A dynamic on-demand all-red clearance interval extension process for stochastic vehicular arrivals at signalized intersections.

摘要

The primary contribution of this research was the development of a decision algorithm for accurate real-time operation of dynamic on-demand extension of the all-red clearance interval. The algorithm provided the process for determining when extension of the all-red clearance interval should occur and the duration of the extension based on the real-time status of arriving vehicles obtained using modern radar detection. The operating parameters and procedures for the algorithm were specified based on findings from a field study of driver behavior.; The algorithm utilized a multi-step zonal classification process to determine if an approaching vehicle needed extended all-red time based on its current speed and position with respect to the intersection. Extended all-red time was provided based on the difference between the estimated time for the vehicle to clear the intersection compared to the amount of time remaining prior to the start of the opposing green indication. The algorithm was designed to manage simultaneous detection of multiple vehicles needing extended all-red time by determining the extension time based on the vehicle requiring the maximum time to clear the intersection.; The efficacy of the algorithm was evaluated through replication of field operations using vehicular data collected as part of this research as well as previous research completed by others. The replicated field evaluation showed the properly calibrated algorithm to provide maximum accuracy for classification of red-light-running vehicles. The evaluation also showed the calibrated algorithm to produce relatively few false extension calls from stopping vehicles, thereby minimizing the expected impacts to signal and traffic operations. Although field installation and testing of the algorithm was not performed, it was anticipated that long-term reductions in targeted red-light-running-related crashes will likely result at intersections implemented with an all-red extension system that utilizes the properly calibrated algorithm. The algorithm will likely provide the greatest safety benefits when used with protected-only lag left-turn phasing on the opposing approach. A proven long-term reduction in crashes would provide overall societal benefits by reducing the occurrence of deaths, injuries, and property damage (and costs associated with each) caused by targeted red-light-running events.