Research on Entropy based Corrective Maintenance Difficulty Estimation of Metro Signaling

Well maintained signaling equipment is critical for the safe and efficient metro operations. Although periodic preventive maintenance is regular for all operators, signal failures always occur from time to time. The duration of signal failures can lead to very different effects on metro operations, from no impact to disastrous traffic disruption. Therefore, duration estimation is crucial for the emergency treatment of all related groups, including dispatchers, station, mechanical maintainers, etc. This paper presents a model to evaluate the difficulty of failure disposal based on information entropy theory and PSO optimization. Furthermore, the duration of corrective maintenance is classified according to the statistics characteristics of difficulty index, which can be used to support the decision making of operation recovery strategies. The model is applied for the case of Beijing metro lines and results are discussed.