Research On Transformer Fault Diagnosis Based On Quantum Genetic Neural Network

In the power system,the role of the transformer is particularly significant,and its operating status is directly related to the stable operation of the power system.The sign of the transformer failure and the type of fault have a complex nonlinear relationship,so the traditional method of fault diagnosis for the transformer has some shortcomings.Artificial neural network has the ability of self-learning and self-adaptability and can deal with nonlinear problems well.Therefore,the artificial neural networks is selected to judge the operating status of the transformer.In this paper,the traditional BP neural network was improved.The quantum theory was applied to the BP neural network to form the quantum neural network.The utilization of superposition state of the quantum theory made the neural network have fuzziness and better ability to process complex data.Taking the characteristic gas dissolved in oil as input and the fault type of transformer as output,the fault diagnosis model of the quantum neural network transformer is constructed.Data was collected for the training of the quantum neural network.Finally,the fault diagnosis of transformer was carried out by using the trained quantum neural network.The simulation results showed that the proposed algorithm solved the problem that the traditional BP neural network was apt to fall into the local minimum.After more than 400 generations of training,the algorithm converges quickly to get the optimal solution,and the correct rate of transformer fault diagnosis increased from 78% to 86%.During the experiment,it was found that the above method still has the phenomenon of misdiagnosis,and the training speed still needs to be further improved.This paper proposed a fault diagnosis method of quantum genetic neural network transformer.Quantum genetic algorithm was the combination of quantum computing and neural network,completed the evolutionary search by updating the quantum gate function and quantum gate,adjust the optimization results and evolutionary speed by adjusting the size of the rotation angle.The RBF neural network was optimized by quantum genetic algorithm and the fault diagnosis of the transformer was carried out by using the optimized network.The simulation results showed that the quantum genetic neural network has greatly improved its convergence speed.When the training reached more than 50 generations,it converged rapidly to get the optimal solution.Moreover,the generalization ability of RBF neural network has also been well improved and the correct rate of fault diagnosis of transformers has been raised from 86% to 93%.