科研之友
微信
新浪微博
Facebook
分享链接
摘要
Snow removal and ice control are critical to the safe and efficient use of roadways in much of the world. Tools that help road authorities plan for, assess, and communicate about winter maintenance activities are increasingly being sought. Weather indices that rate the severity of winter conditions in a given period at a particular location are one subset of such tools. Although several winter severity indices have been developed, many of these are complex and of limited practical use for road authorities. In this research, an approach for developing a simple-to-use, locally calibrated winter severity index (WSI) is outlined. The approach relies primarily on mathematical optimization but is flexible enough to allow expert knowledge to be incorporated. This approach is used in developing a WSI for the provincial highway system of the province of Alberta, Canada. The WSI assigns each day a weather score based on seven weather attributes and one seasonal adjustment factor. By using a classification and regression approach, the eight components are based on attribute-specific thresholds that then trigger a score. These daily scores are aggregated to the 14-day period and are correlated to maintenance activities. The WSI for Alberta highways has strong fit with maintenance activity that occurred, when measured as vehicle hours. At the 14-day level, the R2 values for vehicle hours vary from 0.776 to 0.936 for eight contract areas, and when the data are aggregated over space or time, the fit improves further. This research illustrates how site-specific WSIs can be developed for road authorities through the use of optimization that alters the weather attribute thresholds and scores, reflecting the specific maintenance regimes in that location.
- 出版日期2017-9
