Fault Diagnosis and Prediction of Power System Based on Convolutional Neural Network

Aiming at the low accuracy of power system fault prediction and the limitation of the prediction model by parameters, this study proposes a fault diagnosis and prediction method for power automation systems based on convolutional neural network by optimizing the artificial intelligence algorithm. In the research, we simulated various fault situations in the power system, and established a power system fault model of convolutional neural network, and realized the prediction of power system faults through data preprocessing, model training and result evaluation. In this way, we can achieve accurate prediction and effective diagnosis of power system faults. Experimentally, our method can predict faults in power systems with an accuracy of 99.1%. Compared with other algorithms, our method not only significantly improves the accuracy rate, but also has a significant advantage in predictive effect. Therefore, the application of this method can effectively improve the effectiveness of power system fault prediction.