A Turn-back Fault Diagnosis Method for Urban Rail System Based on Spark
Siqi Ma, Runtong Zhang, Xiaochen Wang, Xin Wang · 2020
Urban rail transit industry has accumulated a mass of data about intercity railways. There are three types of turnback fault in urban rail transit industry data: automatic turnback, end changed automatically, and end changed intermittently. The three type of faults may cause serious accidents, which will damage people's life. However, the end changed automatically and the end changed intermittently are getting much less focus than the automatic turn-back. Recognizing those faults can improve the managers' working efficiency of urban rail transit and protect the safety of passengers. The Apriori algorithm has been frequently used for fault diagnosis under different circumstances, helping managers make effective decisions based on the turn-back fault association rules hidden in the past data. Spark, a fast and general-purpose computing engine, is designed for large-scale data processing. In the era of big data that the amount of data increases exponentially, the memory calculation method of Spark can improve the computing efficiency of the Apriori algorithm. In this paper, we resort to the Apriori algorithm based on Spark implementation to mine feature combinations that appear in message data frequently when faults occur in automatic turnback, end changed automatically, and end changed intermittently respectively.