Hybrid A* Pathfinding with Reeds-Shepp Paths

This project implements a Hybrid A* pathfinding algorithm with Reeds-Shepp paths for planning smooth, collision-free paths in a 2D environment. The algorithm is designed for robotic path planning applications where a non-holonomic robot (one that cannot move directly sideways) navigates from a start to a goal position, avoiding obstacles along the way.

Features

• Hybrid A* pathfinding algorithm for non-holonomic path planning.
• Reeds-Shepp paths for smooth turning and backward movement, allowing the planner to generate realistic paths.
• Collision detection to ensure paths avoid obstacles.
• SFML graphics integration for visualization of the map, the search process, and the final path.
• Precomputed heuristic for faster convergence, enhancing the performance of the pathfinding algorithm.

Dependencies

• SFML: For graphical rendering of the environment and visualization of the pathfinding process.
• OMPL: For computing Reeds-Shepp paths and collision checking, (see here to install).
• Boost: Required by OMPL for serialization and other utilities.

Warning

This code is macOS (silicon) compatible. If you are using a different OS, you may need to modify the build script to link the correct libraries.

Usage

1. Clone the repository:

git clone https://github.com/Faywyn/HybridAstar.git

2. Build the project:

./build.sh release

3. Run the executable:

./build.sh run

4. Press SPACE to restart the search process with a new random start and goal position.

Screenshots