Regression Model For Prioritization Of Corridors When Retiming Traffic Signals

Abstract

Traffic signals are used by transportation engineers to safely and efficiently manage vehicle, bicycle and pedestrian traffic on roadways. However, to maintain the effectiveness of traffic signals, traffic signals should be adjusted or retimed periodically to match the current traffic patterns. Traffic signal retiming is one of the most cost effective ways available to transportation engineers for reducing delays and mitigating congestion. A region may contain a large number of traffic signals all of which cannot be retimed at once due to limited availability of funds and other resources; only a subset of signals can be retimed at any given time. To make the most effective use of (limited) resources, signals to be retimed need to be selected carefully. This thesis proposes several regression models to estimate the monetary benefits obtained by retiming signals along a corridor based on corridor characteristics such as length, number of signals, and average daily traffic. The estimated benefits are used to rank order corridors to be retimed and those expected to yield the most benefit can then be selected for retiming. The thesis also validates the proposed models and investigates the sensitivity of the models to changes in the values of various factors.