Abstract: Wheel tread is the key part of running part of urban rail transit vehicle, which affects the safety and comfort of the vehicle. In order to improve the economy of wheel tread lathing, by analyzing the influencing factors of wheel tread lathing quantity and integrating the data characteristics of on-board fault diagnosis system for running gear and trackside detection system, this paper studied and established a model of economic wheel tread repair based on fuzzy multi-attribute decision analysis, determined the fault type, fault degree and fault development trend of wheel tread based on expert opinions and various indexes determined by fuzzy membership degree, made comprehensive decisions to give the best lathing time and amount, intelligently guide the operation and maintenance of wheel tread, avoid excessive maintenance of wheel tread, prolong the service life of wheels and improve economic benefits.