| In recent years,the dockless bike-sharing system has developed rapidly with its advantages of flexibility,green and low consumption,and sustainable development,which has greatly alleviated urban problems such as traffic congestion,environmental pollution and energy shortage,and effectively solved the last-mile problem at the end of travel.However,with the increase of the number of users and the types of bicycles,the system also revealed many disadvantages,such as disorderly parking encroaching on public road resources,aggravating traffic congestion,etc.At the same time,its environmental benefits need to be assessed and analyzed.In order to plan,design and manage the dockless bike-sharing systems in a sustainable way,it is of positive significance to study the planning and location of the electronic fence and the analysis of environmental benefits.The main contributions are as follows:(1)An electronic fence planning framework based on the Maximum Coverage Model with Complementary Coverage(MCMCC)is proposed.The main purpose of this framework is to plan the electronic fence of the dockless bike-sharing system in the complex shape area,automatically determine the capacity and accurate location of the electronic fence,and improve the site coverage.Firstly,the city is divided into small scale areas according to the spatial parking demand distribution using the dynamic land subdivision algorithm.Then,a mathematical model is established with the goal of maximizing demand coverage,and the MCMCC model based on complementary coverage is used to locate the electronic fence area.Finally,Density-Based Spatial Clustering of Applications with Noise(HDBSCAN)algorithm is used to cluster the Origin-Destination(OD)points in each area,and the accurate location of the electronic fence is further determined according to the centroid of the cluster and the fence capacity.(2)A framework for analyzing the environmental benefits of bicycle-sharing based on the semantic analysis of Point of Interest(POI)types is proposed.The main purpose of this framework is to analyze and visualize the relationship between the changes of environmental benefits of bicycle-sharing and the functions of urban areas.First,map the traffic trajectory data to the urban grid,and calculate the energy consumption,carbon emissions and nitrogen oxide emissions of each grid.Then,make multi-scale geographic weighted regression analysis on the building environmental factors to obtain the influencing factors related to carbon emissions,and preliminarily understand the regional functions.Secondly,the POI data of cities are retrieved and the training data set is constructed.The high-dimensional feature vectors of POI types are extracted through the Word2 Vec model,and then mapped to the two-dimensional semantic space using the t-distributed Stochastic Neighbor Embedding(t-SNE)dimensionality reduction technology to obtain the POI configuration diagram of each grid.Finally,according to the POI configuration diagram and regression results,the functions of urban areas and their impact on the change of environmental benefits of bicycle-sharing are analyzed,and feasible suggestions for improving environmental benefits are put forward.Finally,the two frameworks are verified by using the data of bicycle-sharing trips in Shenzhen.The experimental results show that the proposed location model MCMCC is more stable in coverage calculation than the traditional Maximum Coverage Location Problem(MCLP)model by considering the partial coverage of the space area,and provides a feasible solution for the electronic fence planning.At the same time,the environmental benefit analysis framework reveals the semantic relationship implied by the POI type,which can effectively identify the function of the region,help urban managers analyze the relationship between the energy saving and emission reduction effect generated by bicycle-sharing travel and the regional function,guide people to travel by bike,and help achieve the goal of carbon neutrality. |
PDF Full Text Request