Air-Ground Robot Path Planning via Neighborhood Constraints

The patent segments the task points into neighborhood regions, where each region contains multiple task points that can be completed within a certain radius. This segmentation transforms the complex problem of visiting multiple discrete task points into a simpler problem of visiting fewer neighborhood regions, thereby reducing computational complexity while maintaining task completion capability.

The patent allows robots to visit neighborhood regions that may contain more task points than strictly necessary, rather than precisely targeting each individual task point. This partial action approach simplifies the planning problem by treating neighborhoods as atomic units, reducing the search space and computational burden while still ensuring all task points are covered.

The patent dynamically adjusts the neighborhood radius parameter to balance between calculation efficiency and model accuracy. By making the neighborhood size adaptable rather than fixed, the model can efficiently handle simple cases with larger neighborhoods while maintaining accuracy for complex scenarios by reducing neighborhood sizes, thus resolving the contradiction between speed and adaptability.

The patent changes the parameter representation from individual task point coordinates to neighborhood region definitions with adjustable radii. This parameter transformation allows the mathematical model to maintain calculation efficiency while incorporating more realistic task completion conditions, reducing the need for excessive simplifications and assumptions.

The patent segments the precise target points into broader neighborhood regions, allowing robots to achieve sufficient task completion accuracy by reaching any point within the neighborhood rather than the exact target point. This segmentation maintains adequate positioning accuracy while dramatically improving path planning efficiency by reducing the precision requirements.

The patent creates a simplified neighborhood region model that copies the essential characteristics of task point locations without requiring precise coordinate targeting. This abstracted representation maintains the functional accuracy needed for task completion while enabling more efficient path planning calculations.

The patent segments the air robot's task execution into multiple smaller neighborhood regions rather than requiring long-duration hovering at single locations. This allows the air robot to perform brief hovering tasks at multiple distributed points within neighborhoods, reducing total energy consumption while maintaining task completion capability.

The patent allows ground robots to partially complete tasks by reaching neighborhood boundaries, reducing the need for air robots to hover for extended periods. This partial action approach enables task completion with shorter, more energy-efficient air robot hovering intervals.