Passenger Demand Prediction and Network Planning for Commuters' Bus Systems

Masaki YAMAI


Automobile traffic has various problems, such as congestion during the rush hours. The aim of the research is to plan the network of commuters' bus suppose that commuters' bus systems are operated by a cooperative project of companies in a local city of Sanjo, Niigata. The study predicts passenger demand that sifts from car to bus by using the discrete choice model. Mode choice probability was predicted by using the SP (stated preference) choice model. And bus passenger OD matrixes were computed and assigned on alternative bus network. The results show that one of commuters' bus network plans covering a suburban wide area of Sanjo is most effective assuming that car commuters should pay the parking charge to their working companies, and buses are operating every 20 minutes headway. This plan could be evaluated to be cost recoverable and financially feasible.