Integrated Control of Steering and Braking for Path Tracking Using Multi-Point Linearized MPC

Abstract

In this study, we propose an integrated braking and steering model predictive controller for stable and accurate path tracking. To perform model-based longitudinal control and lateral control at the same time, it is necessary to consider nonlinear characteristics caused by changing vehicle speed and nonlinear tire forces. This paper proposes a multipoint linearization method to minimize the linearization error, and a controller with good computational efficiency and accurate consideration of nonlinear vehicle behavior is also introduced. In addition, the proposed model predictive controller (MPC) actively utilizes the road friction limit constraint for each tire force to ensure vehicle stability. Through this, the proposed controller generates optimal braking and steering inputs for situations such as high-speed turns in which braking must be involved. Due to the proposed linearized model, the controller achieves a significant improvement in computational efficiency and good control performance similar to that of the nonlinear MPC. Comparison with other control methods and performance verification for various road conditions are performed through simulations, and the results show very efficient calculation while performing accurate path tracking.