A real-time nonlinear MPC for extreme lateral stabilization of passenger vehicles

Abstract

Loss of lateral stability remains a major cause of road accidents in recent years. Further improvement of passenger vehicle's active safety requires a more efficient utilization of the tire-road friction. Nonlinear model predictive control (NMPC) is expected to fulfill such a role, as the nonlinear characteristics of the vehicle are included and the control input is optimized. However, the computational load can be excessive for onboard hardware, which hinders the NMPC from practical implementation. To tackle the problem, this study proposes a method to improve the computational efficiency in NMPC. The proposed solution consists of an explicitly stored look-up table for generating initial guesses and an online optimization component. The look-up table is based on the offline solution of a hybrid MPC controller. Through the simulation with multibody vehicle model, impressive control performance has been observed, as the vehicle can be stabilized from a side-slip angle of up to 0.5 rad.