Load extraction fault detection system

• US 5,506,789 A
• Filed: 03/10/1995
• Issued: 04/09/1996
• Est. Priority Date: 10/15/1993
• Status: Expired due to Term

• Current Assignee: Texas A&M University System
• Original Assignee: Texas A&M University System
• Inventors: Benner, Carl L., Russell, B. Don
• Primary Examiner(s): Ramirez, Ellis B.
• Application Number:
  US08/401,644
• Time in Patent Office: 396 Days
• Field of Search: 364/483, 364/492, 364/487, 364/486, 324/534, 324/535, 324/642, 324/509
• US Class Current: 702/58
• CPC Class Codes: H02H 1/0015 Using arc detectors

Abstract

A method and apparatus are provided for detecting an arcing fault on a power line carrying a load current. The apparatus monitors, compiles and analyzes sample parameters taken from successive cycles of the power line load signal indicative of power flow and possible fault events on the line, such as voltage or load current. The analyzed sample parameter is then used to determine a normal parameter corresponding to a normal load pattern for the line. The normal load parameter is extracted from the compiled sample parameter data, and the resulting difference signal waveform has a magnitude value which changes in amplitude corresponding to the load-extracted data which is used to determine whether a fault exists on the power line.

Claims

1. A method of detecting faults on an AC power line, said method comprising the steps of:

(A) sampling a power line load signal to produce a sampled load signal;

(B) identifying a predominant waveform signal component from said sampled load signal;

(C) comparing said predominant waveform signal component with said sampled load signal to produce a difference signal;

(D) assessing said difference signal to identify a fault on said AC power line; and

(E) repeating steps (A) through (D) during changing load conditions.

2. The method of claim 1 wherein said sampled load signal of said sampling step includes said predominant waveform signal component and, when a fault exists on said AC power line, a fault signal component.

3. The method of claim 2 wherein said sampled load signal of said sampling step includes a plurality of load signal cycles of said power line load signal.

4. The method of claim 3 wherein said identifying step includes the step of identifying said predominant waveform signal component with a sample-by-sample comparison of a load signal cycle to subsequent load signal cycles to identify load signal cycles with substantially the same magnitudes.

5. The method of claim 1 wherein said sampling step includes the step of storing said sampled load signal in an input buffer.

6. The method of claim 2 wherein said comparing step includes the step of comparing magnitude values of said predominant waveform signal component to corresponding magnitude values of said sampled load signal to produce said difference signal.

7. The method of claim 6 wherein said comparing step includes the step of storing said difference signal in an output buffer.

8. The method of claim 7 wherein said assessing step includes the step of assessing whether said comparing step resulted in a difference signal with magnitude values exceeding a predetermined threshold.

9. The method of claim 8 wherein said difference signal with magnitude values exceeding a predetermined threshold corresponds to said fault signal component.