Schedule Optimization Methodology Of Customized Bus Based On Travel Demand Analysis

Urban public transport,as one of the most important modes to alleviate traffic congestion in cities,has to be continuously developed to build environmentally friendly cities.Recently,the rapid development of urban customized bus has become a new public transportation mode based on Internet technology.It has the advantages of comfortable,safe,convenient and efficient.It can effectively improve the quality of public transportation service.In recent years,a few cities in China have tried to develop the customized buses,and it has shown a strong market demand.Although the customized buses have successfully developed for several years the wide application of it is still facing challenges in different cities.In some cities,the development of customized buses is in good condition.The number of customized bus routes increases annually.At the same time,however,in some other cities,the development of customized buses is distinctively hindered.This phenomenon suggests that the lack of systematic investigation will result in significantly different results even with superficial same running conditions.As a result,it is necessary to explore a background theory to understand the demand prediction,route planning and vehicle optimal scheduling by using the specific travel preferences of local residents,the strength of government policy support,the capital investment of internet companies as key factors.This study focuses on exploring the fundamental theory of the customized bus planning,management and operation.By looking the customized bus line scheduling as the research objective,the research follows the steps of “analysis of travel behavior?forecast of travel demand?timing scheduling optimization?real-time scheduling optimization?case analysis” to understand the basic theoretical knowledge and thus improve the running and cost efficiency,increase the operation benefit,and optimize the limited public transportation resources.By investigating the demand and classifying the types,the customized bus was divided into the scheduled customized bus and the real-time customized bus.In order to distinguish them from the conventional bus,the orientation of the customized bus under internet reservation has been clarified by refining influencing factors.In this study,via the design of travel survey questionnaire and the application of regression analysis method,the influence factors of travel behavior were systematically analyzed.By using the structural equation model,the causal path and the characteristics of the customized bus was analyzed according to travel demands.Based on the bus IC data and GPS data,the corresponding data description and analysis are carried out.The improved CLIQUE clustering algorithm was used to forecast the demand for customized bus travel through the bus ID data and the GPS data analysis.Based on the bus GPS data,the intersection delay,the links travel time,and the recovery method of the scheduled and the customized bus was established.By using the customized bus stop time and the recovered inter-station travel time as the dividing indicators,an orderly clustering method was adopted to build the operation time division algorithm,and the calculation method of the customized bus departure frequency was proposed.In order to improve the real-time dynamic flexibility of customized bus and face the peak travel demand,a customized real-time scheduling optimization model based on travel cost was established.Different routes and schedules were optimized according to the departure time,expected arrival time,and on-off station submitted by passengers.By balancing the travel time,waiting time,delay and economic cost,the rights and interests of travelers are guaranteed.The simulation platform is built with an example for verification,the results show that the customized bus can save the average travel time,average waiting time and a late penalty.In order to improve the running efficiency of the customized bus,a real-time vehicle scheduling method with passenger time window constraint was developed.The passenger loss caused by bus arriving early and arriving late at the station was transformed into equivalent operating mileage.Taking the total operating mileage of several buses as the target,a vehicle scheduling optimization model is established considering passenger station constraints and bus capacity constraints.By constructing the virtual source station,the scheduling problem of multiple buses is transformed into the traveling salesman problem,the greedy algorithm is used to solve the problem,and the optimal solution is given based on the genetic algorithm.The case study shows that the optimal model can output a reasonable vehicle scheduling scheme,and the model solving method based on the greedy algorithm and genetic algorithm design can effectively reduce the comprehensive running distance of customized bus routes.The research results provide key theoretical guidance for the planning,management,and operation of the bus company's customized bus routes.At the same time,it provides a strong theoretical support for the “public transportation priority” orientation advocated by the government and improves the real-time,accuracy,and efficiency of the urban transportation management.To further promote the transformation and upgrading the “Internet + transportation” from traditional industry to modern service industry,this study provides a theoretical support and scientific basis for the development of green,convenient,cost-effective,and sustainable public transportation mode.