| Because of the convenience and economy of bicycle sharing,it has brought new opportunities for citizens to travel.This new environment-friendly transportation method not only helps travelers solve short-distance travel to a certain extent,but also becomes an effective way to connect rail transit stations in many cities.This research aims to provide a reference for the management and planning of shared bikes from the current situation by analyzing the spatial and temporal distribution characteristics of the bike-sharing used to connect with metro stations in Xi’an and its influencing factors,so as to make shared bikes more effectively solve the problem of "first mile / last mile".In this study,according to domestic and foreign research on the spatiotemporal distribution of shared bike use and its influencing factors,based on the GPS data of shared bicycle within the working day of Xi’an in 2018,the data processing is combined with different visualization methods to identify the hotspots of shared bikes use for rail transit.The geographically weighted regression model was also used to explore the relationship between the type of land use and the use of shared bikes.Specifically,after a series procedures of data preprocessing,such as coordinate system calibration,defining data boundaries,deleting duplicate values,and removing outliers,the information about the starting and ending points of riding is extracted in the Python environment.Based on the O-D extraction,a method for generating riding paths is provided.Besides,combined with the Arc GIS,the shared bike trips that connect rail transit stations are identified,and a method for integrating the O-D data is proposed.After completing the data preparation,various statistical charts based on different visualization software or platforms are used here.Different visualization methods such as heat maps,interactive flow maps,routes maps are also used to intuitively show the spatiotemporal distribution of the shared bikes use for metro.The daily volume changes,the distribution of the starting and ending points,the distribution of the riding directions,and the distribution of the riding paths are visually displayed.Through the identification of the temporal distribution,the peak hours of bike use are determined,thus this research also shows the spatial distribution of cycling for morning and evening peaks.Secondly,this study starts with eight influencing factors that reflect the distribution of POI related to the type of land use.Through global regression and combining the actual land use type distribution in the study area,the selected variables are the number of residences,companies,shopping and recreation.Put the final four independent variables and the use of shared bicycles connected to subway stations into the GWR model to reveal the impact of spatial variables on the use of shared bicycles in different geographical locations in Xi’an.The study concludes that the shared bikes connecting the metro in the morning and evening peaks are mainly used for commuting;the use of shared bikes is distributed around the rail transit network as the axis,radiating outward,and the closer to the metro line,the greater the demand for bikes;Areas with diverse land types have large demand for bikes and are sensitive to changes in the number of residences,companies,shopping,and recreation venues;a simple land use type area has a high cycling travel potential,and there are often cycling hot spots in the area.The number of rides is greatly affected by the main land use types. |
