Travel Research And Simulation Analysis Of Dockless Shared Bikes Based On Spatio-temporal Flow Model

In recent years,due to the continuous growth of urban population,the problems of“urban diseases”such as environmental pollution and traffic congestion have become increasingly prominent.To alleviate urban problems and realize the sustainable development strategy of cities,the state strongly advocates “low-carbon”and “green”modes of travel.The shared bikes was born.In just a few years,the shared bikes has developed rapidly.From the beginning of the development of docked shared bikes to the more convenient dockless shared bikes,mobike and ofo have the most extensive development of the two dockless shared bikes brands,which have the greatest impact on society.However,due to the explosive development of shared bikes,the management shared bikes lags behind,which makes the problems of the shared bikes system in the actual operation process especially prominent.At present,the development of the shared bikes market has shifted from the explosive growth stage to the stable development stage,but the management of shared bikes is still outstanding.Relevant government departments and major shared bikes operators are actively seeking more reasonable management models to achieve healthy and sustainable development of shared bikes.Among them,based on shared bikes riding big data,using data-driven approach,there have been many research results in system planning,forecasting and rebalancing of shared bikes.Based on the time series position data of Beijing mobike dockless shared bikes,this paper proposes a method of urban space division based on point of interest clustering.Then based on the results of the division,an empirical study is carried out on the data of dockless shared bikes.The travel characteristics of shared bikes are analyzed from different dimensions such as time and space,and the reasons for the imbalance of shared bikes are discussed.According to the spatio-temporal flow law of shared bikes,a spatio-temporal flow model of dockless shared bikes is constructed.Finally,based on the spatio-temporal flow model,a preliminary study on the scheduling problem of dockless shared bikes is carried out.The main work of this paper includes the following points:(1)An urban space division method based on point of interest clustering is proposed.In order to study the flow of dockless shared bikes between different areas of the city,it is necessary to divide the urban space.According to the agglomeration effect of interest points of dockless shared bikes,this paper proposes a method of urban space division based on clustering of interest points,so as to obtain the accumulation area of dockless shared bikes.The method is compared with the commonly used uniform grid method,and the urban space division method based on point of interest clustering has better segmentation effect.(2)Empirical research on dockless shared bikes data.According to the results of urban space division,this paper first analyzes the travel characteristics of shared bikes from different dimensions such as time and space,and finds the spatio-temporal flow law of shared bikes.Then,the reasons for the imbalance of shared bikes are further explored,and the three aspects of capacity,circulation and dispersion of the shared bikes gathering area are analyzed.(3)Construct a spatio-temporal flow model of dockless shared bikes.Based on the results of the empirical study of the dockless shared bikes data,this paper considers the two factors of the activity of the shared bikes gathering area and the distance between the gathering areas,and constructs the spatial flow distribution model and the travel time distribution model of the dockless shared bikes respectively.Thereby constructing a spatio-temporal flow model of the dockless shared bikes.Then,the paper uses the maximum likelihood method to evaluate the parameters of the model.Through the design simulation experiment,from the perspective of complex network,the evaluation index of the complex network are used to evaluate the simulation results,and the validity of the model is verified.Finally,a case study was performed based on the real-day shared bikes data utilization model.(4)Research on scheduling problem of dockless shared bikes.In this paper,based on the spatio-temporal flow model of the dockless shared bikes,the research on the scheduling problem of the dockless shared bikes is carried out.First of all,the article elaborates on the current research and methods related to shared bikes scheduling.Then,the paper mainly studies the scheduling problem of dockless shared bikes from two aspects: scheduling area division and scheduling path planning,and proposes a scheduling area division method based on modularity community detection algorithm and a scheduling path planning method based on ant colony algorithm.This paper studies the travel of dockless shared bikes,helps to grasp the flow characteristics of shared bikes,understands the travel rules of urban residents,provides a basis for scheduling of shared bikes,improves the management of shared bikes,and helps improve the urban traffic management system.