Adaptive super-twisting sliding mode control for path tracking in multi-wheel-independent-drive unmanned vehicles

This paper addresses the critical challenge of coordinating path tracking robustness and energy efficiency for multi-wheel independent drive (MWID) unmanned ground vehicles (UGVs) operating in complex off-road environments. Existing approaches often neglect the severe dynamic vertical load transfers induced by rough terrains, leading to compromised lateral stability and excessive energy dissipation. To resolve this limitation, a hierarchical control framework is proposed. The upper layer develops an adaptive super-twisting sliding mode controller (ASTSMC) governed by a β−β˙ phase-plane stability domain. This mechanism enables on-demand control intervention, utilising a Lyapunov-stable adaptive law to ensure the finite-time convergence of tracking errors while effectively suppressing chattering under bounded uncertainties. For the lower layer, an energy-aware distributed model predictive control (DMPC) framework is formulated to optimally allocate wheel torques. By embedding real-time vertical load dynamics as hard constraints, the multi-agent DMPC proactively prevents tire slip and minimises motor power consumption, overcoming the computational bottlenecks of centralised optimisation. Furthermore, to mitigate the multi-objective conflict between upper-layer robustness and lower-layer energy conservation, the whale optimisation algorithm (WOA) is employed for global offline parameter tuning. Comprehensive Simulink/TruckSim co-simulations and Hardware-in-the-Loop (HiL) experiments under extreme steering maneuvers and rough terrains validate the proposed framework.