The Fault Data Mining of Supervision Equipment of Urban Rail Transit Based on Clustering

Based on analyzing characteristics of a large quantities data of equipment fault and alarm of supervision systems of urban rail transit, this paper takes fault characteristic and regulations of monitors as the object to identifying hazard factors that may cause accident. It presents reduction strategy of monitoring data mining and builds clustering feature tree of fault model, in order to figure out hierarchical clustering relations of failure factors. The mutual correlation of equipment and its belonging sub-system or composite system is explored, and algorithm of clustering feature tree is applied. The analysis result shows this method can find out hazard nodes that lead to damage and help make decision for safety management.