Data-Driven Intelligent Management of Energy Constrained Autonomous Vehicles in Smart Cities

Abstract

Intelligent transportation is an important component of future smart cities, and electric autonomous vehicles (EAVs) are envisioned to be the main form of transportation because EAVs can save energy, protect the environment, and improve service efficiency. With limited vehicle-specific energy storage capacity and overall constraint in the smart grid's electric load, we propose a novel intelligent management scheme to jointly schedule the travel and charging activities of the EAV fleet in one geographical area. This scheme not only schedules EAVs to meet the passengers' requests but also explores the matching problem between the energy requirement of EAVs and the deployment of charging piles in smart cities. We minimize the total cruise energy consumption of EAVs under the condition of limited energy supply while guaranteeing the quality-of-service (QoS). Network Calculus (NC) is extended to model the electric traffic flow in this paper. With the real-world electric taxi data in Beijing, simulation results demonstrate that the proposed scheme can achieve substantial energy reduction and remarkable improvements in both the order completion rate and utilization rate of the charging stations.