Robust and Energy-Efficient Torque Vectoring for a Four in-Wheel Motor Electric Vehicle Based on Sliding Mode and Model Predictive Control.

The article focuses on the development of a robust and energy-efficient torque vectoring (TV) controller for four in-wheel motor electric vehicles (4-IWM EVs). It proposes a two-layer control system combining first-order sliding mode control (FOSM) with nonlinear disturbance observers (NDOBs) in the upper layer to enhance robustness against model uncertainties and road disturbances, and model predictive control (MPC) in the lower layer to optimally allocate torque among wheels to minimize total energy consumption. Simulations based on an ISO double lane change scenario on a rough road demonstrate that the proposed controller significantly improves trajectory tracking accuracy, reduces energy consumption by 84.11%, and decreases velocity tracking error by 99.20% compared to a benchmark PID controller. The study highlights the controller's potential for practical application in improving stability and efficiency of 4-IWM EVs under challenging driving conditions.