Optimal Coordination and Control of Connected and Automated Vehicles at Intersections via Mixed Integer Linear Programming

Abstract

This paper addresses the crossing order problem for connected and automated vehicles at an intersection where the problem can lead to higher traffic congestion, especially in high traffic density. First, we formulated the problem with mixed integer linear programming where the solution yields optimal crossing time for each vehicle to cross the intersection. Then, we present an optimization framework to drive each vehicle towards the assigned crossing time while conserving the fuel consumption. Finally, we simulate the intersection scenario with the proposed solution, where it is shown that both congestion and fuel consumption can be reduced significantly.

Autonomous vehicle scheduling at signal-free intersections based on deep reinforcement learning

“…Mathematical optimization techniques typically establish mathematical models based on specific optimization objectives and obtain optimal traffic control strategies by solving the optimal solution. Mixed integer linear programming (MILP) is usually employed to formulate an optimization problem aimed at orchestrating the traffic flow through signal-free intersections in reference [4]. Similarly, X. Pan et al adopted a decentralized model predictive control (DMPC) approach to find optimal speed trajectory for each vehicle, focusing on minimizing of energy consumption and travel time [5].…”