Non-linear model predictive control for cooperative transportation with four wheel drive/steering autonomous driving

Abstract

This paper presents the development of an integrated non-linear model predictive control (NMPC) method for cooperative transportation with four-wheel drive steering autonomous vehicles. An integrated NMPC controller has been devised to simultaneously manage both longitudinal and lateral dynamics, ensuring enhanced coordination between speed and steering. This method effectively reduces the impact of external disturbances and model uncertainties, leading to smoother and more precise trajectory tracking. In this research, the control inputs have been divided into longitudinal and lateral components, allowing the system to be modeled and controlled as a holistic framework. Considering constraints on both steering angles and tire forces, the integrated NMPC aims to improve overall vehicle stability and passenger comfort. The proposed algorithm has been validated through extensive computer simulations using MATLAB and CarSim. The results demonstrated significant improvements in path tracking performance and synchronization between connected vehicles, compared to sliding mode control and decoupled MPC methods. Furthermore, the integrated NMPC showed substantial reductions in both lateral and longitudinal errors, highlighting its superior capability in handling the complex dynamics of cooperative transportation systems.