城市交叉口非机动车交通冲突及其严重程度模型

发布时间：2018-10-31 10:54

【摘要】：为研究非机动车在城市交叉口发生交通冲突的特性,根据呼和浩特市某典型交叉口的实际非机动车运行特点,对统计时段内单位时间某进口道直行非机动车、右转机动车和当量小汽车到达数与相应的机非冲突次数进行一元和多元回归分析。综合考虑并归一化冲突距离、冲突速度和冲突角度对冲突严重程度的影响,建立非机动车冲突严重程度模型,以反映非机动车冲突严重程度。结果显示,在9个一元回归模型中,当量小汽车到达数与冲突次数拟合的二次多项式对冲突的预测准确率最高,为85.3%;在4个多元回归模型中,直行非机动车和当量小汽车到达数与冲突次数的回归方程的预测准确率最高,为92.4%,对非机动车冲突次数的预测效果最好。所建立的非机动车冲突严重程度模型可用于衡量实际发生的交通冲突的严重程度。
[Abstract]:In order to study the characteristics of traffic conflicts between non-motor vehicles at urban intersections, according to the actual non-motor vehicle operation characteristics of a typical intersection in Hohhot, a non-motor vehicle is used directly at an entrance road per unit time in a statistical period. The arrival number of right-turn motor vehicle and equivalent car and the corresponding non-conflict times are analyzed by univariate and multivariate regression analysis. Considering and normalizing the impact of conflict distance, conflict speed and conflict angle on the conflict severity, a non-motor vehicle conflict severity model is established to reflect the non-motor vehicle conflict severity. The results show that the quadratic polynomial fitting the arrival number of the equivalent car and the number of conflicts has the highest prediction accuracy of the conflict among the nine regression models with a precision of 85.3; Among the four multivariate regression models, the regression equation of the number of arrivals and the number of conflicts of non-motor vehicles and equivalent cars has the highest prediction accuracy (92.4), which is the best in predicting the number of conflicts of non-motor vehicles. The non-motor vehicle conflict severity model can be used to measure the severity of actual traffic conflicts.
【作者单位】：北京交通大学城市交通复杂理论与技术教育部重点实验室;
【基金】：国家自然科学基金资助(71210001)
【分类号】：U491

【参考文献】