Design and Implementation of Association Rules in Fault Diagnosis of High-Speed Railway in China

Abstract

With the rapid development of China's high-speed railway network and the increasement of it’s operating mileage, the requirement for the fault diagnosis and safety of high-speed railway has been increasing significantly. In order to enhance the safety and reduce the accident rates, it is essential to analyze the factors that closely related to safety and accident of high-speed railway. Compared with the traditional fault diagnosis method, the association rule algorithm of data mining technology has the advantage of dealing with large scale fault data. In the present study, the parallel apriori algorithm based on MapReduce algorithm was proposed. Furthermore, a subsystem of association analysis regarding fault diagnosis, which was based on mining association rules and integrated with the parallel association rules algorithm, was designed by taking advantage of the sparsity characteristic of the fault data. The function of the designed subsystem was validated by actual operating data collected. The results show that the proposed fault diagnosis subsystem based distributed parallel association rules can be used in the fault diagnosis of China’s high-speed railway.