Research On Hydropower Unit Identification And Optimization For Control Parameters

With the increasing installed capacity of hydropower, the safe and stable operation of hydropower stations has become the primary concern of engineering and technical personnel. The safety and stability of the hydropower station depends largely on the advanced nature of the hydropower station control system. At present, the conventional PID control algorithm is adopted in the hydropower station control system. Although the algorithm is simple and robust, The more frequent the process control effect is not good, which may lead to the accused unit speed deviation and thus a greater impact on the stability of the power grid frequency. In this paper, after studying the development status of hydropower unit control system, combining with the intelligent control theory which is widely researched, fuzzy control and neural network control in intelligent control theory are applied to hydropower unit control system, and the proposed control Strategy and method are analyzed.Based on the complexity, non-linearity and non-minimum phase of the hydro-generator system, two feedforward neural network hydroelectric generating system identification models are established by using the approximation function of feed-forward network. Aiming at the difference of the structure of RBF neural network,the strategy of identifying the number of nodes, the center point vector and the width of the Gaussian base line is studied, and a good system model is obtained.In order to solve the problem that the conventional PID control algorithm can not deal with the frequent change of the working condition, the BP and RBF feedforward neural networks are used to optimize the PID parameters to improve the PID control effect. The RBF network is different from the BP controller in directly outputting PID parameters. The simulation results show that the two controllers are better than traditional PID control, and the RBF network is more effective.Aiming at the problems of the large variety of data and the complicated condition of the hydropower station, it is difficult to test the large amount of data in the neural network control online training. At last, a model-free control method-fuzzy control is studied. On the basis of analyzing the shortcomings of traditional fuzzy control, fuzzy PD control with performance index is proposed, which provides the basis for thecontroller parameter design. At the same time, fuzzy language is used to describe the relationship between the variation of PID parameters and the error and error variation.,The same dynamic PID control is realized. Finally, the simulation results show that the controller has better adaptability and robustness than the traditional PID control.