Modeling And Analyzing Of Pedestrians’ Waiting Time At Signalized Intersection

Urban Road Intersection is an important part of the urban road system, and walking is the most basic and the most extensive travel modal. The priority is to improve vehicles’ efficiency for planning and design of signal intersection, the time and places of pedestrians crossing are random because of pedestrians’ subjective consciousness and the structure characteristics of crossing facilities, which will often cause pedestrian crossing violation.Firstly, the characteristics of pedestrian violation at intersection were analyzed. The pedestrian violation behavior was defined. Moreover, based on the field data of Dongdan North St. & Jinbao St. and Capital Gymnasium South St. & Zengguang St. signalized intersection, the difference of pedestrian crossing violation rate was analyzed in circumstances of different time(peak and off-peak period), different gender and different signal lights(countdown and no countdown) settings. And statistical analysis methods were used to descriptively analyze pedestrian waiting time on these conditions. The analysis shows that the pedestrian waiting time at signal intersection can’t be described by normal distribution, which laid the foundation for modeling pedestrian’s waiting time at signalized crosswalk.Applying Generalized Pareto Distribution(GPD), this paper aims to develop a statistical model of pedestrians’ waiting time at signalized intersections. Heavy-tailed distribution diagnosis was adopted to investigate tail characteristics of pedestrians’ waiting time for street crossings at the signalized intersection. The GPD has three parameters, including the location parameter(or threshold) ?, the scale parameter? and the shape parameter ?. Based on field data we employ Partial L-moments technique to estimate these parameters. A2 and W2 methods were used to test the model. Furthermore, we obtained the GPD fitting figures of excess of pedestrians’ waiting time in difference cases. Results show that the statistical model of pedestrians’ waiting time at signalized intersections is a model with light tail because its shape parameter ? is less than 0. And the value range of pedestrians’ waiting time t is a finite interval, which overcomes the shortcomings of exponential distribution and has good practicability. The GPD is able to mimic pedestrians’ crossing behavior in terms of waiting time.The parameter, maximum pedestrians’ waiting time for street crossings is the main reference for determining whether to take pedestrian signal control and determine the maximum pedestrian red light phase, which has certain guiding significance to the optimization of pedestrian traffic. Based on the GPD, different quantiles of pedestrians’ waiting time at signalized intersections were solved. We find that thresholds of pedestrians’ waiting time at Dongdan North St. & Jinbao St. and Capital Gymnasium South St. & Zengguang St. signalized intersection are 56 s and 42 s respectively. Furthermore, quantile charts of pedestrians’ waiting time were obtained in circumstances of different time, different gender and different signal lights settings, and comparative analysis was carried out. The optimal design program of the intersection for pedestrian signal(red time, continuous green time, green flash time) was proposed, and finally some suggestions on improving pedestrian behavior were put forward.