Research On Parameter Identification Of The Hydro Generator's Excitation System

With the fast acceleration of economic growth and industrialization,we have put forword higher requirements to the sustainable development of energy in China. Our country has rich hydroelectric energy, and the total installed capacity accounts for 20% of the electric power system. At the end of 2015, the total installed capacity of hydropower units in China reached 211 million kilowatts, and it is planned that the installed capacity will rise to 380 million kilowatts for 2020. With the expansion of scale of China's hydropower, the stability of power system dynamic performance is becoming increasingly apparent. As an important part of hydroelectric generator, excitation control system's performance will directly affect the operating characteristics of hydroelectric generator. Therefore, combined with the characteristics of excitation control system of hydro-electric power, the study of parameter identification method is of great significance. Getting the real-time operation of hydro-electric power unit of excitation control system parameter through parameter identification method can helpthe system on-line monitoring and condition assessment process, so the reliability and operating safety and stability function will beimproved.Based on excitation control system parameter identification problem,this paper comes up with the time domain identification method of least squares.Besides by introducing a recursive method to optimize the algorithm, it designsb excitation control system parameter identification method which is based on generalized recursive least squares linear.It is known by us all that traditional identification methods cannot identify linear excitation control system of parameters, so this thesis makes full use of genetic algorithm to achieve the implementation of nonlinear excitation control system parameter identification. According to mathematical model of excitation control system features on fitness function in genetic algorithms, t the identification accuracy of genetic algorithms can be improved through the probability of selection and genetic operators, at last, the universal design of excitation control system parameter identification method in the various links can be achieved. Lastly, by using MATLAB/SIMULINK software for simulation and testing the identification method and evaluating the excitation control systemon the basis of the state performance index to estimate the accuracy of parameter identification of results under operating condition. All these above prove that the model can accurately describe the effect and dynamic performance of excitation control system.What is more,it further confirmsthat through parameter identification for excitation control system for on-line monitoring and condition assessmentis feasible.