Aircraft flight path noise reduction
A method for aircraft flight path generation includes, at a computing system, receiving, for a plurality of waypoints in a geographic area, predicted aircraft noise levels for an aircraft at each waypoint of the plurality of waypoints, the predicted aircraft noise levels predicted based at least in part on a plurality of flight parameters for the aircraft. The predicted aircraft noise levels are input to a flight path prediction system configured to generate a candidate flight path for the aircraft through the geographic area based at least in part on the predicted aircraft noise levels. The candidate flight path is output from the flight path prediction system, wherein the candidate flight path is predicted to result in less ground-level noise when followed by the aircraft as compared to an alternate flight path through the geographic area.
1 . A method for aircraft flight path generation, the method comprising:
at a computing system, receiving, for a plurality of waypoints in a geographic area, predicted aircraft noise levels for an aircraft at the plurality of waypoints in which a predicted aircraft noise level for each waypoint is predicted based at least in part on a plurality of flight parameters for the aircraft passing through the waypoint;
inputting the predicted aircraft noise levels to a flight path prediction system configured to generate a candidate flight path for the aircraft through the geographic area based at least in part on the predicted aircraft noise levels, wherein the flight path prediction system generates the candidate flight path by:
calculating, for each grid cell of a plurality of grid cells of a virtual grid overlaid on the geographic area, a grid-relative predicted noise level, wherein, for at least one grid cell of the plurality of grid cells, the grid-relative predicted noise level is interpolated from two or more predicted aircraft noise levels of two or more of the plurality of waypoints falling within the grid cell;
selecting an initial grid cell of the plurality of grid cells as a starting point of the candidate flight path; and
iteratively selecting one or more subsequent grid cells of the plurality of grid cells as extensions of the candidate flight path from the initial grid cell by, on each of one or more iterations:
identifying a set of candidate grid cells from among the plurality of grid cells, and
selecting a next grid cell from among the set of candidate grid cells as an extension of the candidate flight path based on the grid-relative predicted noise levels of the set of candidate grid cells and further based on whether the set of candidate grid cells satisfy one or more route constraints; and
outputting the candidate flight path generated by the flight path prediction system to a flight system configured to fly the aircraft along the candidate flight path.
2 . The method of claim 1 , wherein, for each grid cell of the virtual grid, the grid-relative predicted noise level of the grid cell is interpolated.
3 . The method of claim 1 , wherein the flight path prediction system implements a dynamic grid-based Viterbi algorithm to generate the candidate flight path.
4 . The method of claim 1 , further comprising generating a second candidate flight path using a weighted graph-based algorithm implemented by the flight path prediction system, wherein the second candidate flight path is generated as a sequence of waypoints connecting a starting waypoint to an ending waypoint within the geographic area.
5 . The method of claim 4 , wherein the weighted graph-based algorithm includes an A* pathfinder algorithm.
6 . The method of claim 5 , wherein the weighted graph-based algorithm further includes a D* Lite algorithm to account for dynamic updates to the one or more route constraints.
7 . The method of claim 1 , wherein the candidate flight path is predicted to cause an amount of ground-level noise at a ground location in the geographic area that is less than a predefined noise target.
8 . The method of claim 1 , wherein the one or more route constraints include one or more of a turning radius of the aircraft, coordinates of restricted airspace within the geographic area, weather conditions in the geographic area, and departure and arrival procedures applying to the geographic area.
9 . The method of claim 1 , wherein the predicted aircraft noise levels are predicted based at least in part on a plurality of historical measured noise levels for a plurality of prior aircraft flights through the geographic area.
10 . The method of claim 9 , wherein the predicted aircraft noise levels are predicted by a machine learning model trained based at least in part on the historical measured noise levels and historical flight parameters for the plurality of prior aircraft flights.
11 . The method of claim 10 , wherein the historical flight parameters include one or more of aircraft type, aircraft speed, altitude, time of day, and weather conditions for the plurality of prior aircraft flights.
