Innovative Traffic Signal Control with Transit Priority on Beacon Street, Brookline

Abstract

The objective of the project is to develop a traffic signal algorithm that can provide near-zero-delay to public transit service by detecting the arrival of a transit vehicle and giving it priority while minimizing the impact on private (motor vehicle) traffic. Innovative signal control algorithms were programmed and tested using the traffic simulation software VISSIM. Priority control is based on predicting a train’s arrival at a traffic signal using detectors located at three points in space, called stages. The first stage is about two signal cycles’ travel time upstream of the signal stage. The second stage is when the train arrives at the last stop before the signal. The third stage is when the train departs from the last stop before the signal. At each successive stage, the ability to predict the train’s arrival time is improved. At the first stage, small adjustments to the signal cycle (e.g. shorten or lengthen the green intervals) are made in order to increase the chance that the train arrives during a (future) green period. At the second and third stage, further adjustments are made. Priority tactics used include green extension, early green, phase rotation, omitting left turn phases, and inserting transit-only phases, while respecting pedestrian phases. Additionally, private traffic signal control aims to improve efficiency and minimize the effects of transit priority. Pedestrian signal are designed to provide green waves to reduce pedestrian’s delay, and dynamic co-ordination is used with private traffic to shorten the pedestrian signal cycle length. Our results show that signal priority can reduce the average intersection delay to public transit but 74 percent, and up to 90 percent at a single intersection. The improvements made with the new control algorithms also allow for the private traffic delay to be kept at current level. Using the pedestrian green wave signal design allows pedestrians moving at 4.2 ft/sec to cross the wide street in a single sage (one signal cycle) instead of two stages (two signal cycles). Overall, the dynamic co-ordination of traffic signals reduces a pedestrians’ delay approximately 0 percent at affected signalized intersections.