Intersection control for automated vehicles with MILP

We present a centralized Mixed Integer Linear Programming (MILP)-based strategy for intersection control in an urban environment of highly automated vehicles. The proposed strategy minimizes vehicle delay at the intersection approaches. This is achieved by decomposing the overall problem in three sub-problems: (i) the computation of feasible arrival times for each vehicle; (ii) a MILP scheduling problem of vehicle arrival time at the intersection; and (iii) a motion planning problem for vehicles to reach the intersection at the scheduled time. The resulting optimal arrival strategy is shown in simulation to adequately manage traffic for the simple intersection considered.