Study On Fault Diagnosis Of Transformer And Comprehensive Management System Of Preventive Test

Power transformer is the hub equipment of power system. Its safe and stable operation isdirectly related to the power supply reliability of the whole system. Therefore, it is necessaryto test transformer running state and accurately judge its testing health status according to testinformation in order to discover the latent fault and take targeted measures to deal with it intime. Thus, vicious damage of defect transformer that will cause power outages and othersafety accidents.is prevented.At present, transformer oil dissolved gas analysis (DGA) which is a more effective wayto early discover transformer internal fault has been widely used. For example, practicalengineering applications used IEC/IEEE recommended three ratio method to judge faultnature. But its has weakness of the incompletement code, too absolute coding boundary, nothigh diagnostic accuracy. Artificial immune optimization of fuzzy kernel clusteringmathematical model for the DGA analysis is introduced to determine the transformer faulttypes according to the characteristics of nonlinear mapping relationship of the gas content andfailure properties.The model considering sample data to be classified as antigen(Ag) and fuzzy clustercenter as a antibody (Ab) can avoid optimization process into a local optimal solution and thepremature problem so that it can quickly obtain the global optimal solution and achieve thegoal of classification of transformer fault diagnosis by using parallel group search strategy ofartificial immune to constantly update clustering center antibodies to identify antigen sampledata. The improvement process of antigen and antibody affinity is the optimization process offuzzy kernel clustering objective function. The method has been proved feasibility and gooddiagnosis accuracy through the example verification and compared with the transformer faultdiagnosis model of fuzzy clustering.In order to obtain reliable diagnostic data sources, a set of standardization, efficiency,information management system of the preventive test to get accurate test information isneeded to support the accurate mathematical model for transformer fault diagnosis.Considering the test management process, test technology and fault types of powertransformer are common with other power equipment, the system greatly enriched the scopeof the function by extending to primary equipment management and some other function to bea complete set of power equipment integrated management system of the preventive test. Thesystem bases on B/S architecture used J2EE technology and MySQL database to fulfillregulation and work instructions, process management, data entry and query, data analysis and evaluation. It can be convenient and efficient to manage the whole test process..