Generalized Predictive Control Of Hydraulic Turbine Governing System Based T-S Fuzzy Model

In order to improve the control performance of the hydraulic turbine governing system, this paper studies on the application of predictive control in the hydraulic turbine control. The predictive control theory and algorithm mainly aimed at linear system, for the hydraulic turbine governing system which is a nonlinear and time-variable complex system. If we treat the nonlinear system as linear for simplicity and design the controller based on it, the control performance will be very finite because of the bias between the linear model and real system. The fuzzy system has the similar pattern with the human brain which can deal with qualitative information. The T-S fuzzy model is a nonlinear model in essence and it can describe the dynamic characteristics of complex system. It has very strong approximation ability and generalization ability. This paper studies on the generalized predictive control of hydraulic turbine governing system based on T-S fuzzy model which combines the advantages of fuzzy model and predictive control.A new single-hidden layer feedforward neural network-ELM is proposed in application of hydraulic turbine modeling. The ELM has the advantages of extreme learning speed, high learning accuracy, good generalization ability and so on. The use of ELM in description of hydraulic turbine flow characteristic and torque characteristic,can truly express the nonlinear property of hydraulic turbine. It's a very effective method of hydraulic turbine simulation modeling.The modeling of hydraulic turbine governing system control object based on T-S fuzzy identification is studied. For the problem of input selection of fuzzy identification modeling, a decision tree search method is used to select the optical input set. The simulation results show that this modeling method has good identification accuracy in the dynamic process of hydraulic turbine.In view of the nonlinear and time-variable property, a generalized predictive controller of hydraulic turbine governing system based on T-S fuzzy model is designed.The T-S fuzzy model is used as predictive model, and online tuning the predictive model.Then control the hydraulic turbine governing system based on the online adaptive model.The simulation experiments on hydraulic turbine governing system shows good control performance and the performance index outperform the conventional PID control and the GPC based on linear model.