Implementation of Adaptive Self-Organizing Fuzzy Wavelet Neural Network: Modeling AC Servo System

In this paper, an adaptive self-organizing fuzzy wavelet neural network (ASOFWNN) controller is investigated for AC servo system, aims to deal with nonlinear problem and time-varying disturbance. In order to acquire better performance, an improved self-organizing wavelet neural network (SOWNN) is proposed as the consequent part fuzzy rule in the training process. Innoviate with adaptive learning, adaptive gradient algorithm (AGA) is used to optimize the parameters in SOFWNN. Furthermore, the Lyapunov approach is implemented to verify the stability of the system. Finally, simulation and experimental results proved that compared with fuzzy wavelet neural network (FWNN), novel ASOFWNN shows fast convergence rate and more precise modeling performance.