Modelling, Simulation and Control of Signalized Intersections under Adverse Weather Conditions
MetadataShow full item record
Adverse winter weather has always been a cause of traffic congestion and road collisions. To mitigate the negative impacts of winter weather, transportation agencies have been introducing weather responsive traffic management strategies such as adaptive control of signalized intersections and variable speed limits. Currently, most traffic signal control systems are designed for normal weather conditions and are therefore suboptimal in terms of efficiency and safety for controlling traffic during winter snow events due to the changing traffic patterns and driver behavior. There is a lack of systemic guidance on weather responsive signal control from signal design manuals and guide books. Existing guidelines do not provide methodical approaches to help traffic operators determine how to deploy weather-responsive signal control strategies for a local network. Additionally, the magnitude of the benefits of implementing weather-responsive signal control strategies is largely unknown due to the lack of reliable evaluation tools. The main objectives of this thesis are therefore to develop quantitative understanding of the effects of winter weather on several key traffic parameters and to investigate the methods and potential of implementing weather-responsive signal control strategies during inclement winter weather conditions. This thesis research consists of three main components. First, we have examined the impacts of winter weather on two key traffic parameters, namely, saturation flow rate and start-up lost time. Field data including traffic video and road weather and surface conditions were collected in the winter of 2015, from which various traffic parameters were extracted from vehicle trajectories. Extensive statistical analyses, including categorical analysis, non-linear regression, and multivariate regression, were followed to develop models for the relationship between each traffic parameter and various influencing factors such as visibility, precipitation and road surface conditions. Second, we have focused on calibrating a microscopic simulation model that can be used to simulate traffic operations under adverse winter weather conditions. A video-based approach was proposed to calibrate three important driver behavior parameters, i.e., mean desired speed, median desired acceleration rate at speed 0, and a parameter reflecting mean safe following distance. This approach is more robust and reliable than the traditional calibration methods due to the fact that the individual parameters are estimated directly from field data in a physically consistent way as opposed to the traditional trial-and-error process. At last, we have investigated the potential benefits of implementing weather-specific signal control plans for isolated intersections as well as arterial corridors based on two case studies. For both case studies, three traffic demand scenarios, i.e., high, medium, and low, were considered. Evaluation results from both deterministic and simulation models show that implementing weather specific signal plans is most beneficial for intersections with a medium level of traffic demand. When the demand is very low or very high, such strategies has little benefit in terms of reducing traffic delay. It has also been found that the benefit of implementing weather-responsive plans is more compelling at an arterial-corridor level with signal coordination than at an isolated-intersection level.