Institute of Transportation, MOTC

       In recent years, as global warming has become more and more serious, our government wants to promote electric buses in urban areas to reduce greenhouse gas emissions. Because electric buses have the characteristic of mileage limitation, which is different from traditional diesel buses, charging schedules must be taken into consideration when arranging electric bus schedules.
        This study proposes a simulated annealing algorithm-based approach to the electric bus scheduling problem. We address a single-depot single-vehicle-type electric bus scheduling problem with multiple objectives to minimize the number of vehicles, deadhead time and the number of charging piles. And each bus is not restricted to be fully charged for each time the vehicle travels back to the depot for charging.
        This study uses 95 bus trips of 5 bus lines in Hsinchu city to test the performance of the algorithm. The testing results show that the proposed algorithm can accept worse neighbors in order to avoid getting stuck in local optima and can only accept better neighbors at the end to find an approximate global optimum.