Physics based neural network for isolating faults

• US 20040064427A1
• Filed: 09/30/2002
• Published: 04/01/2004
• Est. Priority Date: 09/30/2002
• Status: Abandoned Application

• Current Assignee: United Technologies Corporation (Raytheon Technologies Corporation d/b/a RTX)
• Original Assignee: United Technologies Corporation (Raytheon Technologies Corporation d/b/a RTX)
• Inventors: Depold, Hans R., Sirag, David John Jr.
• Application Number:
  US10/261,265
• Publication Number:
  US 20040064427A1
• Time in Patent Office: Days
• US Class Current: 706/26
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology

Abstract

A PBNN for isolating faults in a plurality of components forming a physical system comprising a plurality of input nodes each input node comprising a plurality of inputs comprising a measurement of the physical system, and an input transfer function comprising a hyperplane representation of at least one fault for converting the at least one input into a first layer output, a plurality of hidden layer nodes each receiving at least one first layer output and comprising a hidden transfer function for converting the at least one of at least one first layer output into a hidden layer output comprising a root sum square of a plurality of distances of at least one of the at least one first layer outputs, and a plurality of output nodes each receiving at least one of the at least one hidden layer outputs and comprising an output transfer function for converting the at least one hidden layer outputs into an output.

Claims

1. A PBNN for isolating faults in a plurality of components forming:

a physical system comprising;

a plurality of input nodes each input node comprising;

a plurality of inputs comprising a measurement of said physical system; and

an input transfer function comprising a hyperplane representation of at least one fault for converting said at least one input into a first layer output;

a plurality of hidden layer nodes each receiving at least one first layer output and comprising a hidden transfer function for converting said at least one of at least one first layer output into a hidden layer output comprising a root sum square of a plurality of distances of at least one of said at least one first layer outputs; and

a plurality of output nodes each receiving at least one of said at least one hidden layer outputs and comprising an output transfer function for converting said at least one hidden layer outputs into an output.

2. The PBNN of claim 1 wherein each of said input transfer functions comprise a domain knowledge efficiency and a flow influence coefficient.

3. The PBNN of claim 1 wherein each of said plurality of measurements is comprised of a percent change.

4. The PBNN of claim 3 wherein each of said measurements is normalized with a standard deviation of said measurements.

5. The PBNN of claim 1 wherein each of said plurality of output nodes further comprises at least one weight each associated with one of said at least one hidden layer outputs.

6. The PBNN of claim 5, wherein said at least one weight is altered to a value sufficient to provide increased functionality.