Fault Diagnosis For Power Transformer Based On Genetic Algorithm For Optimization Of Neural Network

Power transformers are key elements in power system. The operation reliability of power transformer is related to the safety and stability of power system directly. Fault of one power transformer may cause long interruptions in supply, costly repairs and loss revenue. Therefore, the fault diagnosis technology is available and reliable to operate and maintain the transformer.Dissolved Gases Analysis (DGA) is one of the main technology methods to diagnose the internal faults in transformers. Artificial Neural Network has many advantages, such as parallel distributed processing, self-adapted ability, association, memory, clustering, faults tolerance etc. It supplys a new way of acquiring knowledge, expression and illation, it can show out some mapped relation through finding out the optimal weights by training different information seriatim, which made it a proper method for the multi-process, multi-fault and multi-mode transformer fault diagnosis. The power transformer fault diagnosis using BP(Back Propagation) Neural Network theory were carried out, but these methods have some disadvantages and localizations.A combining algorithm for neural network training is formed by combining BP algorithm and genetic algorithm in the building process of neural network knowledge bank on fault diagnosis of a certain missile testing equipment, in order to overcome the shortcomings that BP algorithm is usually trapped to a local optimum and it has a low speed of convergence weights, according to the advantage of the globe optimal searching of genetic algorithm. This algorithm can effectively and reliably be used in the fault diagnosis of the missile testing equipment by comparing the two algorithms and analyzing the results of real examples.The results of experiment show that the compared with the Levenberg -Marquardt BP algorithm, the training epochs are decreased by 49.8% and 46.4%, The results explain that this algorithm can effectively improve the neural network convergence speed and decrease the training epochs. the model's fault diagnosis accuracy is above 96.7% and 93.3%,which shows that the model is proper, feasible and correct for the transformer fault diagnosis.