Research On Fault Diagnosis Of Oil-immersed Transformer Based On GWO-BP Neural Network

With the continuous development of national economy,the power demand of China rises year by year.Oil immersion power transformer is in high load operation and extremely vulnerable to power supply fault.At present,the maintenance of oil-immersed power transformers is mainly carried out by periodic maintenance strategy,but the maintenance burden is increasing with the increase of power supply.In order to solve the transformer maintenance dilemma,the best method is to carry out state maintenance and online monitoring of planned maintenance.With the development of transformer maintenance technology,oil-immersed transformer maintenance means are no longer limited to electrical methods,has gradually tended to be combined with other disciplines,the temperature,acoustics,optics and other diversified technology fusion for detection and diagnosis.Therefore,based on BP neural network,this paper proposes an improved BP neural network fault diagnosis method for oil-immersed transformer,and designs an on-line fault diagnosis system for oil-immersed transformer.Firstly,the thermal fault and discharge fault mechanism of oil-immersed transformer are summarized through literature research.According to the gas production characteristics of various faults,the fault diagnosis algorithm of oil-immersed transformer is constructed by BP neural network.The research shows that the diagnosis results are easy to fall into the local optimal solution,resulting in a significant decline in the diagnosis accuracy.Then,in order to solve this problem,GWO algorithm is introduced to optimize,and the diagnosis model of GWO-BP network is constructed.Secondly,the simulation experiment and field application experiment are used to compare and verify the diagnosis methods.The experimental results show that the GWO-BP network can accurately diagnose the faults of various oil-immersed transformers.Compared with IEC three-ratio method and site investigation,it is found that the fault type diagnosis of oil-immersed transformer based on GWO-BP network has higher accuracy than IEC three-ratio method,and can meet the diagnosis needs of actual substation.Compared with the BP network,it is found that the GWO-BP network can avoid the problem that the BP network is easy to fall into the local optimal solution in such nonlinear problems as transformer fault diagnosis.Finally,on the basis of GWO-BP neural network,the design of on-line fault diagnosis system of oil-immersed transformer is completed,and the on-line fault diagnosis system of oil-immersed transformer is realized based on real-time gas monitoring data.