Traffic Signal Timing and Trajectory Optimization in a Mixed Autonomy Traffic Stream

This study introduces a methodology for cooperative signal timing and trajectory optimization at intersections with mix of connected automated vehicles (CAVs) human-driven (HVs). We represent joint control as mixed-integer non-linear program, which is computationally complex. The developed provides balance between computational efficiency solution quality by (a) linearizing the nonlinear constraints reformulating problem tight convex hull solutions (b) decomposing intersection-level program into several lane-level programs. Hence, unique controller jointly optimizes trajectories CAVs on lane parameters associated that lane. setting will allow finding near-optimal small duality gaps complex different demand levels. Case results show proposed finds efficiently most 0.1% gap. compared an existing approach found 13% to 41% reduction in average travel time 1% 31% fuel consumption under scenarios.