Application of Fuzzy-Neural Network Controller on Water Turbine Generator Set Based on T-S Model

The water turbine generator set has the characters of non-linearity, parameter time-variability and non-minimum phase. It is found that PID controller has the problem of integral saturation and poor robustness and the traditional fuzzy controller has a subjective subordinate degree. To improve the running state, a fuzzy neural network based on T-S model is designed by imitating the structure of the conventional PID controller. This model takes a good use of characteristics of nonlinear and interpretation of fuzzy theory. The abilities of self-study and self-organize of neural network can regulate parameters of fuzzy structure. The control effect between the fuzzy neural network controller and the PID controller would be compared in aspects of the exceed adjusting amount, the regulate time, the static state error and the robustness by simulation. The result shows that the fuzzy neural network controller is better than the PID controller.