Research On Station Selection And Route Planning Of Customized Bus Based On Taxi Demand

With the continuous development of our country’s social economy and the rapid improvement of comprehensive strength,the number of motor vehicles has increased rapidly,leading to a series of problems such as urban traffic congestion and environmental pollution.Prioritizing the development of public transportation is one of the effective measures to solve these problems.With the application of "Internet +" technology,customized public transport and online car-hailing have emerged,enriching the content of the public transport system.At present,the scale of online car-hailing operations continues to expand,which better meets the individualized travel needs of citizens and enhances the ability of urban taxis to serve citizens.Therefore,explore the internal relationship between taxis and customized bus travel modes,study the use of taxi demand data to plan customized bus stops and routes is of great significance to innovate bus operation models,implement bus priority strategies,optimize the travel structure of urban residents,and alleviate urban traffic congestion.First of all,in this paper,the related theoretical basis of customized public transportation has been explained,then the characteristics of customized public transportation has been analyzed,the internal relationship between taxis and customized public transportation travel modes has been analyzed,and the preprocessing method of taxi operation data has been presented.Secondly,it analyzes and describes the location of customized bus stops,introduces the principle,process,characteristics and applicability of the DBSCAN clustering algorithm,and proposes an improved DBSCAN algorithm design,which is calculated from the distance between demand points and the selection of clustering parameters.The selection of clustering centers improves the algorithm in three aspects,and uses Silhouette coefficients to evaluate the clustering effect of the algorithm.Then,based on the set of primary selected stations obtained by the clustering algorithm,with the goal of maximizing the overall benefits of enterprises and passengers,a customized bus route planning model was constructed,and a genetic algorithm-based model solution method was proposed.Finally,using the taxi operating data in the urban area of Xiamen City to carry out case analysis,apply the customized bus stop site selection method and route planning model constructed in this article,and compare the K-means algorithm with the improved DBSCAN algorithm,and the analysis results in the algorithm’s advantages and disadvantages in the clustering effect and the selection of the clustering center(site location),etc.,obtain the final site selection and planned route.The results show that compared with the traditional Kmeans algorithm,the improved DBSCAN algorithm proposed in this paper is more accurate in site location location,which makes up for the shortcomings of K-means algorithm that is easily affected by noise points and the site location location is prone to offset,It also makes up for the shortcomings of the traditional DBSCAN algorithm that the cluster centers cannot be directly obtained.Compared with the traditional bus stop and route planning,the customized bus stop site selection and route planning method constructed in this paper is simpler,the solution result is more accurate,and it is more practical.Based on the taxi operation data,this paper takes the setting method of the city customized bus as the research object,and uses the improved DBSCAN clustering algorithm to study and construct the stop location method and route planning method of the customized bus.The research results have a certain supplement to the theory and methods of public transportation system planning.