12 . A computing system, comprising:
a logic subsystem; and
a storage subsystem holding instructions executable by the logic subsystem to:
receive, for a plurality of waypoints in a geographic area, predicted aircraft noise levels for an aircraft at the plurality of waypoints in which a predicted aircraft noise level for each waypoint is predicted based at least in part on a plurality of flight parameters for the aircraft passing through the waypoint;
input the predicted aircraft noise levels to a flight path prediction system configured to generate a candidate flight path for the aircraft through the geographic area based at least in part on the predicted aircraft noise levels, wherein the flight path prediction system generates the candidate flight path by:
calculating, for each grid cell of a plurality of grid cells of a virtual grid overlaid on the geographic area, a grid-relative predicted noise level, wherein, for at least one grid cell of the plurality of grid cells, the grid-relative predicted noise level is interpolated from two or more predicted aircraft noise levels of two or more of the plurality of waypoints falling within the grid cell;
selecting an initial grid cell of the plurality of grid cells as a starting point of the candidate flight path; and
iteratively selecting one or more subsequent grid cells of the plurality of grid cells as extensions of the candidate flight path from the initial grid cell by, on each of one or more iterations:
identifying a set of candidate grid cells from among the plurality of grid cells, and
selecting a next grid cell from among the set of candidate grid cells as an extension of the candidate flight path based on the grid-relative predicted noise levels of the set of candidate grid cells and further based on whether the set of candidate grid cells satisfy one or more route constraints; and
output the candidate flight path generated by the flight path prediction system to a flight system configured to fly the aircraft along the candidate flight path.
13 . The computing system of claim 12 , wherein, for each grid cell of the virtual grid, the grid-relative predicted noise level of the grid cell is interpolated.
14 . The computing system of claim 12 , wherein the instructions are further executable to generate a second candidate flight path using a weighted graph-based algorithm implemented by the flight path prediction system, wherein the second candidate flight path is generated as a sequence of waypoints connecting a starting waypoint to an ending waypoint within the geographic area.
15 . The computing system of claim 14 , wherein the weighted graph-based algorithm includes an A* pathfinder algorithm.
16 . The computing system of claim 12 , wherein the candidate flight path is predicted to cause an amount of ground-level noise at a ground location in the geographic area that is less than a predefined noise target.
17 . The computing system of claim 12 , wherein the one or more route constraints include one or more of a turning radius of the aircraft, coordinates of restricted airspace within the geographic area, weather conditions in the geographic area, and departure and arrival procedures applying to the geographic area.
18 . A method for aircraft flight path generation, the method comprising:
at a computing system, receiving, for a plurality of waypoints in a geographic area, predicted aircraft noise levels for an aircraft at the plurality of waypoints in which a predicted aircraft noise level for each waypoint is predicted based at least in part on a plurality of flight parameters for the aircraft passing through the waypoint, the predicted aircraft noise levels predicted by a machine learning model trained based at least in part on historical measured noise levels and historical flight parameters for a plurality of prior aircraft flights;
inputting the predicted aircraft noise levels to a flight path prediction system configured to generate a candidate flight path for the aircraft through the geographic area based at least in part on the predicted aircraft noise levels, wherein the flight path prediction system generates the candidate flight path by:
calculating, for each grid cell of a plurality of grid cells of a virtual grid overlaid on the geographic area, a grid-relative predicted noise level, wherein, for at least one grid cell of the plurality of grid cells, the grid-relative predicted noise level is interpolated from two or more predicted aircraft noise levels of two or more of the plurality of waypoints falling within the grid cell;
selecting an initial grid cell of the plurality of grid cells as a starting point of the candidate flight path; and
iteratively selecting one or more subsequent grid cells of the plurality of grid cells as extensions of the candidate flight path from the initial grid cell by, on each of one or more iterations:
identifying a set of candidate grid cells from among the plurality of grid cells, and
selecting a next grid cell from among the set of candidate grid cells as an extension of the candidate flight path based on the grid-relative predicted noise levels of the set of candidate grid cells and further based on whether the set of candidate grid cells satisfy one or more route constraints; and
outputting the candidate flight path generated by the flight path prediction system to a flight system configured to fly the aircraft along the candidate flight path.
19 . The method of claim 1 , wherein the flight system is an autonomous flight system configured to fly the aircraft along the candidate flight path without human intervention.
20 . The computing system of claim 12 , wherein the flight system is an autonomous flight system configured to fly the aircraft along the candidate flight path without human intervention.