Purdue University School of Civil Engineering, 550 Stadium Mall Drive, West Lafayette, IN 47907, USA.
J Safety Res. 2011 Aug;42(4):231-9. doi: 10.1016/j.jsr.2011.01.008. Epub 2011 Jul 21.
Highway crash occurrence is a leading cause of unnatural deaths, and highway agencies continually seek to identify engineering measures to reduce crashes and to assess the efficacy of such measures. Most past studies on the effectiveness of roadway improvements in terms of crash reduction considered all rural two-lane sections as a single category of roads. However, it may be hypothesized that the differences in the mobility and accessibility characteristics that are reflected in (and due to) the different design standards between different functional subclasses in the rural two-lane highway system can lead to differences in efficacies of safety improvements at these subclasses. This paper investigates the efficacy of roadway improvements, in terms of crash reduction, at the various subclasses of rural two-lane highways.
An empirical analysis of safety performance at each of the three subclasses of rural two-lane highways was carried out using the negative binomial modeling technique. For each subclass, crash prediction models were developed separately for the three levels of crash severity: property-damage only, injury, and fatal/injury. The crash factors that were considered include lane width, shoulder width, pavement surface friction, pavement condition, and horizontal and vertical alignments. After having developed the safety performance functions, the effectiveness (in terms of the extent of crash reduction, for different levels of crash severity) of highway safety enhancements at each highway subclass were determined using the theoretical concepts established in past literature. These enhancements include widening lanes, widening shoulders, enhancing pavement surface friction, and improving the vertical or horizontal alignment.
The study found that there is empirical evidence to justify the decomposition of the family of rural two-lane roads into its constituent subclasses for purposes of analyzing the effectiveness of safety enhancement projects and thus to avoid underestimation or overestimation of benefits of safety improvements at this class of highways.
高速公路碰撞事故是导致非自然死亡的主要原因，公路管理机构一直在寻求确定工程措施以减少碰撞事故，并评估这些措施的效果。过去大多数关于道路改善减少碰撞效果的研究都将所有农村双车道路段视为单一类别道路。然而，人们可能假设，在农村双车道公路系统中不同功能子类之间反映出的（并由于）不同设计标准的移动性和可及性特征的差异，可能导致这些子类安全改进效果的差异。本文研究了农村双车道公路各子类中道路改善减少碰撞的效果。
使用负二项式建模技术对农村双车道公路各子类的安全性能进行实证分析。对于每个子类，分别为三个碰撞严重程度级别（仅财产损失、受伤和致命/受伤）开发碰撞预测模型。考虑的碰撞因素包括车道宽度、路肩宽度、路面摩擦、路面状况以及水平和垂直线路。在开发安全性能函数后，根据过去文献中建立的理论概念，确定每个公路子类的公路安全增强措施的有效性（针对不同严重程度的碰撞减少程度）。这些增强措施包括拓宽车道、拓宽路肩、增强路面摩擦以及改善垂直或水平线路。
研究发现，有经验证据证明将农村双车道公路家族分解为其组成子类，以便分析安全增强项目的效果是合理的，从而避免低估或高估该类公路安全改进的收益。
