Intelligent Control Of Bulb Turbine Units

At present, PID control strategy is mostly used in Hydro-turbine regulating system. Its control parameters are gained through design and experiments according to the certain operating condition. So it is difficult to keep the stability of the bulb turbine units under isolated network condition. Particularly when operating condition changes largely, stability operation become more difficult. It is because that work head of bulb turbine units is low, the opposite varity of water head is big and turbine output power changes grequently. With the appearance of the intellectual control technology, the intellectral control strategies based on fuzzy logic, neural network and genetic algorithm is more and more applied to every industrial fields. So hydro-turbine regulating system control also faces new opportunity and challenge. Therefore, the paper establish two kinds of intellective controllers, which are fit for bulb turbine units regulating system, through respectively combining CMAC neural network and genetic algorithm with PID control strategy.This paper fistly recommends hydropower resources and status quo of hydropower development in China, and control characteristics and technology advantages of bulb turbine units, and the meaning of studying control strageties. Immediately, the study status quo of control technology of Hydro-turbine regulating system is briefly introduced, and various control strategies are analyzed. Then, The method to establish and simplify the mathematical model of bulb turbine units is analyzed in detail, and conventional PID controller is set up. Aim at the shortage of the conventional PID controller, the CMAC-PID is proposed. It is based on the conventional PID controller and combined with CMAC neural network. Simulation studies show the effectiveness and good performance. And then the basic principle and characteristic of the genetic algorithm is analyzed. Based on genetic algorithms, an optimization design of PID controller parameter is developed. The simulation results show that this method does not depend on the initial values of parameters and the optimal search efficiency is high. The end, all the work and research results are summarized and the shortage and further research work are pointed out.