Research on Uncertainty based Adaptive Model Predictive Control for Autonomous Vehicles under High-speed Operating Conditions

CSME 2025/12
Volume 46 No.6 : 657-668

Research on Uncertainty based Adaptive Model Predictive Control for Autonomous Vehicles under High-speed Operating Conditions

Jia-Cheng Mai ^a, Yi-Fan Zhao ^b, Fei Liu ^a, Xiaofeng Weng ^a, Sheng Zhou ^a and Shaoxiang Feng ^b
^aSchool of Mechanical and Automotive Engineering, Shanghai University of Engineering Science, 333 Long Teng Road, Shanghai, 201620, P.R. China.
^bSchool of Mechanical and Equipment Engineering, Hebei University of Engineering Handan,Hebei,China,Postal Code:056038.

Abstract: This study presents an uncertainty model predictive control (UM-MPC) algorithm that considers the uncertainties inherent in both environmental conditions and model parameters. Its core aim is to bolster the tracking accuracy and control system stability of autonomous vehicles. Establishing a multi-cellular model to mitigate the impact of uncertain parameters in the model. Based on the characteristic that drivers adjust their attention concentration according to road changes, a dynamic rule for the weight matrix has been designed through a large amount of comparative experimental data, achieving a shift in the focus of the algorithm during rolling optimization. Additionally, an adaptive predictive adjustment function for weights is proposed, and the optimal solution is derived through offline analytical optimization and the improved Particle Swarm Optimization (PSO) algorithm. Through a Hardware-in-the-Loop platform, a comparative analysis was conducted with the Adaptive Model Predictive Control (AMPC) based on fuzzy rules, affirming the effectiveness of the algorithm.

Keywords: Model predictive control, Uncertainty, Multicellular model, Trajectory tracking.

Download PDF

TOP