AUTOMATED VEHICLE SCHEDULING SYSTEM: A CASE STUDY OF METROBUS SYSTEM

Abstract

Multiple Depot Vehicle Scheduling Problem (MDVSP) is the problem of preparing vehicle schedules, a task of public transport companies. MDVSP is solved to decide on daily duties of vehicles emanating from multiple depots. It aims to minimize number of vehicles used and total deadhead kilometers. Since there are large number of trips to be covered by vehicles emanating from multiple depots, manually prepared vehicle schedules are usually far from optimality. Therefore, using automatic scheduling systems can reduce number of vehicles used and total deadhead kilometers. In this study, 6254 daily trips belong to 2014-2015 Winter Timetable of IETT General Directorate's Metrobus System are assigned to 476 vehicles instead of existing 496 vehicles by solving a Single Depot Vehicle Scheduling Problem (SDVSP) model, actually a reduced model of MDVSP. As further study, Lagrangian relaxation is going to be used to solve MDVSP to minimize total deadhead kilometers.