Using Shapley value for city bus route scheduling

Abstract

A rational and efficient tool is needed for the scheduling of public transport bus services that will satisfy the needs of passengers for different service time requirements. An intelligent urban bus system should not only be able to reduce carbon emissions and traffic congestion, but also to improve transport services. A model is proposed which computes Shapley values for seventeen different time periods in one day for a bus route. A fair allocation of buses based on the Shapley value is made which entails the minimum amount of deployment cost. This study applies data for the rapid transit bus system in the city of Xiamen in China to verify the Shapley value model. The simulation results show that management can easily use this framework to quantitatively allocate buses along routes within the seventeen time periods in one day and that the proposed approach is more efficient than current bus scheduling methods.