Bayesian-Based Approach for Updating Characteristics of Crash Data: Frequency, Duration, and Lanes Blocked

Crashes cause non-recurring congestions, resulting in vehicle delays and additional emissions on highways. In order to properly characterize crashes, various studies have been carried out on the modeling of crash frequency, duration and lanes blocked based on the historically observed data. However, there has been very limited research on how to update these characteristics in a timely manner in order to capture the profile of crashes more accurately. The characteristics of the statewide crash data file have normally been updated 2-3 years after the occurrences of crashes. This practice has negatively affected the accuracy in modeling the delays and emissions caused by crashes. In this context, the objective of this paper is to develop a Bayesian-based approach for updating the characteristics of crash data in an almost instant manner whenever the new crash information becomes available. By the Bayesian approach, the distribution of the crash frequency, duration and affected lanes can be readily updated based on the new crash data. In the case study, data for crash frequency in Harris County, Texas from 1992 through mid 2000 and data for crash duration and lanes affected in Houston, Texas from 2000 to mid 2002 are used to construct the historical distributions, while data from the late 6 or 7 months are used to update these distributions. Poisson, Lognormal and Binominal distributions are modeled for crash frequency (crash varies with time of day, and day of week), duration, and lanes affected, respectively. The updating procedure is realized through renewing the conjugate distributions of these modeled distributions, and thus the key parameters inside the original modeled distributions. The entire process, though incorporating complex probability analysis, is implemented within Microsoft EXCEL platform in a practically simple approach, which can be easily used by practical engineers and researchers.


  • English

Media Info

  • Media Type: CD-ROM
  • Features: Figures; References; Tables;
  • Pagination: 23p
  • Monograph Title: TRB 86th Annual Meeting Compendium of Papers CD-ROM

Subject/Index Terms

Filing Info

  • Accession Number: 01043542
  • Record Type: Publication
  • Report/Paper Numbers: 07-0424
  • Files: TRIS, TRB
  • Created Date: Feb 8 2007 5:00PM