01593451

Component

Researchers have attempted to rank sites using crash count, severity, expected severity and expected crash count or jointly using two metrics. The top ranked sites are designated as hotspots and are subjected to detailed investigation for design of improvement strategies. The choice of number of candidates for detailed investigation is subjective and depends on the practitioners’ discretion and available fund. No attempt has yet been made to take all the measures together in aggregate form to assess safety scenario of a site. The authors propose a joint fuzzy clustering method to assess safety of sites based on crash count, severity, expected severity and expected crash count. The clusters help identifying hazardous sites or hotspots which represent all possible candidate locations for further investigation.

This paper was sponsored by TRB committee ABJ80 Standing Committee on Statistical Methods. Alternate title: Development of a Comprehensive Fuzzy Cluster-Based HSID Technique

01584066

16-6957

English

17p

2016

Transportation Research Board 95th Annual Meeting

Digital/other

References; Tables

Crash severity; Evaluation and assessment; Fuzzy algorithms; High risk locations; Highway safety; Identification systems; Ranking (Statistics)

Crash counts

Highways; Planning and Forecasting; Safety and Human Factors

Transportation Research Board Annual Meeting 2016 Paper #16-6957

TRIS, TRB, ATRI

Jan 12 2016 7:02PM