Research On Characteristics Of Shared Bicycles Riding And Its Layout Optimization

With the development of the Internet and sharing economy,free-floating bike sharing(FFBS)systems have been widely adopted in major cities over the world.The booming bike-sharing business provides great convenience to people's daily trips and brings notable change to city traffic.However,the popularity of shared bicycles have also brought about some urban problems,such as the fact that some urban corners have a lot of bicycles,and some bicycles are hard to find in some area.Therefore,using the shared bicycle data to fully understand the travel behavior of residents is of great significance for the development and layout of shared bicycles.This paper takes Beijing Fifth Ring as the research area,focusing on the analysis of characteristics of dock-less shared bicycles riding and its layout optimization.The main research contents and conclusions of this paper are as follows:(1)This paper firstly gives a detailed analysis of the spatial and temporal characteristics of shared bicycle riding.On the temporal scale,there are two rush hours every day especially on weekdays.The morning rush hours usually start from 7 am to 9 am,and the evening rush hours range from 5 pm to 7 pm.On the spatial scale,the travel distance of shared bicycles are calculated,and dock-less shared bicycles was mainly used for short-distance travel in cities,especially for commuting and schooling.The hot bike riding streets and areas are identified based on the kernel density estimation.In order to explore the regional difference of the bicycle spatial distribution,this paper analyzes the relationship between shared bicycles and urban elements,such as urban population,city roads,subways and bus stops.Additionally,the paper quantifies the connection effect of shared bicycles with subways and buses,and verifies that the emergence of shared bicycles has greatly enhanced the service range of subways and buses.(2)This paper constructs three-mode tensors of shared bikes' origins and destinations from three dimensions,namely date,hour and space.The Tucker tensor decomposition model under non-negative constraints was used to mine the characteristics of shared bicycle riding modes in different dimensions.Three factor matrices in tensor decomposition results reveal two day modes,three hour modes,and six space modes of shared bicycles in different dimensions:the day dimension has two typical bicycle riding modes,including the weekday mode and the weekend mode;the hour dimension contains three bicycle riding modes,and these modes are the mode of daytime,the mode of morning rush hours and the mode of evening rush hours;six space modes of shared bicycle origin and destination points in the space dimension are analyzed and further explained through the point of interest(POI)density of the study area.Additionally,Single Value Decomposition(SVD)analysis was applied on OD matrices of shared bicycles.After SVD matrix decomposition,top 50 single values with corresponding orthonormal vectors were chosen to give an analysis about spatialization and visualization of bike-sharing OD flows.(3)This paper combines the spatial and temporal distribution and pattern characteristics of shared bicycle riding,and summarizes the layout principle of electronic fence to provide guidance.In this paper,the location allocation location model was used to maximize the coverage of the electronic fence selection objective function,and the electronic fence optimization location model algorithm was constructed to determine the appropriate number of electronic fences in the study area.As a result,a proposal for the layout of the electronic fence was given to improve the service efficiency of shared bicycles.