Customized Bus Route Planning Based On Stochastic Time-depend Road Network

Under the background of the rapid increase in the number of cars in the country,the increasingly serious traffic congestion,the deteriorating environmental quality,the longer and longer commuting time and distance of commuters,and the more personalized and diversified travel needs of passengers.Develop of customized bus is an effective method to alleviate the above problems.The route planning of customized bus is a core part of customized bus services.Scientific and reasonable routes can effectively guarantee passengers’ travel experience and improve customized bus service quality.It is also an issue that needs to be considered when customized bus enter the market and achieve sustainable development.First,this paper analyzes the current traffic environment in china,and elaborates the research background and significance of this research.This paper analyzes and summarizes the current research status of customized bus route planning and vehicle routing problems under stochastic time-depend networks.It is found that there are few researches on customized bus route planning under stochastic time-depend network,so it is necessary to more accurately grasp the operation of customized bus routes Time,reduce the gap between the theoretical travel time and the actual travel time of the routes,and ensure the quality of passenger service.This paper focus on the time-depend network to research the customized bus roueting problem.Secondly,this paper introduced the theories about customized bus,multi-objective optimization problems,Multi-objective ant colony optimization algorithm,uncertainty problem handling methods and robust optimization.The theory are provided for the establishment of models and the design of algorithms in the following sections.Then use the static network as the background of customized bus route planning,taking into account the interests of both passengers and the customized bus operating company,and taking the shortest detour time for passengers and the minimum operating cost of all ODs as the objective function,and OD for the maximum detour time,vehicle load,and station service for constraints,a multi-objective route planning model for singlevehicle and multi-vehicle customized buses under a static network is established.After analyzing the characteristics of the time-depend network and the stochastic time-depend network,in view of the uncertainty of the vehicle travel time under the stochastic timedepend network,the stochastic time-depend is also established with the goal of the shortest detour time for passengers and the minimum operating cost of all OD robust optimization models for customizing single-vehicle and multi-vehicle routes under the network transform the route planning problem under the stochastic time-depend network into the route planning problem under the time-depend network,which reduces the complexity of problem solving.Finally,for the solution of route planning models under the two networks,this paper uses multi-objective ant colony optimization algorithm to solve them.Taking the actual customized bus stops and randomly generated passenger demand data in city B as examples,the single-vehicle and multi-vehicle routes of the customized bus under the static network and the stochastic time-depend network are respectively solved.The analysis of the planned cycling routes under the static network and the stochastic time-depend network shows that there is little difference in the mileage and operating cost of the customized bus line under the two network conditions,and the detour time and the running time of the line under the stochastic time-varying network are greater than The detour time and running time of a static network.Similarly,the static network and the multi-vehicle line under the stochastic time-depend network do not have much difference in mileage and operating cost.The total detour time and the running time of the four lines under the stochastic time-depend network are longer than those under the static network.