基于Logistic回归模型的闯红灯行人的人口特征研究

研究行人人口特征对闯红灯行为的影响.选取某个道路、交通和环境等条件单一的信号交叉口，从而控制这些客观因素对闯红灯行为的影响，采集1 095个过街行人数据，包括121名闯红灯行人、195名红灯时期到达路口等待绿灯通行的行人和779名绿灯期间到达并通行的行人.采用logistic回归模型，分析了行人闯红灯的选择与聚集性、性别、年龄的关系.研究结果表明，群体较之个人，男性较之女性，老年人较之青年人和中年人，闯红灯的可能性更大，从而为制定针对性的交通安全宣传策略提供指导.

Research on Demographic Characteristics of Pedestrians Running the Red Light Based on Logistic Regression

The effect of the pedestrians'' demographic characteristics on their likelihood of running the red light was presented. One signalized intersection with regular road, traffic and environment conditions was singled out to minimize these factors'' effects on the research. Then, the data of 1095 pedestrians crossing the intersection was collected, including 121 pedestrians running the red light, 195 arrived in the red light period waiting till the green light on, and 779 arrived in the green light period. Based on the data, a logistic regression model was built to analyze the relationship between likelihood of running the red light and the pedestrians'' scale, sex and age. The result shows that, 1) compared with the female, the male is more likely to run the red light; 2) compared with the youth and the middle-aged, the elderly people are more likely to cross the red light; 3) compared with a pedestrian alone or the smaller scale, the larger scale are more likely to run red lights. As a whole, the study can provide some guidance when traffic management department implementing traffic safety education.