A NEW LEARNING CONTROLLER AND ITS APPLICATION TO PNEUMATIC SERVO SYSTEM

This paper presents a new learning control law which can obtain learning gain matrices from a recursive identifier alone the learning axis. The controller is designed respectively in continuous and discrete time dynamic system, the system convergence is proved, and it is spread when there exists disturbances in the system and its initial error isn't zero in every iteration. Finally it is applied to pneumatic servo control system, the simulation results show its validity.