Although considerable progress has been made over the past several years in making highway travel safer, the frequency and severity of speed-related crashes on the nation’s highways continue to be of concern. Understanding the factors associated with these crashes enables engineers to identify and implement effective countermeasures to reduce the probability of crashes. A number of studies have been conducted to determine the variation of crash rates as they relate to hourly traffic volumes, geometric characteristics, average speed, and speed variance. However, these studies have not established mathematical relationships that can be used to estimate changes in the crash characteristics as a result of the combined changes in speed, flow, and geometric characteristics. The establishment of direct mathematical models that describe the influence of these factors on crash characteristics would significantly enhance the efforts of traffic engineers to determine suitable countermeasures to reduce the occurrence and severity of crashes. This project develops mathematical relationships that describe the combined influence that traffic and geometric characteristics have on crash occurrences. This study was limited to roadways in the state of Virginia with speed limits of 89 or 105 km/h (55 or 65 mph). The data were obtained from speed monitoring stations established by the Virginia Department of Transportation (VDOT) and from police accident reports from January 1993 to September 1995. Using the variables of mean speed, standard deviation of speed, flow per lane, lane width, and shoulder width to predict crash rates, different types of deterministic models, such as multiple linear regression, robust regression, and multivariate ratio of polynomials were fitted to the data. The multivariate ratio of polynomials was found to be the only mathematical model type that was successful in describing any relationship between the combined effects of changes in the speed, flow, and geometric characteristics of the road on crash rates. Based on this study, all of the models show that under most traffic conditions, the crash rate tends to increase as the standard deviation of speed increases. The effect of the flow per lane and mean speed on the crash rate varied with respect to the type of highway. (Author/publisher)
Samenvatting