신경회로망을 이용한 로보트 매니츌레이터의 Resolved Motion제어기의 설계

Resolved Motion Control of the Robot Manipulator using Neural Network

In this paper we propose the resolved motion controller using a neural network for a robot manipulator. Neural identifier designed by a neural network is trained by using a feedback force as an error signal. The identifier approximates the output of a unknown nonlinear system by monitoring both the input and the output of this system. If the neural network is sufficiently trained well, it does not require either strict modelling of the manipulator or precise parameter estimation. The effectiveness of the proposed controller is demonstrated by computer simulation using a two-link planar robot.