Expert System Research Of Power Transformer Fault Diagnosis Based-on DGA

Power transformer is one of the most important electrical equipments in the electric system. Its operating state attaches importance to system's safety directly. In addition, it is a significant issue for electrical department to find potential faults of the transformer so as to keep it operating safely. Therefore, the fault diagnosis technology is available and reliable to operate and maintain the transformer.First, The paper has deeply analyzed the principle of using gas dissolved in oil as character to diagnose transformer fault and the methods of fault diagnosis. Based on these theories, an expert systerm based on neural network was developed. Neural network was one of the knowledge modules in expert system, which as a principal diagnosis module, the knowledge modules which based on the fuzzy mathematics for the diagnosis module and the knowledge modules which based on modified IEC tri-ratic method.This paper introduces the BP neural network model structure and mechanism for learning, address the problems such as BP neural network convergence slow and difficulty convergence of the fault diagnosis system. The activation transfer function, the different neurons function and training function of neural network were simulated respectively, and also give analysis and research. L-M method as a training function effect best results in the numerical optimization algorithm.In the part of constructing fault diagnosis expert system in this paper, on the issue of program design, calling Matlab neural network toolbox in VB directly. This system has many good characteristics such as the friend interface, easy operation and maintenance, calculating fast and accurate and so on.