SELF-TUNING CONTROLLER FOR NEURAL NETWORK PREDICTIVE DEVIATION COMPENSATION BASED ON DAMPED LEAST SQUARE

In this paper, we propose a kind of neural network predictive deviation compensation self-tuning controller, using linear model to control nonlinear system. The predictive deviation's compensation is completed by neural network. The damped least square method was applied in both identification of the linear model and study of the neural network. The simulation results show that this kind of controller can control nonlinear system effectively and has small overshoot and good stability.