Neural and Adaptive Control of a Rigid Link Manipulator

In this paper a comparison of classical, adaptive and neural control strategies for a robotic manipulator with two revolute joints is presented. The conventional computed-torque method is presented, as a starting point for the design of the adaptive and neural control techniques. Two adaptive control strategies, a non-model based neural control strategy and a model based neural control strategy are implemented. Computer simulations are performed for the control of a rigid manipulator with two revolute joints, in order to verify the performances of the control strategies and to make some useful comparisons. If a classical controller already controls a robot the advantage of proposed structure is that extension to a model based neural controller for performances improvement is easy. Key-Words: robotic manipulator, neural control, computed torque control, adaptive control, trajectory tracking.