Fault detection in artificial intelligence based air data systems

• US 20070130096A1
• Filed: 12/01/2005
• Published: 06/07/2007
• Est. Priority Date: 12/01/2005
• Status: Abandoned Application

• Current Assignee: Rosemount Aerospace, Inc. (Raytheon Technologies Corporation d/b/a RTX)
• Original Assignee: Rosemount Aerospace, Inc. (Raytheon Technologies Corporation d/b/a RTX)
• Inventors: Matheis, Brian, Curtiss, Ryan, Selvig, Andrew
• Application Number:
  US11/291,461
• Publication Number:
  US 20070130096A1
• Time in Patent Office: Days
• US Class Current: 706/45
• CPC Class Codes: G01P 13/025 indicating air data, i.e. flight variables of an aircraft, e.g. angle of attack, side slip, shear, yaw
• CPC Class Codes: G01P 21/025 for measuring speed of fluids; for measuring speed of bodies relative to fluids for measuring volume flow G01F25/10
• CPC Class Codes: G01P 5/14 by measuring differences of pressure in the fluid

Abstract

A method and apparatus for detecting a fault in a sensor for an air data system which uses artificial intelligence to generate air data parameters is disclosed. The method and apparatus generate air data parameters as a function of measured values such as static pressures. The system also generates a fault detection value based upon the received value. The fault detection value is then input into a second network having artificial intelligence to determine if a sensor has experienced a fault.

Claims

1. A method of providing fault isolation in an air data system which uses artificial intelligence to generate air data parameters, the method comprising:

receiving an air data pressure value from at least one air data sensor;

generating a fault detection value based upon the received air data pressure value,

wherein generating the fault detection value further comprises;

generating a predicted value for the received air data pressure value; and

comparing the received air data pressure value with the predicted value to generate the fault detection value; and

determining if a fault is present by processing the fault detection value through a fault

detection network having artificial intelligence.

3. The method of claim 1, wherein generating the predicted value for the received air data pressure value further comprises:

receiving additional air data parameters for the at least one air data sensor; and

searching a database of air data parameter values to identify an air data pressure value associated with the additional air data parameters.

4. The method of claim 3, wherein the database is an inverse look-up table.

5. The method of claim 4, wherein the inverse look-up table is an M-dimensional database, where M is equal to the number of additional air data parameters.

6. The method of claim 3, wherein the data stored in the database is data that is used to train the artificial intelligence of the air data system.

7. The method of claim 1, wherein comparing the predicted value with the received air data pressure value further comprises:

calculating a difference between the received air data pressure value and the predicted value.

2. (canceled)

8. (canceled)

9. A method of providing fault isolation in an air data system which uses artificial intelligence to generate air data parameters, the method comprising:

receiving a plurality of air data pressure values, each air data pressure value associated with a different air data sensor;

generating a fault detection value based upon at least one received air data pressure value, wherein generating the fault detection value further comprises;

creating a non-dimensional value for one of the plurality of received air data pressure values; and

processing the non-dimensional value for the one of the plurality of received air data pressure values through an artificial intelligence fault detection network.

10. The method of claim 9, wherein creating the non-dimensional value for one of the plurality of received air data pressure values further comprises:

averaging the plurality of received air data pressure values; and

dividing the one of the plurality of received air data pressure values by an average value of the plurality of received air data pressure values.

11. An air data system comprising:

a plurality of pressure sensing ports each providing one of a plurality of measured pressures; and

air data computer circuitry configured to use artificial intelligence to generate air data parameters as a function of the plurality of measured pressures, and configured to use artificial intelligence having a fault detection value as an input to identify a fault in one of the plurality of air data sensing ports.

12. The air data system of claim 11, wherein the air data computer circuitry is further configured to generate a predicted value for one of the measured pressures and to identify a fault based on processing a difference between the one of the measured pressures and the predicted value.

13. The air data system of claim 12, wherein the air data computer circuitry is configured to generate the predicted value for one of the measured pressures based on the generated air data parameters for an associated pressure port.

14. The air data system of claim 13, wherein the predicted value for the one of the measured pressures is generated from a database.

15. The air data system of claim 14, wherein the database is an M-dimensional inverse look-up table, where M is equal to the number of air data parameters calculated for a given measured pressure.

16. The air data system of claim 11, wherein the air data computer circuitry is further configured to generate a non-dimensional input, for each of the plurality of measured pressures, to the artificial intelligence to identify a fault.

17. The air data system of claim 16, wherein the non-dimensional input is generated by dividing one of the plurality of measured pressures by an average of the plurality of measured pressures.