Research On Modeling For Motorized Vehicles Running Red Light Considering Random Effects

Running a red light could cause side collision at signal intersection with huge loss of life and property. The number of red light running vehicle has decreased significantly with the installation of red light camera, the countdown timer, and the policy of demerit points and financial penalties. But the phenomenon of this illegal traffic accident is still very prominent, despite repeated prohibition. There are many factors that affect the red light running and the mechanism is still needed to study deeply. Besides, the preventive measures and countermeasures are also need to be further discussed. In this paper, based on open project of key laboratory of ministry of public security for road traffic safety, on the basis of observation data, the paper analyzed the characteristics of the behavior of the vehicle’s running red light. This paper built models of the behavior of red light running, determined the key influence factors of the behavior of the motorized vehicle running red light, and discussed the corresponding prevention methods.Parameters of running red light vehicles were obtained through the observation of the video. According to the above analyses, the traffic characteristics of a red light running vehicle was analyzed in this paper, the statistical distribution of traffic composition, the distance between the vehicle and the stop line at the yellow onset, the speed of the vehicle at the yellow onset and the red onset and the headway were studied.The concept of critical distance was introduced, combining with the start and end of red light, to classify motorized vehicles red light running into three categories: deliberate and unintentional red light violation at red onset and red light violation at red end. Based on video observation data, the red light running rate and the time of entering intersection that commit any of the afore-mentioned three types of red light violation were statistically analyzed, and then their characteristics under various traffic control and management conditions were explored using z-test, f-test and t-test method.According to the characteristics of the three types of red light, combined with the main form of the counting data model and the applicable conditions, considering the influence of random effects, the random-effects negative binomial regression model was selected as the vehicle running red light counting data model. The intersection geometric parameters, signal timing scheme, traffic management control mode and traffic environment characteristic were chosen as independent variables in the model. The key influence factors of the afore-mentioned three types of red light violation were determined with the random-effects negative binomial regression model.According to the three random-effects negative binomial regression models, the key factors that affect the three types of red light running and specific factors that belongs to each type were found out. The results show that countdown timer, green flash, the width of lane and traffic volume can affect the three types. Methods combing the signalized intersection design, traffic signal control device, signal timing scheme and traffic policy were proposed to avoid motorized vehicle running the red light.