Real-time estimation of queue length at signalized intersections

This study develops the method of estimating queue length at a signalized intersection. The method simplifies the past queue length estimation method that was developed using shock wave theory. This simplified method avoids complexity with calculations of shock wave speeds and accounts for the variations in vehicle effective length. The numbers of cars and trucks in each lane were observed upstream of the stop line at a signalized intersection in Windsor, Ontario. Maximum queue length among lanes was estimated in each cycle using second-by-second vehicle count and occupancy data collected from 7 locations of detectors. As a result, the method generally estimated the queue length more accurately than the shock wave method and the estimation errors were relatively consistent regardless of detector locations. The findings provide insights into the development of simpler queue length estimation method and the selection of the optimal location of detectors for accurate queue length estimation