Method for creating and choosing a determinate piloting strategy for an aircraft

Patent metadata

Field	Value
Publication number	US-9852640-B2
Application number	US-201514835114-A
Country	US
Kind code	B2
Filing date	Aug 25, 2015
Priority date	Oct 14, 2014
Publication date	Dec 26, 2017
Grant date	Dec 26, 2017

Abstract

A method includes generating an initial flight context that includes a set of unresolved flight constraints. The method further includes generating a decision tree based on the initial flight context. The decision tree includes a plurality of flight contexts corresponding to leaves of the decision tree. The method also includes expanding the decision tree according to a greedy best-first strategy that selects a flight context of the decision tree for expansion based on a fitness value assigned to the flight context. The method includes selecting, as a piloting strategy, a flight context of the decision tree that includes a sequence of actions to resolve each of the unresolved flight constraints. The method further includes flying an aircraft according to the piloting strategy.

First claim

What is claimed is:

1. A method comprising:
generating a set of flight contexts including at least an initial flight context, wherein each flight context of the set of flight contexts includes a fitness value;
repeatedly performing operations comprising:
selecting a flight context that has a highest fitness value from the set of flight contexts;
resolving at least one flight constraint selected from a set of unresolved flight constraints of the selected flight context to obtain at least one expanded flight context;
allocating a new fitness value to the at least one expanded flight context; and
adding the at least one expanded flight context to the set of flight contexts;

after repeatedly performing the operations, selecting, from the set of flight contexts, the flight context that has the highest fitness value as a piloting strategy for flying an aircraft; and
automatically piloting the aircraft according to the piloting strategy.

2. The method of claim 1 , wherein each flight context of the set of flight contexts further includes a set of resolved flight constraints, the set of unresolved flight constraints, a sequence of actions to be performed by the aircraft, or a combination thereof.

3. The method of claim 2 , wherein the set of resolved flight constraints includes a set of resolved altitude flight constraints, a set of resolved speed flight constraints, or both, and wherein the set of unresolved flight constraints includes a set of unresolved altitude flight constraints, a set of unresolved speed flight constraints, or both.

4. The method of claim 2 , wherein the flight constraints are expressed in a flight intent description language (FIDL) comprising lexemes that univocally express a way of piloting the aircraft and univocally lead to a determinate aircraft trajectory when all configurable parameters involved in the FIDL lexemes are determined.

5. The method of claim 4 , wherein the configurable parameters include:
a range of numerical values of certain air vehicle state variables;
a set of discrete values for defining the reference for the numerical value of the aircraft state variables; and
a set of discrete values for defining different air vehicle settings.

6. The method of claim 2 , wherein each action of the sequence of actions is expressed in an intent composite description language (ICDL) comprising lexemes or intent composites that univocally express a way of piloting the aircraft and univocally lead to a determinate aircraft trajectory when all configurable parameters involved in the ICDL lexemes are determined.

7. The method of claim 6 , further comprising translating the sequence of actions from the ICDL into an aircraft intent description language (AIDL) that is a low level description language that univocally defines a precise way of commanding configurable parameters of the aircraft, yielding to a determinate aircraft trajectory.

8. The method of claim 2 , further comprising leaving at least one configurable parameter of at least one of the actions of the sequence of actions as undefined or expressed as a range of selectable values to enable tuning according to a user preference.

9. The method of claim 1 , wherein resolving the at least one flight constraint includes:
selecting an action that addresses the at least one flight constraint;
adding the selected action to a sequence of actions to be performed by the aircraft;
removing the at least one flight constraint from the set of unresolved flight constraints; and
adding the at least one flight constraint to a set of resolved flight constraints of the selected flight context.

10. The method of claim 1 , wherein the piloting strategy is selected from a subset of the set of flight contexts, wherein each flight context of the subset has an empty set of unresolved flight contexts.

11. The method of claim 1 , wherein the operations further comprise computing a partial trajectory corresponding to at least one action of a sequence of actions associated with the selected flight context.

12. The method of claim 11 , wherein an amount of time devoted to resolving the at least one flight constraint and to computing the partial trajectory is configurable to enable compromise between computing time and optimality of the piloting strategy.

13. The method of claim 11 , wherein the piloting strategy is selected from a subset of the set of flight contexts, wherein for each flight context of the subset, the partial trajectory corresponds to each action of the sequence of actions and is feasible.

14. The method of claim 13 , wherein the partial trajectory is feasible when the partial trajectory is flyable and the partial trajectory meets all the resolved flight constraints, wherein the partial trajectory is flyable if it is compatible with an aircraft performance model of the aircraft.

15. The method of claim 1 , wherein the new fitness value is:
proportional to a number of resolved flight constraints of the at least one expanded flight context,
proportional to an inverse of the number of resolved flight constraints of the at least one expanded flight context,
proportional to the number of unresolved flight constraints of the at least one expanded flight context,
proportional to the number of unresolved flight constraints of the at least one expanded flight context,
equal to the sum of an estimated cost of a computed partial trajectory corresponding to the at least one expanded flight context and an estimated cost of an estimated remaining trajectory,
or any combination thereof.

16. A method comprising:
generating an initial flight context;
adding the initial flight context to a set of flight contexts stored at a context database;
selecting a flight context that has a highest fitness value from the set of flight contexts, the selected flight context including a set of unresolved flight constraints, a set of resolved flight constraints, and a sequence of actions;
computing a partial trajectory for all the actions of the sequence of actions when the partial trajectory can be computed;
determining whether the partial trajectory is feasible when the partial trajectory is computed;
determining whether the set of unresolved flight constraints is empty when the partial trajectory is feasible;
yielding a piloting strategy corresponding to the selected flight context when the set of unresolved flight constraints is empty, otherwise:
resolving at least one flight constraint of the set of unresolved flight constraints of the selected flight context to form at least one expanded flight context;
adding the one or more expanded flight contexts to the set of flight contexts; and
selecting again the flight context that has the highest fitness value from the set of flight contexts; and

automatically piloting an aircraft unmanned according to the piloting strategy.

17. The method of claim 16 further comprising:
when the partial trajectory cannot be computed:
determining whether the set of unresolved flight constraints is empty.

18. The method of claim 16 further comprising:
when the partial trajectory is not feasible:
discarding the selected flight context from the set of flight contexts; and
yielding no solution when the set of flight contexts is empty, otherwise, selecting again the flight context that has the highest fitness value from the set of flight contexts.

19. A method comprising:
generating an initial flight context that includes a set of unresolved flight constraints;
generating a decision tree based on the initial flight context,

Assignees

• Boeing Co

Inventors

• Besada Portas Juan Alberto
• Lopez Leones Javier
• Frontera Guillermo

Classifications

• G08G5/025

  Physics · mapped topic

• G08G5/0052

  Physics · mapped topic

• G05D1/101 Primary

  specially adapted for aircraft · CPC title

• G08G5/0095

  Physics · mapped topic

• B64C19/00

  Aircraft control not otherwise provided for · CPC title