On Line Monitoring And Fault Diagnosis Of Power Primary Equipment

Transformer in electric power system is one of the most important and valuable electrical equipment, its running state or not is directly related to the safety of the whole power system security and stability. To carry out the transformer condition monitoring and testing, and on the basis of studying the fault diagnosis method is currently very urgent work, is helpful for early detection of transformer latent fault, reduce the economic consequences of the transformer faults, improve the safety and reliability of power system operation, is an issue of great significance and practical value.In the study of transformer fault diagnosis based on dissolved gas analysis (DGA) in oil is the most effective way. Based on power transformer oil dissolved gas monitoring method, on the basis of thorough analysis of dissolved gas in transformer oil theory, constructs the support vector machine was optimized based on the searcher of power transformer fault diagnosis model, the main work includes the following several aspects:(1) In power transformer oil dissolved gas theory research. With object, the analysis of the power transformer of transformer internal gas source, analyzed the generating mechanism of dissolved gas in transformer oil. Further analysis when the transformer fault occur, the internal mechanism of gas and the basic characteristics of gas under different failure modes. Finally, the paper analyzes the common transformer fault diagnosis method based on DGA.(2) Based on search optimization algorithm of the support to the two model parameters optimization. In view of the SVM model parameter optimization problem, this chapter puts forward the SVM model parameter optimization method based on SOA. First of all, the support vector machine (SVM) theory is discussed in this chapter, according to the model parameter optimization problem, optimization method based on seekers algorithm is put forward, the basic theory of SOA algorithm and performance is analyzed, and the steps of the SVM model parameter optimization is discussed.(3) Based on SOA - power transformer fault diagnosis model of SVM. According to the characteristics of the power transformer monitoring data, put forward and construct the model of fault diagnosis based on SOA - SVM. Analysis on how to construct the model and process of fault diagnosis, an introduction to the basic process of the model, and analyze the key issues and clear.4) The transformer fault diagnosis model to empirical research. SOA - SVM model based on the established, using the actual monitoring data of transformer is analyzed to validate, the results show that the established model of transformer fault diagnosis based on SOA - SVM, has high discriminated accuracy, flavor and low misjudgment ratio, has the good scientific and application effect.