Hydraulic Power Station Dynamic System Models Identification

Along with the development of the power system in our country, the accuracy of grid analysis computing and the precision of automation have been set higher requirements; also, with the building and operating of large-scale hydro power stations and pumped storage plants, the water power proportion in our country is increasing year by year. Establishing accurate hydraulic power station dynamic system models have a great significance in power system analysis and designing governors with better control strategies. But the hydroturbines' characters are complex, the status of the operation changes greatly and frequently, and the conduit systems are often quite complicated. So it is difficult to establish the mathematical models totally depended on mechanics, electricity, and other physical axioms. Therefore, research on establishing accurate mathematical models of hydraulic power station dynamic system has great significance.Methods, characteristics, and experiment schemes of hydraulic power station dynamic system models identification have been firstly researched in this thesis. Then, applications of Least square algorithms, Genetic algorithm, and Pattern search algorithm to hydraulic power station dynamic system models identification have been researched.In hydraulic system linear models identification, a method using field test data under step disturbance of given power has been proposed in this thesis. In simulation experiment, results of identifying models using step disturbance of given power signal, white noise signal, and pseudorandom signal as input signals have been compared. Feasibility of using step disturbance of given power signal as identification input signal has been validated. In identification experiment of ShiSanling pumped storage power station, reliable results has been achieved by using step disturbance of given power signal. In addition, a method of identifying hydraulic system linear models based on random load fluctuation has been researched. Random load fluctuation could be considered as white noise disturbance signal and provide adequate identification conditions when the hydro generating set under grid connected condition. In identification experiment of ShiSanling pumped storage power station, results of identification using filed test data under random load fluctuation and step disturbance of given power have been compared and validated by variety methods. After having...