Energy-efficient driving for distributed electric vehicles considering wheel loss energy: A distributed strategy based on multi-agent architecture

Abstract

Distributed electric vehicles equipped with four-wheel independent drive (4WID) and four-wheel independent steering (4WIS) systems offer trajectory tracking performance and energy-saving potential. However, the challenge remains in how to coordinate the steering angles and torques of the four wheels to balance both tracking accuracy and energy efficiency. Distributed control, which trades design complexity for control flexibility, is able to differentiate the control of different wheels according to the vehicle's driving state to reduce wheel loss energy, providing a new perspective for improving the energy-efficient potential of vehicles. In this paper, a physical-data-driven distributed predictive control strategy is proposed within a distributed control framework, and multi-agent vehicle and wheel energy consumption models are constructed. To address the increased energy consumption and reduced trajectory tracking accuracy caused by model mismatches, a novel physical-data-driven predictive model-building approach is introduced, with real-time updates facilitated by the Givens Rotation and forgetting mechanism. The weights of the optimization objective function are dynamically adjusted according to changes in the wheel states to achieve comprehensive optimization of trajectory tracking and energy efficiency. Experimental results demonstrate that the proposed control strategy significantly reduces driving energy consumption while improving trajectory tracking performance. Under the CLTC-P cycle condition, energy loss is reduced by 11.5 %; under S-curve and double lane change steering conditions, energy losses are reduced by 15.0 % and 16.6 %, respectively. These results fully validate the effectiveness and superiority of the proposed strategy in practical applications.