US 12,117,806 B1Granted

Cloud and edge integrated method and system for identifying parameters of servo motor

AssigneeGuangdong Polytechnic Normal University

InventorBitao Zhang

Life of this patent

Estimating the term…

Loading patent…

Abstract

The present invention discloses a cloud and edge integrated method and system for identifying parameters of a servo motor. The method comprises the following steps: acquiring parameters of the servo motor; transmitting the parameters of the servo motor to a cloud identification engine and an edge identification engine; performing data processing and identification analysis, by the cloud identification engine, the parameters of the servo motor to obtain an internal resistance of the motor Rs, and sending an identification result to the edge identification engine through a network; identifying, by the edge identification engine, the cloud identification result and the parameters of the servo motor to obtain identification results of d-q axes inductances and a flux linkage. The system comprises a servo motor parameter acquisition unit, a sending unit, a cloud identification engine, and an edge identification engine.

Claims6

Dependent claims

A cloud and edge integrated method for identifying parameters of a servo motor, comprising the following steps:

acquiring parameters of the servo motor;
transmitting the parameters of the servo motor to a cloud identification engine and an edge identification engine;
performing data processing and identification analysis, by the cloud identification engine, the parameters of the servo motor to obtain an internal resistance of the motor Rs, and sending an identification result to the edge identification engine through a network;
performing identification analysis, by the edge identification engine, the cloud identification result and the parameters of the servo motor to obtain identification results of d-q axes inductances of the motor and a flux linkage;
wherein a neural network identification algorithm in the cloud identification engine is as follows:
Rs=fk(·)*G2+fk−1(·)*G1
wherein G1 and G2 are weights of corresponding output functions, and G1+G2=1;
fk(·) represents a network output value at a time k, fk−1(·) represents a network output value at a time k−1, defined as follows:
fk(·)=iq*w1+id*w2+uq*w3+ud*w4+T*w5+w*w6
wherein wi represents a weight of an input quantity, i=1, 2, . . . , 6, and w1+w2+w3+w4+w5+w6=1, w is a motor speed, id and iq are d-q axes currents, ud and uq are d-q axes input voltages, and T is a motor temperature;
the edge identification engine identifies the parameters Lq, Ld and φf by using the following equations:
A
=
∑
i
=
1
n
⁢
y
i
(
x
i
-
1
n
⁢
∑
i
=
1
n
⁢
x
i
)
∑
i
=
1
n
⁢
x
i
2
-
1
m
⁢
(
∑
i
=
1
m
⁢
x
i
)
2
,
B
=
1
m
⁢
∑
i
=
1
n
⁢
(
y
i
-
Ax
i
)
,
wherein dimensions of yi and xi are m*n dimensions,
y
i
=
(
i
d
i
q
)
,
x
i
=
(
u
d
u
q
)
,
A
=
[
R
s
-
ω
⁢
L
q
ω
⁢
L
d
R
s
]
-
1
,
B
=
[
R
s
-
ω
⁢
L
q
ω
⁢
L
d
R
s
]
-
1
[
0
ω
⁢
φ
f
]
,
id and iq are the d-q axes currents, ud and uq are the d-q axes input voltages, T is the motor temperature, Rs is the internal resistance of the motor, ω is an angular velocity of the motor, Ld and Lq are the d-q axes inductances, and φf is the flux linkage.

147 more elements

A cloud and edge integrated system for identifying parameters of a servo motor, comprising:

a servo motor parameter acquisition unit used for acquiring parameter data of the servo motor;
a sending unit used for receiving the parameter data of the servo motor and sending it to a cloud identification engine and an edge identification engine so as to enable the cloud identification engine and the edge identification engine to identify the parameters of the servo motor respectively;
the cloud identification engine used for identifying the parameter data of the servo motor and sending an identification result to the edge identification engine;
the edge terminal identification engine used for performing identification analysis of the acquired parameter data of the servo motor and the cloud identification result to obtain target parameter results;
the cloud identification engine comprises a data processing module, a neural network identification module and an output module;
wherein,
the data processing module is used for processing and denoising the acquired parameter data of the servo motor;
the neural network identification module is used for identifying the acquired parameter data of the servo motor to obtain Rs;
the output module is used for transmitting the identification result to the edge identification engine;
the neural network identification module identifies the internal resistance of the motor Rs based on the following neural network identification algorithm:
Rs=fk(·)*G2+fk−1(·)*G1
wherein G1 and G2 are weights of corresponding output functions, and G1+G2=1;
fk(·) represents a network output value at a time k, fk−1(·) represents a network output value at a time k−1, defined as follows:
fk(·)=iq*w1+id*w2+uq*w3+ud*w4+T*w5+w*w6
wherein wi represents a weight of an input quantity, i=1, 2, . . . , 6, and w1+w2+w3+w4+w5+w6=1, w is a motor speed, id and iq are d-q axes currents, ud and uq are d-q axes input voltages, and T is a motor temperature;
the edge identification engine identifies the parameters Lq, Ld and φf by using the following equations:
A
=
∑
i
=
1
n
⁢
y
i
(
x
i
-
1
n
⁢
∑
i
=
1
n
⁢
x
i
)
∑
i
=
1
n
⁢
x
i
2
-
1
m
⁢
(
∑
i
=
1
m
⁢
x
i
)
2
,
B
=
1
m
⁢
∑
i
=
1
n
⁢
(
y
i
-
Ax
i
)
,
wherein dimensions of yi and xi are m*n dimensions,
y
i
=
(
i
d
i
q
)
,
x
i
=
(
u
d
u
q
)
,
A
=
[
R
s
-
ω
⁢
L
q
ω
⁢
L
d
R
s
]
-
1
,
B
=
[
R
s
-
ω
⁢
L
q
ω
⁢
L
d
R
s
]
-
1
[
0
ω
⁢
φ
f
]
,
id and iq are the d-q axes currents, ud and uq are the d-q axes input voltages, T is the motor temperature, Rs is the internal resistance of the motor, ω is an angular velocity of the motor, Ld and Lq are the d-q axes inductances, and φf is the flux linkage.

152 more elements

Dependencies are inferred from the claim text.