Circuit Breaker Fault Diagnosis Method Based on Improved One-Dimensional Convolutional Neural Network

Abstract

Aiming at the problems of manual feature extraction and poor generalization ability of model in traditional circuit breaker fault diagnosis technology, a circuit breaker fault diagnosis method based on improved one-dimensional convolutional neural network is proposed. Firstly, the input feature sequence is adaptively weighted by self-attention mechanism to highlight the weight of important information; Secondly, 1  1 convolution layer and global average pooling layer are used to replace the full connection layer, which reduces the model training parameters, improves the training efficiency and prevents the phenomenon of over-fitting. Aiming at the problem of small number of data samples, the data is enhanced by Generative Adversarial Network. After adding the generated data to the original data, the accuracy of fault identification is further improved. The experimental results show that this method can effectively and accurately identify different fault types of circuit breaker, and verify the feasibility of its engineering application.

Keywords

• circuit breaker
• data enhancement
• global mean pooling
• one-dimensional convolution
• self-attention