Recent research in advanced traffic management systems has emphasized incident detection and response to mitigate nonrecurring congestion. Existing incident response decision-making algorithms do not account for the expected losses associated with false alarms, undetected incidents, and delayed incident response. A freeway incident response decision-making system based on sequential hypothesis testing techniques is presented. The primary feature of this decision-making system is that it minimizes the sum of the expected losses associated with false response, nonresponse, and delayed responses to incidents through a dynamic programming algorithm. The results of simulation tests indicate that this algorithm performs better than typical Bayesian incident response algorithms for mean response time, false response rate, and nonresponse rate.
|Original language||English (US)|
|Number of pages||8|
|Journal||Transportation Research Record|
|State||Published - 1997|
ASJC Scopus subject areas
- Civil and Structural Engineering
- Mechanical Engineering