Effect of stop line detection in queue length estimation at traffic signals from probe vehicles data

Stop line detectors are one of the most deployed traffic data collection technologies at signalized intersections today. Newly emerging probe vehicles are increasingly receiving more attention as an alternative means of real-time monitoring for better system operations, however, high market penetration levels are not expected in the near future. This paper focuses on real-time estimation of queue lengths by combining these two data types, i.e., actuation from stop line detectors with location and time information from probe vehicles, at isolated and undersaturated intersections. Using basic principles of statistical point estimation, analytical models are developed for the expected total queue length and its variance at the end of red interval. The study addresses the evaluation of such estimators as a function of the market penetration of probe vehicles. Accuracy of the developed models is compared using a microscopic simulation environment-VISSIM. Various numerical examples are presented to show how estimation errors behave by the inclusion of stop line detection for different volume to capacity ratio and market penetration levels. Results indicate that the addition of stop line detection improves the estimation accuracy as much as 14% when overflow queue is ignored and 24% when overflow queue is included for less than 5% probe penetration level.

► Develops models for queue lengths from stop-line detectors and probe vehicles.
► Gives a closed-form equation when overflow queue is ignored.
► Compares the estimation errors of the models with VISSIM-microscopic simulation.
► Shows numerical examples for various probe levels and volume-to-capacity ratios.
► The impact of stop-line detector reaches up to 24% especially at low probe levels.