Classification of urban rail transit stations based onpassenger flow time series(PDF)
长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]
- Issue:
- 2021年6期
- Page:
- 113-126
- Research Field:
- 交通工程
- Publishing date:
Info
- Title:
- Classification of urban rail transit stations based onpassenger flow time series
- Author(s):
- MA Zhuanglin1; YANG Xing1; TAN Xiaowei2; MA Fei3; WANG Jin4
- (1. College of Transportation Engineering, Chang’an University, Xi’an 710064, Shaanxi, China;2. School of Automobile, Chang’an University, Xi’an 710064, Shaanxi, China;3. School of Economics and Management, Chang’an University, Xi’an 710064, Shaanxi, China;4. Yunnan Science Research Institute of Communication Co., Ltd., Kunming 650011, Yunnan, China)
- Keywords:
- traffic engineering; urban rail transit; classification of station; smart card data; passenger flow characteristic; principal component analysis; Gaussian mixed model
- PACS:
- -
- DOI:
- -
- Abstract:
- Amid at different types of stations have significant differences in the aspects of regional characteristics, traffic functions and land use functions, and scientific and reasonable station classification was used to helpful to understand urban functional zoning, interpret residents’ travel characteristics, comprehend urban pattern and evolution, and evaluate the construction of rail transit infrastructure. The principal component analysis (PCA) was used to extract the characteristics of passenger flow data, and the Hopkins statistics was used to analyze the clustering trend and determine the number of clustering categories. The advantages and disadvantages of Gaussian mixture model (GMM) and Kmeans clustering were compared by using CH coefficient, Silhouette coefficient and DB index. GMM was used to identify the types of rail transit stations based on the smart card data of two consecutive working days of Nanjing rail transit in 2017. The results show that the effectiveness of three indicators of GMM method are better than that of Kmeans clustering method under the same input variables and iterations. Nanjing rail transit stations are divided into six types, namely residentialoriented station, employmentoriented station, spatial mismatched station, mixed mainly residentialoriented station, mixed mainly employmentoriented station, and hub comprehensive station by using the GMM. Nanjing rail transit stations have obvious characteristics of ring structure from the perspective of spatial distribution. The urban center area with densely populated is dominated by employmentoriented and mixed mainly residentialoriented station, the main urban area with moderate population density is dominated by mixed stations (such as spatial mismatched station, mixed mainly residentialoriented station, and mixed mainly employmentoriented station), and the outer suburbs with low population density have relative single function. There are 14 rail transit stations with a population density of more than 20 000 people per km2, which are mainly employmentoriented and mixed mainly employmentoriented stations. There are 16 rail transit stations with a population density of 10 000~20 000 people per km2, which are mainly employmentoriented, spatial mismatched, and mixed mainly employmentoriented stations. There are 12 rail transit stations with a population density of 5 000~10 000 people per km2, which are mainly spatial mismatched, mixed mainly residentialoriented, and mixed mainly employmentoriented stations. It can identify the peak period of passenger flow in urban rail transit stations on weekdays, and provide guidance for the department of operation and management of urban rail transit to formulate effective inbound and outbound passenger flow organization of the station and avoid the mass incidents like stampedes caused by impact of large passenger flow. 5 tabs, 14 figs, 37 refs.
Last Update: 2021-12-14