Rolling Horizon Approach for Real-Time Charging and Routing of Autonomous Electric Vehicles

The adoption of autonomous electric vehicles (AEVs) offers an opportunity to decarbonize the transportation sector while eliminating the human errors in driving accidents. However, adopting AEVs may impose challenges to the operation of power distribution systems to ensure the availability of power for charging a growing number of AEVs at different times and locations. This paper takes an opportunistic look at this problem and develops a rolling horizon model for coordinating the operation of electric autonomous ride-hailing systems with power distribution systems. The proposed model incorporates the most recent real-time information and the future expected value of energy level, spatial and temporal location of AEV fleet, traffic data, and passenger demand. Using this data, the proposed model adopts a rolling horizon approach to optimize the routing of AEVs to serve spatio-temporal passenger demand across the transportation network, while optimizing the time and location of AEVs charging to ensure the availability of energy to serve the passenger demand, and satisfying the operational constraints of the power distribution system. The proposed model is implemented on a test transportation system, coupled with the IEEE 33-bus test power distribution system. The numerical results demonstrate the capability of the proposed model in ensuring the reliability and quality of service for both electric autonomous ride-hailing and power distribution systems.