Research on Fault Diagnosis Method of Train Wheelset Based on Deep Learning and Big Data Analysis

摘要

With the continuous upgrading of the intelligent train wheel-set system, the number and types of the wheelset monitoring data have gradually increased, the research on the fault diagnosis of train wheelset is moving towards the era of big data. 11 is a geat challenge to collect and sort out the useful fault diagnosis information in the chaotic monitoring data of train wheelset. This paper proposed an improved fault diagnosis method of train wheelset based on the LSTM network and MapReduce framework, which takes advantage of deep learning and big data analysis. This novel method can directly process the raw monitoring data without any preprocessing or traditional feature extraction, also can process the large-scale data quickly and get a higher accuracy of diagnosis results.