Research On Condition Monitoring System Of Generator Excitation Transformer

Excitation transformer is an important part of generator excitation system.Due to its large variable ratio,complex working conditions and high load during operation,it is prone to failure or internal deterioration,therefore,monitoring its operating condition is an important part to ensure safe and stable generator operation.However,the current monitoring methods of excitation transformers in most power plants still have problems such as less data and poor real-time performance.To address the above problems,this paper researches and designs a set of condition monitoring and analysis system for excitation transformers,which is mainly as follows:(1)The hardware system for excitation transformer condition monitoring is designed and built,and the software for online monitoring and analysis of excitation transformer operation condition is developed.Through the analysis of the principle and working environment of the excitation transformer,the hardware system is designed,the transmission path of the electrical parameter data is built,and the real-time infrared images of the excitation transformer are collected in a non-contact measurement mode.The software system realizes real-time and comprehensive monitoring of excitation transformer oper ating status,including the host software and mobile software,in which the host software realizes remote real-time status monitoring and analysis,device parameter setting,data query and management,etc.;the mobile software acquires and displays cloud data,which is convenient for maintenance personnel to view the unit’s operating status anytime and anywhere.(2)A health assessment and prediction method based on Gaussian mixture model(GMM)and long and short term memory neural network(LSTM)is proposed for the deterioration trend existing in the operation of excitation transformers.The correlation analysis and removal of redundant variables were performed by maximum information coefficient theory(MIC),a GMM-based health benchmark model was established,a health deterioration index(HDI)based on the Mahalanobis distan ce was designed,and the multivariate prediction of HDI was performed using LSTM.The experimental results show that the designed HDI can accurately evaluate the operational health,and the prediction for HDI can detect the unit deterioration trend earlier,which can provide a reference for the periodic maintenance and overhaul of power plants.(3)In order to verify the reliability and functional integrity of the excitation transformer condition monitoring system,on-site installation and commissioning were completed,and the long-term operation data of the system were analyzed.The results show that the monitoring system can realize the functions of accurate data acquisition,reliable transmission and real-time monitoring,and the accuracy of system temperature measurement and analog acquisition can meet the requirements of the national standard,and the system can comprehensively monitor the operation status of the excitation transformer,and has good stability under long-term operation,which ensures the safe and stable operation of the excitation transformer unit.