Method and device for diagnosing wear of an electrical switching unit, and electrical unit comprising such a device

Patent metadata

Field	Value
Publication number	US-11404223-B2
Application number	US-201715831541-A
Country	US
Kind code	B2
Filing date	Dec 5, 2017
Priority date	Dec 16, 2016
Publication date	Aug 2, 2022
Grant date	Aug 2, 2022

Abstract

A method for diagnosing the state of wear of an electrical switching unit including an electrical unit monitoring phase. The monitoring phase uses learning data loaded previously and representative of the type of electrical unit, and initialization data corresponding to the unit to be monitored and stored in an initialization phase. The monitoring phase includes the measurement and the acquisition of a measurement curve on opening the electrical unit, the determination of the value of local descriptors of the measurement curve as a function of values of the measurement curve, of initialization data and of learning data, the determination of the positioning of local descriptor values, the determination of an overall state class as a function of the positioning values. The device and the electrical unit implement the method.

First claim

The invention claimed is:

1. A diagnostic method for diagnosing a state of wear of an electrical switching circuit component, comprising:
monitoring of said electrical circuit, wherein said monitoring includes using:
first learning data loaded previously and corresponding to a type of product representative of said electrical circuit, and
initialization data corresponding to said electrical circuit to be monitored and stored in an initialization phase and comprising at least one reference curve specific to the type of product to be monitored saved at a start of a life of the type of product,

said monitoring further comprises:
measuring and acquiring a measurement curve on opening of said electrical circuit,
determining values of local descriptors of said measurement curve as a function of values of said measurement curve, of saved initialization data, and of the first learning data loaded,
determining a positioning of values of local descriptors of the measurement curve relative to second learning data, the second learning data including descriptor trend curves of
a deviation or a difference in value between two points of the measurement curve,
a deviation of a variation, of a difference, or of a correlation between the measurement curve and the initialization data,
a deviation, of a variation, of a difference, or of a correlation between the measurement curve and the first learning data,
an integral of a signal of the measurement curve between two points, and
a value of a point of the measurement curve, and

determining an overall status class as a function of said values of the positioning of the values of local descriptors of the measurement curve relative to said second learning data including the descriptor trend curves,

wherein the initialization data is stored when variations of the measurement curve become stable, and
wherein the first learning data sets comprise at least one reference curve of a worn circuit component corresponding to an electrical signal during an opening of a worn circuit taken from the measurement curves at the end of life of electrical units used in a previous learning.

2. The diagnostic method according to claim 1 , wherein the determining of an overall status class comprises:
at least three determinations of intermediate status classes as a function of positionings closest to said local descriptors, and
selecting said overall class as a function of said intermediate status classes.

3. The diagnostic method according to claim 1 , wherein a first descriptor is representative of a difference in value between two noteworthy points of said measurement curve.

4. The diagnostic method according to claim 1 , wherein
a second descriptor is representative of a deviation between a measurement curve and a specific reference curve, the more distant the measurement curve is from the specific reference curve, the more the electrical circuit will be considered to be worn.

5. The diagnostic method according to claim 1 , wherein:
a third descriptor is representative of a deviation between a measurement curve and said worn reference curve, a closer the measurement curve is to the worn reference curve, the more the worn circuit will be considered to be worn.

6. The diagnostic method according to claim 1 , wherein a fourth descriptor corresponds to a curve surface or to an integral between two noteworthy points of the measurement curve or to two thresholds of values in a first direction of variation and in a second, opposite direction of variation.

7. The diagnostic method according to claim 1 , wherein a fifth descriptor corresponds to a variation of value of a noteworthy point including a change of sign of the variation of a signal on the measurement curve.

8. The diagnostic method according to claim 1 , wherein the second learning data are representative of trend curves of descriptors representative of predefined characteristics of said electrical signal during a life of said electrical circuit.

9. The diagnostic method according to claim 1 , wherein trend curves of descriptors are loaded in a straight line segment definition format to reduce memory space needed for said learning data.

10. The diagnostic method according to claim 9 , wherein values of said trend curves of descriptors that are preloaded are associated with status or wear classes of circuits to be monitored.

11. The diagnostic method according to claim 1 , further comprising:
a preliminary learning phase for recording, during a predetermined number of manoeuvres, learning data representative:
of at least one reference curve of a worn circuit corresponding to an electrical signal during an opening of said worn circuit, and
of at least two trend curves of descriptors representative of predefined characteristics of said electrical signal during a life of said electrical circuit,

storing said learning data,
loading said learning data, and
using the learning data in said electrical circuit monitoring.

12. The diagnostic method according to claim 1 , further comprising an initialization phase for determining said initialization data comprising a specific reference curve of the product to be monitored determined:
after a predetermined number of measurements, and/or
when the measurement curve varies very little between two consecutive measurements,
saving said specific reference curve, and
using the initialization data in said electrical circuit monitoring.

13. The diagnostic method according to claim 1 , wherein the descriptor trend curves are a plurality of curves generated based on the measurement curve.

14. A device for diagnosing the state of wear of an electrical switching circuit connected to an electromagnetic coil actuating contacts of said electrical switching circuit, wherein the diagnostic device comprises a processing circuit for implementing the diagnostic method according to claim 1 .

15. The diagnostic device according to claim 14 , wherein the processing circuit comprises a local diagnostic processing module close to said electrical switching circuit linked to an external processing module remote from said electrical switching circuit.

16. An electrical circuit, comprising:
power electrical contacts actuated by an electromagnetic control coil; and
a device for diagnosing a state of wear of an electrical switching circuit component connected to said electromagnetic coil actuating said contacts and configured to:
monitor said electrical circuit using first learning data loaded previously and corresponding to a type of product representative of said electrical circuit, and initialization data corresponding to said electrical circuit to be monitored and stored in an initialization phase and comprising at least one reference curve specific to the type of product to be monitored saved at a start of a life of the type of product,
measure and acquire a measurement curve on opening of said electrical circuit,
determine values of local descriptors of said measurement curve as a function of values of said measurement curve, of saved initialization data, and of the first learning data loaded,
determine a positioning of values of local descriptors of the measurement curve relative to second learning data, the second learning data including descriptor trend curves of
a deviation or a difference in value between two points of the measurement curve,
a deviation of a variation, of a difference, or of a correlation between the measurement curve and the initialization data,
a deviation, of a variation, of a difference, or of a correlation between the measurement curve and the f

Assignees

• Schneider Electric Ind Sas

Inventors

• Remy Mael Eugene Alain
• Desforges Jose
• Rolard Francois Jean Maurice

Classifications

• G06F15/00

  Digital computers in general (details G06F1/00 – G06F13/00); Data processing equipment in general · CPC title

• H01H47/002

  Monitoring or fail-safe circuits · CPC title

• H01H1/0015 Primary

  Means for testing or for inspecting contacts, e.g. wear indicator (measuring circuits G01R31/3274) · CPC title

• H01H51/04

  with single armature; with single set of ganged armatures · CPC title

• G01R31/3274 Primary

  Details related to measuring, e.g. sensing, displaying or computing; Measuring of variables related to the contact pieces, e.g. wear, position or resistance (measuring contact resistance G01R27/205) · CPC title