Bearing Fault Diagnosis using Graph Sampling and Aggregation Network

Abstract

Bearing fault diagnosis technology has a wide range of practical applications in industrial production, energy and other fields. Timely and accurate detection of bearing faults plays an important role in preventing catastrophic accidents and ensuring product quality. Traditional signal analysis techniques and deep learning-based fault detection algorithms do not take into account the intricate correlation between signals, making it difficult to further improve detection accuracy. To address this problem, we introduced Graph Sampling and Aggregation (GraphSAGE) network and proposed GraphSAGE-based Bearing fault Diagnosis (GSABFD) algorithm. The original vibration signal is firstly sliced through a fixed size non-overlapping sliding window, and the sliced data is feature transformed using signal analysis methods; then correlations are constructed for the transformed vibration signal and further transformed into vertices in the graph; then the GraphSAGE network is used for training; finally the fault level of the object is calculated in the output layer of the network. The proposed algorithm is compared with five advanced algorithms in a real-world public dataset for experiments, and the results show that the GSABFD algorithm improves the AUC value by 5% compared with the next best algorithm.