摘要: |
Recent research suggests that traditional safety evaluation methods may be inadequate in accurately determining the effectiveness of roadway safety measures. In recent years, advanced statistical methods are being utilized in traffic safety studies to more accurately determine the effectiveness of roadway safety measures. These methods, particularly Bayesian statistical techniques, have the capabilities to account for the shortcomings of traditional methods. Hierarchical Bayesian modeling is a powerful tool that more fully identifies a given problem than a simpler model could. This report explains the process wherein a hierarchical Bayesian model is developed as a tool to analyze the effectiveness of two types of road safety measures: raised medians and cable barrier. Several sites where these safety measures have been implemented in the last 10 years were evaluated using available crash data. The results of this study show that the installation of a raised median is an effective technique to reduce the overall crash frequency and crash severity on Utah roadways. The analysis of cable barrier systems shows that they are effective in decreasing cross-median crashes and crash severity. The tool developed through the research can now be utilized for additional analyses, including hot-spot analysis, before-after change, and general safety modeling. This tool will be an asset to the Utah Department of Transportation Traffic and Safety Division for data analysis in the years to come. |