effectiveness and robustness of the proposed controller. Comprehensive comparisons with other latest approaches show that the proposed approach is superior in terms of tracking performance and robustness.