FAULT PREDICTION AND CONDITION-BASED MAINTENANCE METHOD FOR URBAN RAIL TRAIN BOGIE

摘要

Provided is a fault prediction and condition-based maintenance method for an urban rail train bogie. An optimal longevity distribution model of a framework, a spring apparatus, a connecting apparatus, a wheel set, an axle box, a driving mechanism and a basic braking apparatus of a bogie is determined using a method based on survival analysis. A reliability feature function of each subsystem is obtained. Then, a fault rate of each subsystem of the bogie is calculated using a neural network model optimized by an evolutionary algorithm. Finally, the fault rate and the number of safe operation days of each subsystem of the bogie are taken as concomitant variables so as to perform proportional risk model modelling. On the basis of cost optimization, threshold values and control limits for condition-based maintenance of a bogie system are obtained.