The Time-Space Analysis Of Shared Bikes And Site Selection Of Electronic Fence

Shared bike is an emerging bike-sharing service.Recent years,the traffic problems in urban have become increasingly serious,such as air pollution,traffic congestion and so on.In order to solve the problems above,the shared bike emerged.Shared bike has features such as scan the two-dimension code intelligently,no fixed parking piles,high flexibility and density.The point-to-point way to travel of shared bike,not only effectively solves the problem of "the last kilometer" of urban traffic system,but also alleviates the phenomenon of traffic congestion to a certain extent.However,as more and more companies have joined in the interest pursuit of shared bike,the number and types of delivery gradually increase,also,various problems appear in the operation and management process.The phenomenon of random parking caused by the tension or ambiguity of parking areas has a relatively serious impact,which is easy to cause traffic accidents,affect the city appearance and the operation efficiency of urban rail transit.The essence of the problem could boil down to how to ensure the shared bike could provide the most services while occupying the least public resource space.The scientific location of electronic fence is one of the important methods to solve this problem,which guide the user to park scientifically by delimiting the virtual parking frame.It can not only reduce the burden for urban management departments,but also expand the influence scope and attraction of urban rail transit,to attract more travelers to adopt this approach.In order to solve the location problem of electronic fence,this paper designs a location model for the maximum coverage problem of electronic fence based on simulated annealing algorithm,and takes Beijing city as an example to apply the algorithm.The main work is as follows:(i)Space-time analysis of shared bike.Based on the trip data provided by Beijing Mobike company after Geohash decoding,using Excel,Python and other data analysis methods.In terms of time dimension,the changing trend of shared bike users in various time periods was analyzed,in order to obtain the characteristics of their daily distribution.In terms of spatial dimension,Folium library of Python was called,to display the dynamic change of spatial distribution of shared bike in Beijing at each moment in way of heat map.(ii)Location selection of electronic fence.Under the theoretical guidance of the maximum coverage problem,the algorithm was designed,and the simulated annealing algorithm in the heuristic algorithm was applied to solve the problem.Taking xicheng district,Beijing as an example,we obtained the location selection results of electronic fence under different number of sites,set 75% as the minimum coverage.The coverage,calculation cost and mean distance were compared,the spatial visualization of the calculated results was carried out with the combination of GIS.The conclusions of this paper are as follows:(i)The shared bike in Beijing has shown the tidal distribution on weekdays,but not on weekdays;The bikes are distributed roughly randomly along the street;During the morning rush hour,the hot spots of Shared bikes gradually develop from the outer ring to the inner ring over time;From the morning and evening peak OD heat map,it can be seen that the heat difference between the vehicle volume and the vehicle return volume during the peak period in the region indicates that the number of vehicles is not equal.(ii)Taking xicheng district as an example,the site selection algorithm constructed in this paper is applied.In the case of different electronic fence coverage radius(200m,300 m,400m)and the combination of the number of sites,the calculation process of this model can converge quickly and reach the maximum coverage,which indicates that this model is efficient and has high availability.Through the combination of different conditions of coverage,repetitive coverage rate and the average distance between all the indexes analysis shows that this model could get the reasonable scheme in various cases,which means this method has good universality,can be applied in different sizes and demand planning problem of site selection of electronic fence to provide decision support,and determine the most appropriate location problem.