Authors:
Zhixu Gao
1
;
Guyue Tian
2
and
Fengsi Yu
2
Affiliations:
1
School of physics and electronic information, Yunnan Normal University, Yunnan 650500, China, China
;
2
College of Science, Chongqing University of Technology, Chongqing, 401320, China, China
Keyword(s):
Survival analysis model; Traffic congestion; Predictive model.
Abstract:
Urban transportation is the core of urban social activities and economic activities. However, due to the increase of population and motor vehicles, traffic congestion is caused by many factors. This paper established a traffic congestion duration model based on survival analysis. The purpose is to use the model to obtain the relationship between congestion index and congestion time, and improve the accuracy of prediction. Using the nonparametric method to calculate, after defining the Shanghai Expressway survival function and risk function, combined with the compiled data, calculate whether it is the impact of the working day on traffic congestion, and the difference between the early, middle and late peaks for traffic congestion. The result can be obtained: Traffic congestion on workdays is higher than on weekends, and traffic congestion is longer than weekends.