Abstract:

In order to dealing with transit scheduling problem with several lines, an improved transit driver scheduling model was presented. Its objective was to schedule drivers in such a way that the deadheading was minimized while the operational constraints such as work time and spread time were satisfied. An ant colony algorithm (ACA) was presented to solve transit scheduling problem based on principle of ACA used to solve vehicle routing problem (VRP). Improvement on route construction rule and Pheromone updating rule was achieved on the basis of former algorithm. An example was analyzed to demonstrate the correctness of the application of this algorithm. It is proved that ACA is efficient and robust in solving transit scheduling problem.

Key words: transit scheduling problem, ant colony algorithm, vehicle routing problem (VRP)