Attention Mechanism-Wavelet Transform Based Performance Monitoring of Maglev System

Maglev trains have the advantages of high speed, low noise, safety and environmental protection, and are triggering a research boom in the field of transportation. It is crucial to monitor the performance of the levitation system to ensure safety. However, the information content of the levitation system is huge but the density is low, which makes it difficult to extract its important features directly. Aiming at this problem, this paper proposes a performance monitoring method based on the framework of Attention Mechanism-Wavelet Transform. Attention mechanism is utilized to capture the feature fragments. Wavelet transform is utilized to extract the features of time series. Combined with machine learning to complete the performance monitoring. This study is successfully applied to the Changsha Maglev Express Line in China.