Establishment of a two-axis commercial vehicle 6DOF prediction model for nonlinear MPC controller

Abstract

Most of the commercial vehicle dynamics models used by nonlinear MPC controllers are 3-DOF models, which do not reflect the motion constraints of sprung and unsprung masses, the coupling relationship between translational and rotational motion, and the coupling characteristics of lateral and longitudinal tire forces, so it is difficult to meet the model prediction requirements under extreme working conditions. Therefore, according to the dynamics theory, this paper firstly establishes a 14DOF model which can reflect the movement characteristics of the sprung mass in the lateral, longitudinal, and vertical space, the suspension characteristics and the movement characteristics of the wheels on both sides. Then, in order to reduce the computational burden of the model prediction and make it applicable to the nonlinear MPC controller, the 14DOF model is simplified to the 6DOF model. Finally, the established 6DOF model and the classical 3DOF commercial vehicle model were compared and evaluated based on the error in prediction horizon and performance in trajectory tracking. The results show that the model that, in extreme working conditions, the new 6DOF model has obvious accuracy advantages compared with the classical 3DOF model in the prediction horizon, and can achieve convergence to the reference states faster, so as to obtain a solution efficiency similar to or even higher than 3DOF.