Integrating A* and APF for Dynamic Path Planning for the Mars Rover: A Fusion Approach to Avoid Deceptive Terrain

Mars exploration rovers must navigate treacherous deceptive terrain, which consists of surfaces that may appear easy to traverse but pose a significant risk of wheel slip, sinking, or rollover. Currently, these hazards are identified manually by mission operators. We propose a path-planning framework that integrates the A* algorithm with an artificial potential field (APF) method. By overlaying a repulsive field onto deceptive terrain, the rover is automatically guided to circumvent hazardous areas, thereby enhancing mission safety. To mitigate the excessive detours sometimes generated by a simplistic A*–APF coupling, we introduce a dynamic gain-scheduling rule that reduces the repulsive force as the rover successfully navigates away from danger. Further, we extend the conventional 2D curvature penalty into three dimensions to dampen oscillations on uneven ground. The effectiveness of this approach is validated through simulations in a virtual-reality model of the Martian surface. Experimental results indicate that the proposed enhanced fusion planner generates trajectories that are shorter, less risky, and significantly smoother in comparison to baseline methods, the experimental results show that the improved A*-APF algorithm reduces path length by up to 24.31%, path oscillation by up to 82.61%, and cumulative risk by up to 36.51%.