High-performance collision avoidance parallel parking strategy of the autonomous vehicle

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Autonomous vehicles face real-world driving uncertainties like uneven roads and climate change.
• These uncertainties impact the precision and safety of autonomous parallel parking maneuvers.
• Existing control strategies may not adequately address time-varying, unknown bounds of uncertainty.

Purpose of the Study:

• To develop a robust parallel parking strategy for autonomous vehicles operating under uncertain conditions.
• To enhance the accuracy and safety of autonomous parking through advanced control techniques.
• To validate the proposed strategy against conventional methods.

Main Methods:

• Determining the dynamic model of the uncertain autonomous vehicle using the hierarchical constraint approach.
• Developing a self-regulating leakage type adaptive robust control algorithm.
• Utilizing the Udwadia-Kalaba method for uncertainty analysis and control design.

Main Results:

• An analytical form of the dynamic model for the uncertain autonomous vehicle was successfully obtained.
• The proposed adaptive robust control strategy demonstrated effective parallel parking.
• The strategy proved superior to other control methods in simulations and tests.

Conclusions:

• The developed parallel parking strategy effectively manages uncertainties in autonomous driving.
• The adaptive robust control approach ensures accurate parking and collision avoidance.
• This research contributes to the advancement of reliable autonomous vehicle navigation systems.