Online 3-Dimensional Path Planning with Kinematic Constraints in Unknown Environments Using Hybrid A* with Tree...

Area of Science:

• Robotics
• Artificial Intelligence
• Ocean Engineering

Background:

• Autonomous underwater vehicles (AUVs) require sophisticated path planning for 3D navigation.
• Existing path planners struggle with dynamic environments and complex motion constraints.

Purpose of the Study:

• To extend the hybrid A* (HA*) path planner for 3D environments.
• To enhance AUV safety by incorporating motion constraints and collision avoidance.
• To improve path planning efficiency in unexplored or partially known environments through tree pruning.

Main Methods:

• Extension of the hybrid A* (HA*) algorithm for 3D path planning.
• Integration of vehicle motion constraints for feasible path generation.
• Development of a tree pruning procedure for dynamic replanning in unknown environments.

Main Results:

• The enhanced HA* planner successfully generates safe and feasible paths in 3D environments.
• Simulations demonstrate superior performance of HA* in known environments compared to other algorithms (planning time, path length, success rate).
• The tree pruning method significantly reduces total planning time in unknown environments.

Conclusions:

• The extended HA* path planner offers a robust solution for AUV navigation in complex 3D terrains.
• The tree pruning mechanism provides a significant advantage for AUVs operating in dynamic and uncertain underwater settings.
• This research contributes to more efficient and reliable autonomous underwater operations.