US 12,500,532 B1Granted

Online fault identification method for current sensor, diagnosis apparatus, and fault-tolerant control system

AssigneeSouthwest Jiaotong University

InventorsShuai Xu, Guangtong Ma, Chunxing Yao, Sijia Wu

Life of this patent

Estimating the term…

Loading patent…

Abstract

Disclosed are an online fault identification method for a current sensor, a diagnosis apparatus, and a fault-tolerant control system, relating to the technical field of diagnosis of faults of current sensors. The method includes defining a fault of the current sensor, and setting a corresponding label; then simulating a motor drive system, so as to obtain operation data in a normal mode and a fault mode, and making a data set; designing a neural network, and performing optimization processing on the neural network through the data set, so as to obtain an intelligent diagnosis model; and finally deploying the intelligent diagnosis model in an edge apparatus, so as to perform real-time online diagnosis on the fault of the current sensor; where the fault of the current sensor includes a saturation fault and a noise fault.

Claims12

Dependent claims

An online fault identification method for a current sensor, comprising defining a fault of the current sensor, and setting a corresponding label;

then simulating a motor drive system, so as to obtain operation data in a normal mode and a fault mode, and making a data set;
designing a neural network, and performing optimization processing on the neural network through the data set, so as to obtain an intelligent diagnosis model;
and finally deploying the intelligent diagnosis model in an edge apparatus, so as to perform real-time online diagnosis on the fault of the current sensor;
wherein the fault of the current sensor comprises a saturation fault and a noise fault,
wherein the saturation fault is defined as follows:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
Max
c
,
i
x
(
t
)
≥
Max
c
,
t
≥
t
f
i
x
(
t
)
,
Min
c
≤
i
x
(
t
)
≤
Max
c
,
t
≥
t
f
Min
c
,
i
x
(
t
)
≤
Min
c
,
t
≥
t
f
;
the noise fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
i
x
(
t
)
+
φ
⁡
(
t
)
,
t
≥
t
f
;
wherein
i(t) denotes an output current, ix(t) denotes a normal current without a fault, tf denotes a moment when the fault occurs, Maxc and Minc denote a maximum limit and a minimum limit of the current sensor respectively, and φ(t) denotes a noise of the sensor;
wherein the fault of the current sensor further comprises a stuck fault, a disconnection fault, a zero drift fault, and a gain fault;
the stuck fault is defined as:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
K
s
,
t
≥
t
f
;
the disconnection fault is defined as follows:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
0
,
t
≥
t
f
;
the zero drift fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
i
x
(
t
)
+
K
z
,
t
≥
t
f
;
and the gain fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
Gi
x
(
t
)
,
t
≥
t
f
;
i(t) denotes an output current, ix(t) denotes a normal current without a fault, tf denotes a moment when the fault occurs, and Ks, Kz, and G denote coefficients of the stuck fault, the zero drift fault, and the gain fault respectively;
wherein deploying the intelligent diagnosis model in an edge apparatus, so as to perform real-time online diagnosis on the fault of the current sensor, comprises:
in actual operation, transmitting real-time data to the edge apparatus via a serial communication line, so that the intelligent diagnosis model deployed in the edge apparatus is used to perform real-time online diagnosis on the fault of the current sensor.

242 more elements

A diagnosis apparatus, comprising a memory, a compiler, and a microprocessor;

wherein the memory stores an intelligent diagnosis model generated through an online fault identification method for a current sensor;
the compiler compiles the intelligent diagnosis model in the memory into an inference engine;
and the microprocessor performs online real-time diagnosis on a fault condition of the current sensor through the inference engine;
wherein the online fault identification method for the current sensor comprises:
defining a fault of the current sensor, and setting a corresponding label;
then simulating a motor drive system, so as to obtain operation data in a normal mode and a fault mode, and making a data set;
designing a neural network, and performing optimization processing on the neural network through the data set, so as to obtain an intelligent diagnosis model;
and finally deploying the intelligent diagnosis model in an edge apparatus, so as to perform real-time online diagnosis on the fault of the current sensor;
wherein the fault of the current sensor comprises a saturation fault and a noise fault
wherein the saturation fault is defined as follows:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
Max
c
,
i
x
(
t
)
≥
Max
c
,
t
≥
t
f
i
x
(
t
)
,
Min
c
≤
i
x
(
t
)
≤
Max
c
,
t
≥
t
f
Min
c
,
i
x
(
t
)
≤
Min
c
,
t
≥
t
f
;
the noise fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
i
x
(
t
)
+
φ
⁡
(
t
)
,
t
≥
t
f
;
wherein
i(t) denotes an output current, ix(t) denotes a normal current without a fault, tf denotes a moment when the fault occurs, Maxc and Minc denote a maximum limit and a minimum limit of the current sensor respectively, and φ(t) denotes a noise of the sensor;
wherein the fault of the current sensor further comprises a stuck fault, a disconnection fault, a zero drift fault, and a gain fault;
the stuck fault is defined as:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
K
s
,
t
≥
t
f
;
the disconnection fault is defined as follows:
i
⁡
(
t
)
=
{
i
x
(
t
)
,
0
≤
t
≤
t
f
0
,
t
≥
t
f
;
the zero drift fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
i
x
(
t
)
+
K
z
,
t
≥
t
f
;
and the gain fault is defined as follows:
i
⁡
(
t
)
=
⁢
{
i
x
(
t
)
,
0
≤
t
≤
t
f
Gi
x
(
t
)
,
t
≥
t
f
;
wherein
i(t) denotes an output current, ix(t) denotes a normal current without a fault, tf denotes a moment when the fault occurs, and Ks, Kz, and G denote coefficients of the stuck fault, the zero drift fault, and the gain fault respectively;
wherein deploying the intelligent diagnosis model in an edge apparatus, so as to perform real-time online diagnosis on the fault of the current sensor, comprises:
in actual operation, transmitting real-time data to the edge apparatus via a serial communication line, so that the intelligent diagnosis model deployed in the edge apparatus is used to perform real-time online diagnosis on the fault of the current sensor.

248 more elements

Dependencies are inferred from the claim text.

Facts

Application: 19/090,954Filed 26 Mar 2025
Priority: CN 202411279790.112 Sep 2024
Granted: 16 Dec 2025
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

H02P 6/12
Arrangements for controlling synchronous motors or other dynamo-electric motors using electronic commutation dependent on the rotor position;…:…
G01R 19/0092
Arrangements for measuring currents or voltages or for indicating presence or sign thereof: measuring current only
G01R 31/34
Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by…
G01R 31/343
Testing dynamo-electric machines: in operation