Data Link Fault Location Model Based on Machine Learning

Abstract

As the scale of national grid data continues to grow, data link fault location becomes more and more important. The traditional fault location method has many shortcomings due to its cognitive and technical limitations. With the rapid development of artificial intelligence technology today, using machine learning technology for data link fault location has become an effective method. Therefore, this paper proposes an autoencoder-BP neural network model, which uses the autoencoder to extract data features, and then uses the BP neural network for classification. Finally, it is proved through experiments that the combination of the two deep learning algorithms can effectively improve Accuracy of fault location.