Research On Probability Estimate Of Urban Traffic Congestion Based On The Cumulative Logistic Model

Evaluation of traffic condition evaluates the level of urban traffic operation. Based on real data, assessing the traffic operation condition accurately is very helpful for understanding the traffic operation situation of a city. Under the background of today’s serious traffic congestion, the accurate evaluation and forecast of traffic condition, especially estimate of traffic congestion, can help people to get a comprehensive understanding of a city’s traffic situation. It will allow more useful solution to fix the problem of traffic congestion. The current research is focusing on the theory, discussing the modeling of evaluation or prediction model. This thesis aims to accurately evaluate the traffic operation condition base on real data, and gives a comprehensive analysis of the real traffic situation of Beijing. On this basis, this thesis does probability estimate of traffic congestion.This thesis mainly adopts questionnaire and research to acquire evaluation of the traffic condition and the latest actual traffic condition data. According to the real situation of traffic congestion in Beijing, this paper classifies the time period and section type.7:30-900is defined as morning peak period.16:30-19:00is defined as evening peak period. Based on the traditional grade of route, this thesis further divides routes into six types using cluster analysis:all-day congestion section; peak period congestion section; morning peak period congestion section; evening peak period congestion section; all-day better traffic condition route; all-day good traffic condition route. Finally, this paper is based on the cumulative logistic regression model, establishing section probability estimate model of traffic congestion to calculate the probability of traffic congestion. Given the analysis about the traffic congestion probability, including: summary of congestion probability of all kinds sections in each period; differences between each kind of sections and each period; differences about traffic congestion probability between each kind of road grade and etc.