Two-Stage Request Scheduling for Autonomous Vehicle Logistic System

James J.Q. Yu
IEEE Transactions on Intelligent Transportation Systems, in press

Autonomous vehicles are expected to play an important role in handling the last mile logistics in intelligent transportation systems thanks to their unmanned nature and full-fledged controllability. Recently, an autonomous vehicle logistic system (AVLS) was proposed, which employs autonomous vehicles to serve logistic requests and utilize the excessive renewable energy generated by distributed generations. In this paper, we propose an optimization problem for AVLS to develop schedules for request allocation, vehicle routing, and battery charging. By considering the unique characteristics of AVLS, the proposed scheduling problem can exploit its advantages in goods delivery and renewable energy utilization over existing logistic request allocation algorithms. We formulate the problem as a mixed integer non-linear program. To improve its scalability, we also devise a two-stage scheduling methodology to approach the optimal solutions of the original problem. We conduct comprehensive simulations to assess the performance of the proposed request scheduling problem and two-stage scheduling methodology. The results indicate that the proposed problem can improve the efficacy of AVLS in terms of total travel distance and utilized renewable energy, and the two-stage methodology can develop near-optimal solutions with notably reduced computation time.