Power Transformer Fault Fusion Diagnosis Method For Test Data And Symptom Phenomenon

Large power transformer is one of the most important and expensive equipment in the power system, its operation statements has a direct impact on the power system security. At present, the maintenance system transforms from regular maintenance to condition-based maintenance. The condition-based maintenance is based on comprehending the device operational status. To learn more about equipment status, those equipments status needed be analyzed and diagnosed. This paper is carried out in this context; its significance lies in providing technical support for the condition based maintenance of power transformers.Power transformer fault diagnosis is an issue with assessments of multi-parameter fault indicators, fault causes have a complex relationship with corresponding indicators, and many reasons need to be considered synthetically. The fault diagnosis for large oil-immersed power transformer is generally carried out through preventive tests. However, the preventive test data must be obtained in power off and maintenance time, and the collection amounts and accuracy of the working site data is limited. Yet, when the transformer goes wrong in the running time, there is always accompanying the variations of the appearance for color, sound, temperature and oil lever, etc. on some certain parts. That is, the symptom phenomenon for power transformer. It is a summary of a large number of practices and experiences, to some extent, it can reflect the transformer failures quickly and accurately. Therefore, an idea that integrates both the preventive test data and the symptom phenomenon in the transformer fault diagnosis is proposed. According to a large number of data which reflect the transformer operation states, the corresponding relationships of fault modes, fault characteristics indicators, and instances is build up through composition structure. The diagnostic results of the preventive test data and the symptom phenomenon will be fused, and so forth, the corresponding results will be obtained. The diagnostic results show that the method improves the fault diagnosis accuracy and enhances the fault diagnosis robustness. This paper mainly studies the following points.â‘ The transformer is divided into eight major parts, and the relationship between the parameter indicators and fault of each part is deeply analyzed.â‘¡By using fuzzy mathematical theory to test data, we use the undetermined coefficients method for the analyzed membership function of different electric test according to the relationship between the membership distribution and the actual physical meaning of the experimental data. Therefore, we avoid the randomicity of choosing different membership functions. For the DGA parameters, the failure probability is determined based on DGA decomposition energy. For the transformer symptoms phenomenon, quantify the qualitative indicators, and then use extreme value theory to determine the failure probability.â‘¢The paper gives transformer fault fusion diagnosis model. It can be used as a kind of effective mean through the given example. The indicator parameters are divided into the experimental data and symptoms phenomenon according to its composition. Finally, the DS evidence theory is used to synthesize the result.The given example shows that the established model can integrate various status information.It can be used as a kind of simple, accurate mean through the on-site application.