Optimizing the Crossing Sequence in Autonomous Intersection Management with Travel Time and Energy Considerations

Connected and Automated Vehicles (CAVs) offer fine-grained control and connectivity with other vehicles and the surrounding infrastructure that enable safe and energy efficient navigation. In this paper, mathematical programming is employed to construct optimal CAV crossing sequences to minimize a combination of fuel consumption and travel time. The developed approach is based on the construction of the minimum energy profile for each CAV which allows the derivation of efficient crossing times, using Mixed Integer Programming, without the need to consider vehicle dynamics. Safety is then guaranteed by ensuring that the vehicles achieve the corresponding crossing times using Quadratic Programming. The effectiveness of the derived CAV sequences compared to FIFO is illustrated through extensive simulations. The results indicate that resequencing yields a minor reduction in travel time but a significant improvement in fuel consumption. Fuel consumption improves for increasing number of CAVs arriving at the intersection. CAV rescheduling further supports observations of previous works that small sacrifices in travel time result in considerable fuel savings.