Research On Online Monitoring And Fault Diagnosis Of High Power Excitation System Based On Multi-source Information Fusion

With the rapid development of the world economy and the continuous progress of the power industry,the single unit capacity of the generating unit has become larger and larger,which also puts forward higher standards for the safety,reliability,economy and other aspects of the unit operation.As the core of generator control,the performance of excitation system directly affects the power quality and reliability.Especially when the excitation system fails,if the failure can't be found and eliminated in time,it is very likely to cause serious electric accidents,so it is necessary to conduct online monitoring and fault diagnosis research on the excitation system.After consulting a large amount of literature,this article first elaborates the current research status of on-line monitoring and fault diagnosis of the excitation system,and introduces the working principle,basic structure and function of the excitation system.Secondly,a simulation model of the power unit of the excitation system is established on the MATLAB simulation platform,and the open-circuit,short-circuit fault and normal operation of the thyristor are simulated on this model,and the detailed analysis of each fault waveform is performed.Based on the power unit of a single excitation cabinet,a simulation model of the excitation power unit operating in parallel with multiple bridges was built,and the open-circuit fault of the thyristor was simulated,and the characteristics of different types of faults were classified and analyzed.This paper also models and simulates the more complicated and serious demagnetization faults of the excitation system,and analyzes the fault characteristics such as the excitation voltage and the excitation current.Based on the simulation analysis of three typical faults of the excitation system,a fault diagnosis strategy for the excitation system based on multi-source information fusion is proposed.Then by analyzing the failure mechanism and extracting the fault features,the BP neural network fault diagnosis model is constructed on the MATLAB platform by using the extracted fault features,and the open-circuit and short-circuit faults of the single-excitation power unit and the open-circuit faults of the multi-excitation power unit And diagnosis of loss of field fault.Finally,the excitation system online monitoring and fault diagnosis system is divided into two levels,one is local level,the other is plant level.At the local level,the diagnosis and monitoring of the simple faults such as the open circuit fault and the short circuit fault of the thyristor in the excitation system are realized,and the results are uploaded to the background;at the plant level,the MATLAB intelligent fault diagnosis system is established,and the BP neural network fusion algorithm is used to diagnose the excitation system loss of excitation fault and other complex comprehensive faults,and the diagnosis results are displayed.