Vehicle Torque Model Predictive Control Based on Adaptive Soft Boundary Constraint

Two separate electric motors are mounted on the rear axis of a hybrid vehicle. This configuration enables torque distribution between the rear wheels, which can generate torque rotating the vehicle. The front axis is driven by a hybrid system composed of an engine, an electric motor and a generator. Model predictive control is used in this paper to distribute the torque of the front axis, the rear left wheel and the rear right wheel. Some boundaries are regarded as soft constrains to avoid unsolved cases. This paper also proposes an evaluation method based on the improvement of understeering degree of vehicle steady-state response. Finally, compared with the method of average torque distribution, it effectively realizes the differential torque and assists the driver in steering, and reduces the steering wheel angle by up to 8% under the same path tracking target.