ABSTRACT

Poor coordination between geometric elements of horizontal curves can lead to unsafe speed. This study evaluated the effect of horizontal curve geometry on vehicle speed distribution for a four-lane divided highway and developed prediction models for mean speed and standard deviation. Spot speed data of passenger cars (PC), heavy commercial vehicles and light commercial vehicles were collected at 24 horizontal curves. Statistical analysis of the speed data of randomly selected 10 sites revealed PC as the critical vehicle category. Hence, the effects of horizontal curve geometry on PC speed were analyzed and radius was identified as the most influencing parameter. Radius beyond 400 m had similar influence on speed distributions. Appropriate models were developed using regression analysis. Degree of curvature was found as the most significant predictor. Benefits of the proposed models over traditional 85th percentile speed in designing and evaluating horizontal curves for consistency and safety were discussed.

摘要

水平曲线的几何元素之间的协调不良会导致速度不安全。这项研究评估了四车道分割公路的水平曲线几何形状对车速分布的影响，并开发了平均速度和标准偏差的预测模型。沿24条水平曲线收集乘用车（PC），重型商用车和轻型商用车的现货速度数据。对随机选择的10个站点的速度数据的统计分析表明，PC是关键车辆类别。因此，分析了水平曲线几何形状对PC速度的影响，并确定了半径是影响最大的参数。半径超过400 m对速度分布也有类似的影响。使用回归分析开发了合适​​的模型。发现弯曲度是最重要的预测指标。在设计和评估水平曲线的一致性和安全性时，讨论了所提出的模型优于传统的第85个百分点的速度的好处。