Research On Demand Responsive Customized Bus Real-Time Route Generation

With the acceleration of the urbanization process,the population continues to concentrate in the city,the urban area continues to expand,and the travel needs of urban residents continue to tend to be randomized and diversified,putting forward new demands and new challenges for the urban public transportation system,so major cities have begun to develop customized buses.However,the current customized buses are mostly static reservation type,their service scope and service population are relatively fixed,the attraction to passengers is not high,and the profit margin of bus companies is low.With the rapid development of big data and Vehicle Networking technology,customized buses break down technical barriers,and increase the operating costs of enterprises while meeting the random travel needs of passengers and reducing passenger travel time,so customized buses are gradually developing in the direction of responsive customized buses that can meet the needs of random travel of passengers.Demand-responsive custom bus line generation is a key link for providing services to passengers and bringing benefits to enterprises,so this paper will study the real-time line generation problem of demand-responsive customized bus.This paper first introduces the definition,operation mode and service conditions of demand-response customized bus,determines the necessary conditions for customized bus to pick up passengers,and uses the structured model algorithm to calculate the OD amount of a single bus line based on the number of people getting on and off,combined with Chengdu Tianfu The travel data of 32 bus lines in the new area is used to calculate the points of high passenger travel demand,and as fixed stations in the road network,the two-step optimization location method of customized bus ride-sharing stations is used to determine the passenger ride-sharing stations.Then,based on the travel cost of on-board passengers,the penalty cost when the vehicle does not arrive within the time window,and the vehicle operating cost,a weighted method is used to establish the initial route objective function of the demandresponse customized bus.Penalty cost when serving passengers,establish real-time route objective function.Finally,a line control coefficient constraint is introduced into the real-time line model to control the total length of the vehicle’s travel line,and the NSGA-II algorithm is used to solve the demand-response customized bus line generation model.At the end of the paper,a calculation case is designed to randomly generate a number of travel requirements within the Tianfu New District of Chengdu.The requirements are divided into two parts: advance submission and real-time submission.For the travel demand submitted in advance,5 demand-response customized bus initial operation routes are generated in the road network,and passengers on each route can receive punctual services.For the travel demand submitted in real time,the real-time response is made on the basis of the 5 initial lines.It is calculated that when the line control coefficient is not added to the model,line IV and line V will have a large delay,especially the delay time of line V.Reached the 990 s.When the line control coefficient is added to the model and the value is 1.2,the objective function of the model,the total length of the vehicle travel line,the total number of vehicle services,and the delay time for vehicles to arrive at each station are in the optimal range.And when the service area of the demand-response customized bus is the outer area of the city,if the total length of the initial route has exceeded 40 km,the bus is no longer suitable for responding to the realtime travel demand of passengers.