The research of neural networks combined with fuzzy control algorithm applied in hydraulic servo system

Abstract

The mathematic model of the hydraulic servo system can be simplified as a series connection of a servo valve friction dead zone nonlinear and a linear three-order system. The servo valve friction dead zone nonlinear is always static. The paper proposed a method that an inverse model which realized on neural networks is used for compensting the dead zone nonlinear. On the other hand, the parameters of the part of the linear is always perturbation with the influence of pressure of oil supply, the viscosity of oil and the temperature of environment. So fuzzy and PI control algorith are used for solving the above problem. The result of simulation show that the method can effectively deal with the servo valve friction dead zone nonlinear and resist the parameters perturbation of controlled object.