Path Planning of Inboard Space Flying Robot in a Virtual 3D Environment

The paper considers the task of motion path planning for inboard space flying robots in a virtual environment with static and dynamic obstacles. This task involves a three-dimensional map construction of the virtual scene, as well as finding the robot's path based on this map. An approach was developed for mapping that uses bounding volumes surrounding the geometry of objects, as well as bounding boxes aligned along the world coordinate system axes. Based on this approach, original algorithms have been developed for constructing and updating a scene map with the implementation of fast intersection tests between containers and map cells. In turn, pathfinding of robot's motion with an adaptation of the heuristic algorithm A* for the three-dimensional case is proposed that is based on the representation of map in the form of a regular decomposition graph. In the proposed solution, path graph nodes are formed dynamically, and obstacle nodes are processed using special sets with their updating taking into account map changes. The approbation of the developed algorithms and approaches was carried out in virtual environment complex and showed their effectiveness and adequacy for real time inboard space flying robots motion simulation without collision with obstacles.