Hierarchical Fuzzy Controller for Robot Manipulator

Adaptive fuzzy controllers are potential candidates for controlling robot arms. However, traditional fuzzy controllers suffer from the rule explosion phenomenon. We propose here a hierarchical scheme to implement the fuzzy system and consequently relax the rule explosion problem. Furthermore, since payloads are normally varying, robotic models are uncertain. We propose a fuzzy-logic based dynamic observer to estimate the modeling uncertainties and external disturbances. The estimated model is incorporated within the control law to guarantee an acceptable tracking performance. We prove that the proposed adaptive fuzzy controller assures that all signals are uniformly ultimately bounded. Simulation results of a two-link robot arm are reported to confirm the validity of our theoretical results.