Research On Optimization Of Customized Bus Network

In recent years,the rapid development of information and science technology has led to a new wave of diversified transportation.Conventional buses with fixed operating routes,fixed operating schedules,and the characteristics of one-stop and one-stop cannot meet people’s travel needs,nor can they attracting more private car travelers,thus failing to achieve the goal of energy conservation,emission reduction,and low-carbon travel that we advocate.Therefore,a high-quality public transport service model “tailored” for passengers named customized public transport was born.Because of the characteristics of fast,efficient,comfortable,and economic,it quickly swept all major cities across the country.The network planning process is the basis of the Customized bus operation,and its rationality plays an important supplementary role to the urban bus network.In this paper,we first analyze the general background of the customized public transport and clarify the value and practical significance of the research.Then,we summarize the relevant research at home and abroad,and summarize some of the works and results that the researchers have completed.The development of this new type of public transport mode is an inevitable trend.Based on this result,we determine the main content of the research.Secondly,we introduce the relevant theoretical overview of the customized bus,including the definition and characteristics of the customized bus,the comparison of the customized bus with other modes of transportation,the objective analysis of advantages and problems,and the related standards and methods of the customized bus network planning.Then,according to the characteristics of the customized bus passenger travel demand data,the paper uses the improved K-means clustering method to plan the customized bus sharing stations.To improve passenger satisfaction,this paper takes the shortest total service time and the minimum penalty cost of the time window as the goal,and establishes a dual-objective customized bus network optimization model with mixed time windows.Taking into account that the same station may have multiple customized bus stops and the orderly pairing of bus stops,virtual nodes are used to convert multiple stops into a traditional single stop problem.By constructing paired and orderly chromosome codes to complete the data processing and design two optimization algorithms to solve the problem.Taking into account the disadvantages of genetic algorithms that are easy to fall into local optimality and early convergence,this paper combines the strong local search characteristics of the tabu search algorithm with the excellent global search characteristics of the genetic algorithm,and improves the local search ability of the genetic algorithm through the tabu search mutation operator.Finally,MATLAB simulation is used to analyze the feasibility of the modeling process and the solution process taking Lanzhou City as an example.The results of the genetic and tabu hybrid algorithm are better than those of the genetic algorithm through comparison.Total time,time window penalty costs,travel distance,and average vehicle rate have all been reduced.Moreover,the number of vehicles has also been reduced.,The rationality of the optimization method and the effectiveness of the solution algorithm are verified.