Tube-based multi-objective robust model predictive steering and torque vectoring control of 4-IWD electric vehicles

Abstract

The four-independent-wheel-drive electric vehicle is a typical over-actuated system. With the increase of actuators, the system complexity has brought great challenges to steering control and torque distribution. Existing works mostly employ model-based approaches since they are able to handle multiple control outputs while accommodating constraints. However, the investigated works usually focus on a single objective, and simplified prediction models are typically utilized in practice to mitigate computational burdens. In that case, it is difficult to fulfill the drivers’ diverse driving needs, and the robustness of the controllers is inevitably affected by the model mismatch. In this work, considering stability, motor energy loss, and tire slip energy, a multi-objective control framework is proposed. Furthermore, to address robustness against the model mismatch, robust model predictive control is devised. Compared with the state-of-the-art, the effectiveness of the proposed method has been validated in simulation. In the full-throttle acceleration scenario, the energy consumption is effectively suppressed. The motor energy loss and tire slip energy are reduced by 18.1% and 13.7%, respectively. Under the double lane change maneuver, vehicle stability is enhanced. The sideslip angle and yaw rate tracking errors are reduced by 12.5% and 16.1%.