Optimizing Energy Efficiency for Connected and Autonomous Electric Vehicles in the Context of Vehicle-Traffic Interaction

Abstract

The operational efficiency of connected and automated electric vehicles (CAEVs) is significantly impacted by the interplay between vehicle dynamics and traffic conditions. This study presents an energy-conscious optimization (ECO) approach aimed at enhancing the energy efficiency of CAEVs. This is achieved by addressing the dynamic constraints of the traffic environment and the vehicle's powertrain limitations within a unified framework. To develop the ECO approach, a novel bias deep compensative estimator is introduced to determine the parameters of the vehicle dynamics model. Utilizing these identified parameters, the traffic environment's constraints are translated into corresponding powertrain constraints for CAEVs. In the pursuit of optimal energy efficiency while adhering to powertrain limitations, a fresh velocity-torque coordinate system is established to normalize the constraints. Additionally, an iterative neighborhood search algorithm is proposed to systematically explore the coordinate system and identify the optimal efficiency point. With this newfound optimal efficiency point, a torque tracking control strategy is formulated. This strategy serves to guide the electric powertrain, ensuring its operation within the high-efficiency region. Real-world experiments are conducted to validate the effectiveness of the proposed approach, with a direct comparison against two prevailing state-of-the-art methods.