A GWRbased study on spatialtemporal differentiation oftaxi pickup activities(PDF)
长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]
- Issue:
- 2020年5期
- Page:
- 77-86
- Research Field:
- 交通工程
- Publishing date:
Info
- Title:
- A GWRbased study on spatialtemporal differentiation oftaxi pickup activities
- Author(s):
- JIAO Ping; WANG Lixuan; YUAN Changwei; GENG Xinrui
- (1. School of Economics and Management, Xian Aeronautical University, Xian 710077, Shaanxi, China;2. School of Economics and Management， Changan University， Xian 710064， Shaanxi， China)
- Keywords:
- traffic engineering; spatialtemporal characteristic; GWR model; kernel density estimation; taxi trajectory; GPS
- PACS:
- -
- DOI:
- -
- Abstract:
- In order to mine the taxi operation law that reflected by the taxi trajectory data as much as possible, the POI data of the city was taken as the influence parameter into the data research of the taxi trajectories when the phenomenon of timespace difference of the taxi operation was discussed. The times of the pickup in each time interval were extracted and counted to confirm that the taxicab pickup events have time nonstationary, and the pickup points in each peak period were clustered by using the kernel density estimation, to confirm that the taxicab pickup events have the spatial nonstationarity. The hot spot extraction model was established to extract hot spots and defined the heat value, facilities that significantly affect the distribution and heat value of hot spots were picked from the urban POI data. In order to study the temporal and spatial characteristics of hot spots and various facilities of urban, the geographical weighted regression (GWR) model was established in different time periods by calculating the shortest distance between facilities and the hot spots. Additionally, a case study was conducted based on GPS data of taxies in Xian. The results show that comparing to the ordinary least squares（OLS）model，GWR provided a much better fit(with the average value of goodness of fit increasing from 0.29 to 0.57). At different time periods in the study area, the mean values of regression coefficients of various facilities have obvious temporal differences, in addition, the distribution of regression coefficients of similar facilities has obvious spatial differences. The spatialtemporal differentiation of hot spots in the study area generally conforms to the travel regularities of urban residents, and taxis play a more important role as a substitute for public transport system at the peak hours of working days. 〖JP2〗The research results can provide implications to the prediction of taxi spatial demand and fine management of taxi dispatch and price control. 3 tabs, 9 figs, 21 refs.
Last Update: 2020-10-12