Drone obstacle avoidance using real-time wind estimation

Patent metadata

Field	Value
Publication number	US-10909864-B2
Application number	US-201816020338-A
Country	US
Kind code	B2
Filing date	Jun 27, 2018
Priority date	Jun 27, 2018
Publication date	Feb 2, 2021
Grant date	Feb 2, 2021

Abstract

System and techniques for drone obstacle avoidance using real-time wind estimation are described herein. A wind metric is measures at a first drone and communicated to a second drone. In response to receiving the wind metric, a flight plan of the second drone is modified based on the wind metric.

First claim

The invention claimed is:

1. A device for drone obstacle avoidance using real-time wind estimation, the device comprising:
processing circuitry; and
non-transitor machine readable media including instructions that, when executed by the processing circuitry, cause the processing circuitry to:
receive a set of wind metrics measured by several support drones operating in a three-dimensional wind field;
model the wind field by estimation of a set of differential equations based on the set of wind metrics received from the several support drones, wherein the estimation of the set of differential equations use characteristics of the several support drones including measured characteristics and intrinsic characteristics, the measured characteristics including thrust generated by the motor outputs and attitude of a drone; and
modify, a flight plan through the wind field based on the modeled wind field.

2. The device of claim 1 , wherein the wind metrics are measured by the several support drones by each of the several support drones being configured to:
measure flight control inputs to maintain a flight orientation for; and
calculate a wind metric based on the flight control inputs.

3. The device of claim 2 , wherein the flight control inputs are motor outputs to a plurality of vertically oriented motors.

4. The device of claim 3 , wherein the flight orientation is a hover.

5. The device of claim 1 , wherein the set of differential equations are represented by {dot over (x)}=f(ϕ, T, g, a)+h(ν, γ, ϕ, b), where f is a force from the vertically oriented motors to counteract gravity and accelerate, ϕ is the attitude towards the wind, T is a total thrust from the vertically oriented motors, g is gravity, a are the intrinsic characteristics, h is external force from the wind acting, ν is a velocity of the wind, γ is a direction of the wind, and b are aerodynamic characteristics.

6. The device of claim 1 , wherein, estimation of the set of differential equations includes providing, motor outputs and attitude of the several support drones are provided to an artificial neural network trained on a dataset that includes a plurality of wind velocities and directions, and corresponding motor outputs and attitudes to stabilize the orientation.

7. The device of claim 1 , wherein, to modify the flight plan, a path in which the wind field threatens to force a collision with an obstacle is avoided.

8. The device of claim 1 , wherein the wind field includes wake turbulence of an aircraft, and wherein the flight plan is modified to avoid the wake turbulence.

9. A method for drone obstacle avoidance using real-time wind estimation, the method comprising:
receiving a set of wind metrics measured by several support drones operating in a three-dimensional wind field;
modeling the wind field by estimation of a set of differential equations based on the set of wind metrics received from the several support drones, wherein the estimation of the set of differential equations use characteristics of the several support drones including measured characteristics and intrinsic characteristics, the measured characteristics including thrust generated by motor outputs and attitude of a drone; and
modifying a flight plan through the wind field based on the modeled wind field.

10. The method of claim 9 , wherein, to measure the wind metrics, each of the several support drones:
measures flight control inputs of the first drone to maintain a flight orientation for the first drone: and
calculates the wind metric based on the flight control inputs.

11. The method of claim 10 , wherein the flight control inputs are motor outputs to a plurality of vertically oriented motors.

12. The method of claim 11 , wherein the flight orientation is a hover.

13. The method of claim 9 , wherein the set of differential equations are represented by {dot over (x)}=f(ϕ, T, g, a)+h(ν, γ, ϕ, b), where f is a force from the vertically oriented motors to counteract gravity and accelerate, ϕ is the attitude towards the wind, T is a total thrust from the vertically oriented motors, g is gravity, a are the intrinsic characteristics, h is external force from the wind acting, ν is a velocity of the wind, γ is a direction of the wind, and b are aerodynamic characteristics.

14. The method of claim 9 , wherein estimation of the set of differential equations includes providing motor outputs and attitude of the several support drones to an artificial neural network trained on a dataset that includes a plurality of wind velocities and directions, and corresponding motor outputs and attitudes to stabilize the orientation.

15. The method of claim 9 , wherein modifying the flight plan includes avoiding a path in which the wind field threatens to force a collision with an obstacle.

16. The method of claim 9 , wherein the wind field includes wake turbulence of an aircraft, and wherein the flight plan is modified to avoid the wake turbulence.

17. At least one non-transitory machine readable medium including instructions for drone obstacle avoidance using real-time wind estimation, the instructions, when executed by a machine, cause the machine to perform operations comprising:
receiving a set of wind metrics measured by several support drones operating in a three-dimensional wind field;
modeling the wind field by estimation of a set of differential equations based on the set of wind metrics received from the several support drones, wherein the estimation of the set of differential equations use characteristics of the several support drones including measured characteristics and intrinsic characteristics, the measured characteristics including thrust generated by motor outputs and attitude of a drone; and
modifying a flight plan through the wind field based on the modeled wind field.

18. The at least one non-transitory machine readable medium of claim 17 , wherein, to measure the wind metrics, each of the several support drones:
measures flight control inputs of the first drone to maintain flight orientation for the first drone; and
calculates the wind metric based on the flight control inputs.

19. The at least one non-transitory machine readable medium of claim 18 , wherein the flight control inputs are motor outputs to a plurality of vertically oriented motors.

20. The at least one non-transitory machine readable medium of claim 19 , wherein the flight orientation is a hover.

21. The at least one non-transitory machine readable medium of claim 17 , wherein the set of differential equations are represented by {dot over (x)}=f(ϕ, T, g, a)+h(ν, γ, ϕ, b), where f is a force from the vertically oriented motors to counteract gravity and accelerate, ϕ is the attitude towards the wind, T is a total thrust from the vertically oriented motors, g is gravity, a are the intrinsic characteristics, h is external force from the wind acting, ν is a velocity of the wind, γ is a direction of the wind, and b are aerodynamic characteristics.

22. The at least one non-transitory machine readable medium of claim 17 , wherein estimation of the set of differential equations includes providing motor outputs and attitude of the several support drones to an artificial neural network trained on a dataset that includes a plurality of wind velocities and directions, and corresponding motor outputs and attitudes to stabilize the orientation.

23. The at least one non-transitory machine readable medium of claim 17 , wherein modifying the flight plan includes avoidin

Assignees

• Intel Corp

Inventors

• De La Guardia Gonzalez Rafael
• Pohl Daniel
• Gomez Gutierrez David
• Hippelein Andreas
• Campos Macias Leobardo
• Aldana Lopez Rodrigo
• Parra Vilchis Jose
• Vervoorst Jan Willem

Classifications

• B64U2201/10

  autonomous, i.e. by navigating independently from ground or air stations, e.g. by using inertial navigation systems [INS] · CPC title

• B64U2201/00

  UAVs characterised by their flight controls · CPC title

• G08G5/55

  for a single aircraft · CPC title

• G08G5/54

  for approach or landing · CPC title

• G08G5/52

  for take-off · CPC title