A Hierarchical Shared-Control Framework for Distributed Drive Electric Vehicles Considering Driving Ability and Energy Efficiency

To address the problem of personalized and ability-aware shared control for distributed drive electric vehicles, this paper proposes a hierarchical shared control framework that simultaneously accounts for driving ability and energy efficiency. First, a driving-ability metric is constructed by jointly considering path-tracking accuracy and handling stability, and a long short-term memory network (LSTM)-based short-horizon predictor is developed to capture its rapid variations under dynamic conditions. Then, an adaptive model predictive control (MPC) scheme is designed to coordinate the driver, active front-wheel steering (AFS), and direct yaw moment control (DYC) within a multi-agent formulation, dynamically adjusting multi-objective weights and authority allocation according to the predicted ability. At the lower layer, an energy-aware Nash-game torque vectoring allocation (TVA) strategy is introduced, where each in-wheel motor minimizes its own energy loss while satisfying global longitudinal and yaw-moment demands. Driver-in-the-loop experiments demonstrate that the proposed strategy significantly improves the level of human-machine cooperation, achieving approximately a 14.5% improvement in tracking accuracy and a 4.6% improvement in energy efficiency, while maintaining vehicle stability.