Lu G.,Beihang University |
Cheng B.,Tsinghua University |
Kuzumaki S.,Toyota Motor Corporation |
Mei B.,Beijing Traffic Management Bureau
Traffic Injury Prevention | Year: 2011
Objective: Road traffic conflicts can be used to estimate the probability of accident occurrence, assess road safety, or evaluate road safety programs if the relationship between road traffic accidents and conflicts is known. To this end, we propose a model for the relationship between road traffic accidents and conflicts recorded by drive recorders (DRs). Methods: DRs were installed in 50 cars in Beijing to collect records of traffic conflicts. Data containing 1366 conflicts were collected in 193 days. The hourly distributions of conflicts and accidents were used to model the relationship between accidents and conflicts. To eliminate time series and base number effects, we defined and used 2 parameters: average annual number of accidents per 10,000 vehicles per hour and average number of conflicts per 10,000 vehicles per hour. A model was developed to describe the relationship between the two parameters. Results: If A i = average annual number of accidents per 10,000 vehicles per hour at hour i, and E i = average number of conflicts per 10,000 vehicles per hour at hour i, the relationship can be expressed as (α > 0, β > 0). The average number of traffic accidents increases as the number of conflicts rises, but the rate of increase decelerates as the number of conflicts increases further. Conclusions: The proposed model can describe the relationship between road traffic accidents and conflicts in a simple manner. According to our analysis, the model fits the present data. © 2011 Taylor & Francis Group, LLC.
Wang S.,Tsinghua University |
Li R.,Tsinghua University |
Guo M.,Beijing Traffic Management Bureau
Transport | Year: 2015
Predicting the duration time of incidents is important for effective real-time Traffic Incident Management (TIM). In the current study, the k-Nearest Neighbor (kNN) algorithm is employed as a nonparametric regression approach to develop a traffic incident duration prediction model. Incident data from 2008 on the third ring expressway mainline in Beijing are collected from the local Incident Reporting and Dispatching System. The incident sites are randomly distributed along the mainline, which is 48.3 km long and has six two-way lanes with a single-lane daily volume of more than 10000 veh. The main incident type used is sideswipe and the average incident duration time is 32.69 min. The most recent one-fourth of the incident records are selected as testing set. Vivatrat method is employed to filter anomalous data for the training set. Incident duration time is set as the dependent variable in Kruskal–Wallis test, and six attributes are identified as the main factors that affect the length of duration time, which are ‘day first shift’, ‘weekday’, ‘incident type’, ‘congestion’, ‘incident grade’ and ‘distance’. Based on the characteristics of duration time distribution, log transformation of original data is tested and proven to improve model performance. Different distance metrics and prediction algorithms are carefully investigated. Results demonstrate that the kNN model has better prediction accuracy using weighted distance metric based on decision tree and weighted prediction algorithm. The developed prediction model is further compared with other models based on the same dataset. Results show that the developed model can obtain reasonable prediction results, except for samples with extremely short or long duration. Such a prediction model can help TIM teams estimate the incident duration and implement real-time incident management strategies. Copyright © 2015 Vilnius Gediminas Technical University (VGTU) Press
Guo M.,Beijing Traffic Management Bureau |
Lan J.-H.,University of Science and Technology Beijing |
Li J.-J.,University of Science and Technology Beijing |
Lin Z.-S.,University of Science and Technology Beijing |
Sun X.-R.,University of Science and Technology Beijing
Jiaotong Yunshu Xitong Gongcheng Yu Xinxi/Journal of Transportation Systems Engineering and Information Technology | Year: 2012
The quality of the raw traffic data detected from traffic sensors directly affect the follow-up benefits of the intelligent transportation systems. In view of the widespread failure problems of collected traffic data, the paper takes the traffic flow data of intersection detector as the research object. A traffic flow data recovery algorithm based gray residual GM (1, N) model is proposed. First, the grey relational analysis is conducted on the traffic flow of four links at an intersection. Then a grey model GM (1, N) is developed for the estimated recovery of failure data. The residual modification is used to improve the accuracy of the repaired data. The results indicate that the proposed traffic flow data recovery algorithm is feasible. It is able to solve the post-processing difficulties due to data failure and it serves as a good method for failure data recovery in other areas as well.
Li R.,Tsinghua University |
Guo M.,Beijing Traffic Management Bureau
Journal of Advanced Transportation | Year: 2015
Summary Different clearance methods in traffic accident management lead to varied duration distributions. Apart from investigating the influence of various factors associated with accidents on the duration of such accidents using different clearance methods, this study also examines the cumulative incidence probability. We used traffic accident data obtained for 12 months from the Fourth Ring Expressway main line in Beijing to develop a subdistribution hazard regression model, which can assess the risk factors of two clearance methods. The regression results show that the different factors have statistically significant effects on the duration of two accident groups with different clearance methods; furthermore, opposite effects occur even for some factors that have a strong effect on both accident groups. For example, an accident involving a taxi extends the duration time with clearance method 1; in comparison, the accident is shorter with clearance method 2. The predicted cumulative incidence curves of the two types of clearance methods are shown as examples, with stratification based on the influence factors (taxi involved, season). Finally, the Gray test of the cumulative incidence functions and the log-rank test of the Kaplan-Meier estimates of the survival functions are compared, in order to demonstrate the importance of using proper methods for analyses. Copyright © 2014 John Wiley & Sons, Ltd.
Gu Y.L.,Beijing Jiaotong University |
Ma Y.N.,Beijing Jiaotong University |
Li J.W.,Beijing Traffic Management Bureau
Advanced Materials Research | Year: 2014
Dynamic traffic data quality evaluation can provide reliable data support for the traffic management system. This paper aim at three kinds of common failure data, put forward the traffic flow data quality control methods and process. First time evaluate the data quality in three stages, and each stage respectively for all kinds of ITS detector, and builds the six index system. Finally take Beijing as an example to analyze and verify the method of this paper. © (2014) Trans Tech Publications, Switzerland.