CPC G01C 21/3833 (2020.08) [G01C 21/3461 (2013.01); G05D 1/6445 (2024.01); G05D 2109/20 (2024.01)] | 12 Claims |
1. A method of on-board navigation through an unknown environment for an unmanned aircraft system (UAS), comprising:
using a sensor system to generate point cloud data representing an environment of the UAS as the UAS flies;
using a processing system to perform the following steps: processing the point cloud data to generate both a two-dimensional occupancy grid and a three-dimensional voxel map;
wherein the two-dimensional occupancy grid comprises cells having known and unknown data;
determining an amount of known-to-unknown data in each cell of the two-dimensional occupancy grid, thereby determining unexplored regions of the environment;
assigning a cost to each unexplored region, based at least in part on the amount of known-to-unknown data within cells of the region;
sorting the unexplored regions based on their costs, thereby determining an optimal region to explore;
using the three-dimensional voxel map to determine a safe area within the optimal region where the UAS may fly;
calculating a flight path in the safe area provided by the three-dimensional voxel map;
generating control signals to navigate the UAS based on the flight path; and
using the sensor system and the processing system to continuously repeat the method to generate control signals to navigate the UAS based on the flight path.
|
7. A navigation system for an unmanned aircraft system (UAS) operating in an unknown environment, comprising:
an on-board sensor system operable to generate point cloud data representing an environment of the UAS as the UAS flies;
an on-board processing system having a process for processing the point cloud data to generate both a two-dimensional occupancy grid and a three-dimensional voxel map; wherein the two-dimensional occupancy grid comprises cells having known and unknown data;
the on-board navigation system further having processes for determining an amount of known-to-unknown data in each cell of the two-dimensional occupancy grid, thereby determining unexplored regions of the environment; for assigning a cost to each unexplored region, based at least in part on the amount of known-to-unknown data within cells of the region; for sorting the unexplored regions based on their costs, thereby determining an optimal region to explore; for using the three-dimensional voxel map to determine a safe area within the optimal region where the UAS may fly; and for calculating a flight path in the safe area provided by the three-dimensional voxel map; and for generating control signals to navigate the UAS based on the flight path.
|