Combined dynamic traffic assignment and signal control: An application to a zone in Guangzhou

To enable an intelligent traffic light system (ITLS) to consider the interactions between the signal controls and the traffic flow distribution resulting from the selfish-routing behaviors of travelers, a dynamic origin-destination (O-D) demand estimation model and a dynamic combined traffic assignment and signal control (CTA-SC) model are needed. However, the ITLS may collect inaccurate and incomplete link flow data. How to estimate the dynamic O-D demands based on the inaccurate and incomplete dynamic link flows and guarantee their reliabilities is a big challenge. How to optimize the traffic signal controls when the dynamic O-D demands and the route choices of travelers are uncertain is also a popular research topic worldwide. This research treats the dynamic link flows as fuzzy sets and proposes a bi-level optimization problem to estimate the upper and lower bounds of the dynamic O-D demands. This research also assumes that the dynamic O-D demands and route choices of travelers are random parameters and establishes a bilevel robust optimization model for the signal controls. Both proposed bi-level models adopt Cell Transmission Models (CTMs) as the lower level problems. This research adopts iterative solution algorithm to solve the first bi-level demand estimation model and adopts heuristic to solve the robust optimization model for the signal controls.