Research On Fault Diagnosis And Prediction Technology Of Excitation System Of Large Generator Set

With the continuous expansion of the power grid scale and the increasing number of electrical power system operating equipment,the complexity of electrical power system operation has also appeared to increase.A safe and stable operation has become more and more important in recent years.As a result,the status monitoring technology,as well as the fault diagnosis technology in the operation process and electrical power system,have attracted increasing attention in the power grid.The excitation system is the core control equipment of the generator,which is directly related to the safety and reliability of power production.Therefore,it is far from enough to realize the basic status monitoring and safety protection functions of the excitation control system,but increasingly necessary and urgent to carry out real-time monitoring and rapid diagnosis of faults in the excitation control system of large-scale generators.In this paper,indepth research was performed for the fault diagnosis and prediction technology of the excitation system of large-scale generator sets.The specific research contents are as follows:Firstly,it analyzed the typical fault cases of the excitation system and researched the fault diagnosis methods,constructed the fault diagnosis of the excitation system by using the model reference method,and carried out fault diagnosis for the parameter setting error of the excitation regulator.The experimental results show that the method can quickly and accurately determine whether there is a failure in the excitation system.In addition,it studied the excitation system fault diagnosis method based on Fast Fourier Transform(FFT),constructed the excitation power unit model by using MATLAB,simulated the output of fault through simulation experiment,and analyzed the output waveform with FFT,thus diagnosing the location of the faulty equipment.Secondly,it designed the software and hardware models of the fault diagnosis device for excitation system on the basis of D5000 platform,introduced the hardware circuit and software model of the system in detail,analyzed the algorithm flow of the system and the design ideas of the main modules,and showed the main functions of human-computer interface and the running cases.This device provided technical support for the realization of the fault diagnosis and prediction function of the excitation system.The experimental results show that the system can run stably and effectively monitor and diagnose the faults of the excitation system online.Thirdly,an excitation system fault diagnosis device was designed in this paper by combining the model reference method and the neural network method for the defect that the model reference method could not accurately locate the faulty equipment.Instead of the amplitude characteristic parameter,the phase characteristic parameter of fault waveform FFT was used as the training input sample of the neural network to realize the fault diagnosis of a single excitation power unit.The experimental results show that the improved intelligent diagnosis model can effectively improve the accuracy of fault diagnosis.In practical application,most excitation control systems adopt the multi-bridge parallel power unit circuit.Therefore,the effectiveness and accuracy of fault diagnosis device were verified through the fault diagnosis of a doublebridge parallel excitation power unit and the characteristic parameters of thyristor failure prediction were determined through the analysis of the aging failure of thyristor,it built a fault prediction model based on the neural network,so as to realize the effective prediction of excitation system fault.