The subway engineering department regularly operates track inspection vehicles to detect the state of the tracks, which is crucial for residents' safe travel. The operational path of track inspection vehicles mainly relies on expert judgment, which is not only a timeconsuming practice but is also ineffective. To address the shortcomings of the current lack of systematic planning for the operational paths of track inspection vehicles, this study, set against the backdrop of the urban rail transit network, constructs a largescale subway inspection vehicle routing optimization model named Urban Track Inspection Vehicle Routing Problem (UTIVRP), under the conditions of a complex network. Considering the characteristics of subway networks, a cultural genetic algorithm with a special encoding method is designed and validated using practical examples from the Beijing subway. The computational results indicate that under the conditions of meeting the established inspection requirements, the optimization solution can not only reduce the idle mileage of vehicles by 48.88%, but also decrease the maximum deviation rate of the network's inspection interval by 93.33%.

urban rail transit  /  inspection velaicle  /  track inspection  /  routing problem  /  memetic algorithm