Intelligent vehicle lateral control strategy research based on feedforward + predictive LQR algorithm with GA...

Area of Science:

• Automotive Engineering
• Control Systems Engineering
• Robotics

Background:

• Intelligent vehicle autopilot systems require precise lateral motion control for safety and performance.
• Existing control algorithms often face challenges with system lag and adaptability to varying conditions.

Purpose of the Study:

• To develop an advanced lateral motion control algorithm for intelligent vehicle autopilots.
• To improve tracking accuracy and vehicle stability using a novel control strategy.

Main Methods:

• A feedforward + predictive Linear Quadratic Regulator (LQR) algorithm was designed, incorporating Genetic Algorithm (GA) for parameter optimization.
• A Proportional-Integral-Derivative (PID) steering angle compensation controller was developed to correct lateral errors.
• The proposed controller was validated using Carsim-Simulink co-simulation for Double Lane Change (DLC) and Circular Condition Tests (CCT).

Main Results:

• The optimized LQR controller demonstrated over 50% improvement in lateral and heading error control compared to traditional LQR controllers.
• Vehicle sideslip angle and yaw rate were maintained within strict stability constraints (-0.05° to 0.05° and -0.15 rad/s to 0.10 rad/s, respectively).
• Enhanced tracking accuracy and adherence to vehicle stability constraints were achieved.

Conclusions:

• The proposed feedforward + predictive LQR algorithm, optimized with GA and PID compensation, offers superior lateral motion control for intelligent vehicles.
• This approach effectively addresses system lag and enhances adaptability, leading to improved vehicle stability and tracking performance.