Study on unified control of drifting and conventional driving for path tracking
Active safety control under extreme driving conditions, such as transient drifting, is crucial for autonomous vehicles. However, standard Nonlinear Model Predictive Control (NMPC) frequently encounters infeasibility when strictly enforcing hard constraints under highly nonlinear dynamics, and traditional hybrid switching strategies often induce transient instability. Targeting distributed-drive vehicles, this paper proposes a unified, mode-free optimal control framework for full-condition path tracking based on the iterative Linear Quadratic Regulator (iLQR). First, by incorporating the distribution characteristics of tire slips under combined conditions, a continuous mapping relationship between tire forces and slip velocities across arbitrary slip states is established. This unifies the control problem at the fundamental dynamics level, thereby completely eliminating the need for controller switching. Second, a hierarchical constraint management strategy is introduced: soft constraints via penalty functions are employed during the optimization phase to ensure recursive feasibility under sudden disturbances, followed by strict dynamic hard constraints at the output level to guarantee actuator safety. Finally, the proposed framework is validated through CarSim-Simulink co-simulations and Hardware-in-the-Loop (HIL) tests under various road adhesion coefficients. The results demonstrate that, compared to standard NMPC, the proposed iLQR framework achieves high-precision path tracking across all conditions with millisecond-level computational efficiency. Furthermore, the actuator saturation trigger rate under extreme conditions is kept below 10.2%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$10.2\%$$\end{document}, exhibiting outstanding actuator protection, dynamic smoothness, and high engineering practicality.