Research On Parameter Identification Of Excitation System Of Synchronous Generator

With large capacity units and UHV transmission lines put into use, it become the main characteristic of modern power system that introduce the modern control theory and computer technology into the power system, but whether control or calculation are required to establish a reliable mathematical model. Excitation system parameters is one of the four parameters of power system, so an accurate appropriate excitation system model has a great effect on improving the system stability and the operating conditions, which requires researchers to conduct in-depth research on excitation system. Early people use a constant potential excitation model, which can meet the practical needs in a certain extent. But with the introduction of advanced control systems into the system, this model can no longer simulate the real system.So this paper identifies parameters by using simulation data in order to achieve real values of the parameters.This paper first introduces the structure model of synchronous generator excitation system and several common methods of parameter identification, then focuses on the research of identifiability of the Parameter, analyses the relationship between identifiability of the parameter and system and algorithm and system input.Based on time-domain identification, this paper identifies the parameter of two excitation systems and it identifies the parameter of a single segment of three different links of the first system and the whole link of the second system.We found that the identification is reliable and high precision in single segment identification, but the results are not stable and has a lot of discrete in the identification of the entire link. Which shows that it can’t identify all of the parameters by the measured input and output data, namely the result is not unique. Through the analysis of the parameters found it exit implicit functions between these parameters, and called these associated parameters. Therefore, the most important is the identification of associated parameters. It selects associated parameters representative through the method of trajectory sensitivity, and given the representative typical value to lift the implicit functions and identify all parameters. At the same time, we can increase the number of known conditions by increase the intermediate measurement to increase equations of parameters identification, which can lift the the implicit functions and identify all parameters too. It uses this method to identify a excitation system and simulation analysis with the original system, which found the identification result is stable and higher precision and it achieves a relatively good result.