Fault Diagnosis of Oil-immersed Transformer Based on Improved Seagull Optimization Algorithm to Optimize Wavelet Neural Network

Abstract

Aiming at the problems that the fault diagnosis accuracy of oil-immersed transformers in power systems is low and the diagnosis results are difficult to cover the entire transformer. A new fault diagnosis algorithm is proposed. ISOA is used to improve the extraction effect of fault data features, and WNN algorithm is used to realize the fault information classification and prediction. The fault diagnosis application results show that. The performance test results of the two algorithms are excellent, and the correct rate of fault information feature extraction is 94.65%, which is better than other algorithms. Predict and analyze the failure to reduce the difficulty of maintenance by technicians. The research content will effectively solve problems such as fault diagnosis of oil-immersed transformers, and has great value for the development of power systems.