A Lyapunov-based model reference control scheme with CMAC neural network

Abstract

This paper presents a novel model reference control scheme. The CMAC(Cerebellar Model Articulation Controller) neural network is used to minimize the differences between the reference model and the plant which is influenced because of parameter variation and disturbance online. Moreover the neural network iterative algorithm based on Lyapunov stability is provided in the paper. By this, the output of the plant can accurately follow the the nominal model, videlicet the plant has a characteristic of linear and certain, whose dynamic performance is the same as the nominal model's. So using the traditional control methods can make the system have perfect transient and steady-state performance. Simulation results demonstrate that the proposed control scheme can reduce the plant's sensitivity to parameter variation and disturbance.