On-demand Bus Route Generation And Operation Optimization Based On Random Users' Demands

At present,the main commuter modes of transportation in China are traditional public transportation,taxis and private cars.The cost of traditional public transportation is low,but its travel time is long and experience is poor.Taxi and private cars have short travel time and good travel experience,but their costs are high.Traditional public transportation and taxis/private cars are the two “extremes” of commuter transportation.The on-demand bus is a “more economical and more comfortable” mode.It improves the service quality of public transportation to a certain extent.However,the route planning and operation of opened ondemand bus are mostly based on experience.There is no reliable theoretical to support them.Therefore,based on the research of on-demand public transportation,a theoretical study on the route generation and operation optimization of on-demand bus is made and checked by examples.Based on the characteristics and connotation of on-demand bus,the key elements of ondemand bus system are analyzed.Then,the operation mechanisms under three cases are proposed.The route planning and operation optimization models are established under different operating mechanisms.(1)The case of fixed starting points and fixed end points: The optimization model with the maximum bus income and the minimum passengers' waiting time is established,and the model is checked by the online car data in Beijing.(2)The case of flexible starting points and fixed end points: The starting area is divided into several communities.The optimization model with the shortest running distance,the shortest waiting time,the shortest on-bus time and the maximum income is established,and the model is checked by the online car data in Chengdu.(3)The case of fixed starting points and flexible end points: The ending area is divided into several communities.The optimization model with the shortest running distance and the shortest on-bus time is established,and the model is checked by the online car data in Chengdu.Finally,the results are analyzed.Compared with traditional public transportation,the proposed on-demand bus operation mechanisms can greatly reduce the travel time at the cost of a small increase in travel expenses.Compared with taxi/private mode,the proposed operation mechanisms can greatly reduce travel expenses at the cost of a small increase in travel time.The research of this paper can lay a certain foundation for the route planning and operation optimization of real-time response on-demand bus.It can also provide some theoretical basis and support for the practical operation of on-demand bus.