Application improvement of A* algorithm in intelligent vehicle trajectory planning

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence in Transportation

Background:

• Autonomous driving relies heavily on effective trajectory planning.
• The classical A* algorithm faces limitations in vehicle applications, including ignoring vehicle contours, producing non-smooth paths, and lacking speed planning.

Purpose of the Study:

• To address the limitations of the A* algorithm for autonomous vehicle trajectory planning.
• To propose an enhanced A* algorithm that considers vehicle contours, path smoothness, and speed planning.

Main Methods:

• A path processing method incorporating safe redundancy space considering vehicle contours.
• Path smoothing using Bessel curves and speed planning based on path curvature.
• A trajectory tracking algorithm combining an expert system for speed control and pure tracking theory for path control.

Main Results:

• The proposed method effectively smooths paths and plans speed based on curvature.
• The trajectory tracking algorithm improves path tracking accuracy by correcting steering angle based on speed influence.
• An expert system enhances speed tracking using vehicle acceleration characteristics.

Conclusions:

• The enhanced A* algorithm significantly improves applicability for automated vehicle trajectory planning.
• The integrated approach provides safer, smoother, and more accurate trajectory planning and tracking for autonomous vehicles.