Scholarlink

This study presents a comprehensive methodological framework to identify and explore the factors associated with road traffic injury frequency at urban midblock sections in Hyderabad, India. To explore the relationship between crash frequency at different levels of severity (fatal, non-fatal, and total) and roadway and traffic and geometric attributes, three sets of exclusive Safety Performance Functions (SPF) are formulated. For SPF model calibration, (a) historical crash data (2015–2019) collected from Hyderabad police and (b) primary geometric, traffic, and roadway-built environment-specific information collected through detailed road inventory survey across Hyderabad were utilized. Subsequently, Negative Binomial Regression models, an extensively adopted technique for count-data-based modelling are used to develop three sets of SPFs (i.e. Total Crash-SPF, ii. Fatal-Crash-SPF and iii. Non-Fatal-Crash-SPF) for midblock sections in Hyderabad. Results clearly indicate that speed, median width, and midblock segment length are found to significantly influence the fatal/non-fatal/total crash frequency at midblock roadway segments. The identification of potential influencing factors that are relevant to the frequency of crashes could help in developing necessary mitigation measures.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Investigation of Risk Factors for Midblock Road Traffic Crashes Using Negative Binomial Model: A Case Study of Hyderabad, India

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Siddardha Koramati,
  • Bandhan Bandhu Majumdar,
  • Prasanta K. Sahu

摘要

This study presents a comprehensive methodological framework to identify and explore the factors associated with road traffic injury frequency at urban midblock sections in Hyderabad, India. To explore the relationship between crash frequency at different levels of severity (fatal, non-fatal, and total) and roadway and traffic and geometric attributes, three sets of exclusive Safety Performance Functions (SPF) are formulated. For SPF model calibration, (a) historical crash data (2015–2019) collected from Hyderabad police and (b) primary geometric, traffic, and roadway-built environment-specific information collected through detailed road inventory survey across Hyderabad were utilized. Subsequently, Negative Binomial Regression models, an extensively adopted technique for count-data-based modelling are used to develop three sets of SPFs (i.e. Total Crash-SPF, ii. Fatal-Crash-SPF and iii. Non-Fatal-Crash-SPF) for midblock sections in Hyderabad. Results clearly indicate that speed, median width, and midblock segment length are found to significantly influence the fatal/non-fatal/total crash frequency at midblock roadway segments. The identification of potential influencing factors that are relevant to the frequency of crashes could help in developing necessary mitigation measures.