Fault Diagnosis for Power Transformer Based on Uniform Distribution of Grasshopper Optimization Algorithm Combined with Neural Network

The fault of power transformer endangers the safety of power system. In order to improve the fault diagnosis accuracy and avoid the inherent defects of traditional algorithms, an improved grasshopper optimization algorithm (IGOA) with back propagation (BP) neural network is proposed for power transformer diagnosis in this paper. Firstly, the IGOA optimization algorithm was improved and combined with BP neural network to optimize the model parameters. Then the optimized network model is substituted into transformer fault identification to improve the accuracy of fault diagnosis. The comparison between the prediction results of different models shows that the proposed model and method can accurately predict transformer faults.