Adaptive model predictive control for robust lateral motion tracking of semi-autonomous vehicles with dynamic...

Area of Science:

• Control Systems Engineering
• Robotics and Autonomous Systems
• Vehicle Dynamics

Background:

• Traditional controllers face challenges in maintaining vehicle stability and accuracy due to nonlinear dynamics and external disturbances.
• Semi-autonomous vehicles require robust control systems capable of adapting to real-time variations in vehicle parameters and environmental conditions.

Purpose of the Study:

• To develop and evaluate a novel adaptive model predictive control (AMPC) framework for robust lateral motion tracking in semi-autonomous vehicles.
• To enhance the adaptability and robustness of vehicle control systems operating under dynamically uncertain conditions.

Main Methods:

• Integration of real-time parameter estimation using recursive least squares with a predictive optimization structure.
• Development of a comprehensive simulation environment in MATLAB/Simulink for rigorous testing.
• Evaluation across diverse scenarios including aggressive lane changes, crosswind disturbances, and low-friction conditions (μ=0.4).

Main Results:

• AMPC demonstrated a 43% reduction in lateral tracking error and a 37% improvement in yaw angle RMSE compared to conventional MPC and LQR.
• Maintained peak yaw errors below 0.275 radians even under severe disturbances, with smooth steering control signals.
• Human-in-the-loop simulations confirmed stability and trajectory tracking with delayed driver intervention (1-3s).

Conclusions:

• The proposed AMPC framework offers superior robustness and adaptability for complex lateral control tasks in semi-autonomous vehicles.
• This scalable solution enhances safety and reliability in shared-control driving environments and advanced driver assistance systems (ADAS).