Mixed-objective robust hierarchical lateral control of autonomous distributed drive electric vehicles considering parametric uncertainties and energy loss

Abstract

To overcome the issues of parametric uncertainties, external disturbances, input saturation and energy loss in the lateral control of autonomous distributed drive electric vehicles (A-DDEVs), a mixed-objective robust hierarchical control strategy is proposed. Firstly, considering the uncertain tire cornering stiffness and velocity, a control-oriented constrained polytopic system model with disturbance is established, which integrates active front wheel steering (AFS), direct yaw moment control (DYC), and anti-rollover control (ARC) to maximize vehicle safety. Subsequently, an upper-level mixed-objective constrained robust model predictive control (RMPC) algorithm is developed, transforming the robust control problem–satisfying input/state constraints, desired quadratic performance, and H∞ performance–into a finite-dimensional convex optimization problem in terms of linear matrix inequalities (LMIs). For lower-level control, an optimal torque vectoring control (TVC) algorithm balancing safety and economy is proposed, employing a weighted sum of tire load rate and energy loss as the objective function. Here, the weight is dynamically adjusted via phase plane and fuzzy rules for holistic optimization. Finally, simulations across diverse driving scenarios validate the strategy's effectiveness and robustness.