Review of trajectory tracking control for intelligent vehicles considering model uncertainty

In the process of intelligent vehicles moving toward advanced autonomous driving, trajectory tracking control, as the core link of safety and precise positioning, is facing performance bottlenecks caused by the interweaving of model uncertainty, system delay, and chassis coupling effects. This article considers the problem of intelligent vehicle trajectory tracking control under model uncertainty, and systematically reviews the latest developments in three research directions: vehicle driving state estimation, trajectory tracking control algorithms, and chassis cooperative control. Firstly, based on the background of distributed drive electric chassis and wire controlled chassis technology, the coupling effect mechanism of model uncertainty on vehicle dynamics and execution system was analyzed. Secondly, the research status of high robustness state estimation methods in critical vehicle state estimation is reviewed from dimensions such as Kalman filtering, sliding mode observer, and data-driven estimation. Furthermore, the theory and methods of trajectory tracking control that balance vehicle stability, comfort, and energy consumption constraints while ensuring path tracking accuracy were summarized. Finally, focusing on the chassis coordination control framework, the integration trend of multi-dimensional actuator collaborative optimization, limit envelope control, fault-tolerant control, and digital twin technology for steering/braking/driving was explored. Based on the analysis of domestic and foreign literature and comparison of typical cases, this article points out that: (1) The mechanism model and data-driven fusion of the “vehicle road cloud” digital twin system can effectively compress uncertainty boundaries; (2) The adaptive state estimation framework coupled with multimodal sensors and deep learning has higher robustness in complex working conditions; (3) The collaboration of reinforcement learning and model predictive control is expected to achieve trajectory stability integrated optimization in extreme scenarios.