Research Of Intelligent Diagnosis Method And Its Application In Transformer Faults

With the technical level, productivity and automation of modern large-scale equipment improves continually, the fault probability increases greatly which promoted the development of diagnostics of modern large-scale equipment.Power transformer as the complex and important equipment of the power system has a very significant influence to power system, enterprise's production and people's life. How to diagnose the transformer fault, and find out the fault reason quickly is a good way to increase work efficiency and lighten the economy losing. Therefore, it is significant to diagnose the transformer fault for ensuring the security, reliability, and efficiency of power system.Insulation faults account for most of the power transformer faults, and the technology of oil dissolved gas analysis is commonly used to online monitor the operating status of transformer. The indefinability of transformer fault reasons and test data make it difficult to use the common ratio method to satisfy the requirement of engineering applications. In recent years, artificial intelligence methods have made successful application both in the transformer fault diagnosis and other fields, and start a new way to overcome shortcoming of the traditional methods of transformer fault diagnosis.This paper summarizes the theory fault diagnosis of modern equipment and the shortcomings of traditional technology of transformer fault diagnosis first of all. In view of the successful application of artificial neural network technology and support vector machine algorithm in the field of fault diagnosis of modern equipment, this paper proposes two models. One is a model of wavelet neural network based on modified particle swarm optimization and the other is a model of least squares support vector machine based on relative transformation to preprocess the data. The application of them to diagnose the transformer faults achieves fairly good results.At last, this paper sums up the excellent performances and shortcomings of the two artificial intelligence diagnostic methods of this paper, and analyzes the prospects and development of artificial intelligence method in transformer fault diagnosis.