Path Planning for Wheeled Mobile Robot in Partially Known Uneven Terrain

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Path planning for wheeled mobile robots in partially known, uneven terrain presents significant challenges due to incomplete environmental data.
• Robot motion is heavily influenced by terrain characteristics like slope, steps, and unevenness, complicating navigation.

Purpose of the Study:

• To propose a novel hierarchical path planning approach for wheeled robots operating in partially known, uneven terrain.
• To enhance robot navigation by effectively integrating global and local path planning strategies.

Main Methods:

• Modeling the partially known uneven terrain environment, considering features such as slope, step, and unevenness.
• Utilizing the A* algorithm for global path planning based on the partially known terrain map.
• Employing Q-learning for local path planning to manage locally detected obstacles and impassable areas during global path tracking.

Main Results:

• The proposed hierarchical path planning approach generated satisfying paths in simulations and experiments.
• The method successfully avoided locally detected obstacles and impassable terrain areas.
• Performance was compared favorably against the classical A* algorithm and the artificial potential field method.

Conclusions:

• The hierarchical path planning approach is effective for wheeled robots in partially known uneven terrains.
• Integrating global planning (A*) with local planning (Q-learning) enhances navigation safety and efficiency.
• This method offers a robust solution for complex terrain navigation challenges.