Research on 2D Variable Learning Controller and Application in Pneumatic Servo System

In this paper a novel learning control law is proposed which can obtain learning gain matrices from a recursive identifier along the learning axis. This algorithm convergence is proved with random noise disturbance. The control law can online regulate learning gain matrices along the learning axis by using RLS technique and doesnpsilat need system accurate model. Finally it is applied in pneumatic servo control system, the simulation and experimentpsilas results show its validity, and then receive the idea results.