A Stochastic Optimal Control Approach to Real-time, Incident-Responsive Traffic Signal Control at Isolated Intersections

Real-time incident-responsive traffic control and management is vital to development of advanced incident management systems in ITS. More importantly, it provides, from an academic point of view, the linkages between incident detection, incident management, and traffic signal control. This study explores the application of a stochastic optimal control approach to real-time incident-responsive traffic control at isolated intersections. In the methodology development, time-varying lane traffic state variables and control variables are specified to characterize section-wide inter-lane and intra-lane traffic states under conditions of lane-blocking incidents. Following specification of system states, we formulated a discrete-time nonlinear stochastic model which comprises four types of equations, namely (1) recursive equations, (2) measurement equations, (3) incident-induced delay equations, and (4) boundary constraints. From the proposed stochastic model, we then developed a stochastic optimal control algorithm to update the time-varying control variables and lane traffic state variables in real time with lane-blocking incidents at isolated intersections. To generate efficiently traffic data used in model tests, we employed an advanced microscopic traffic simulator, Paramics, Version 3.0, which is developed to model and analyze ITS traffic flow conditions. The preliminary test results indicate that the proposed method can accomplish the goal of real-time incident-responsive traffic signal control. In addition to proposing a new methodology, we hope that this study can initiate investigation into real-time incident-responsive traffic control and management to achieve the final goal of network-wide incident-responsive traffic optimal control for incident management.