CN113656977A - Coil fault intelligent diagnosis method and device based on multi-mode feature learning - Google Patents

Coil fault intelligent diagnosis method and device based on multi-mode feature learning Download PDF

Info

Publication number
CN113656977A
CN113656977A CN202110980867.8A CN202110980867A CN113656977A CN 113656977 A CN113656977 A CN 113656977A CN 202110980867 A CN202110980867 A CN 202110980867A CN 113656977 A CN113656977 A CN 113656977A
Authority
CN
China
Prior art keywords
coil
fault
model
modal
feature learning
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110980867.8A
Other languages
Chinese (zh)
Other versions
CN113656977B (en
Inventor
郭丽
张锐
李�杰
刘胜涛
李长春
王钤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mianyang Weibo Electronic Co Ltd
Original Assignee
Mianyang Weibo Electronic Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mianyang Weibo Electronic Co Ltd filed Critical Mianyang Weibo Electronic Co Ltd
Priority to CN202110980867.8A priority Critical patent/CN113656977B/en
Publication of CN113656977A publication Critical patent/CN113656977A/en
Application granted granted Critical
Publication of CN113656977B publication Critical patent/CN113656977B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Algebra (AREA)
  • Computing Systems (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)

Abstract

The invention discloses a coil fault intelligent diagnosis method and a device based on multi-mode feature learning, wherein the method comprises the following steps: acquiring various characteristic data information of a coil, wherein each characteristic data is used as data of one mode; constructing a multi-feature data sparse constraint model of the coil according to the information of the various feature data of the coil; modeling the coil fault according to various characteristic data information of the coil to obtain a reconstructed fault model of the coil; fusing data processed in each mode according to a multi-feature data sparse constraint model of the coil and a reconstructed fault model of the coil to obtain a coil fault diagnosis model based on multi-mode feature learning, and solving the model to obtain a fault E of the coil under a certain mode signal of the coil; and judging the fault reason of the coil according to the size of the E. The invention improves the diagnosis accuracy of coil faults, is not limited to the fault diagnosis of the sensor coil, and can be popularized and applied to motor coils, engine coils and the like.

Description

Coil fault intelligent diagnosis method and device based on multi-mode feature learning
Technical Field
The invention relates to the technical field of intelligent diagnosis of coil faults of magnetic balance sensors, in particular to a coil fault intelligent diagnosis method and device based on multi-mode feature learning.
Background
Since the magnetic balance sensor is often used in a complicated and varied industrial environment, various failures of the coil inevitably occur during a long-term continuous operation due to the influence of the power supply condition and the load condition. These faults can seriously affect the reliability and safety of the coil operation. If not diagnosed and repaired in a timely manner, they can lead to sensor damage, resulting in serious consumer damage and property damage.
Coil faults are various, but the two categories are summarized, one category is electric faults such as current, voltage, frequency, power and the like; the other is mechanical failure, frame looseness, coil falling or breaking, turn-to-turn short circuit and the like. Usually, people pay attention to only one of the coil fault characteristics, such as only one of a current signal or a vibration signal, but the performance of the coil is manifold when the coil is in fault, and the method is time-consuming and has fault omission; the existing intelligent coil fault diagnosis method has low performance and low accuracy.
Disclosure of Invention
The invention aims to solve the technical problems of low diagnosis performance and low accuracy of the existing intelligent coil fault diagnosis method, and aims to provide an intelligent coil fault diagnosis method and device based on multi-modal feature learning, which can extract information from multiple possible factors to serve as fault features for diagnosis (the method is explained by extracting current signals and vibration signal features), and provide an intelligent fault diagnosis method based on multi-modal feature learning and sparse representation.
The invention is realized by the following technical scheme:
in a first aspect, the invention provides a coil fault intelligent diagnosis method based on multi-modal feature learning, which comprises the following steps:
acquiring various characteristic data information of a coil, wherein each characteristic data is used as data of one mode;
constructing a multi-feature data sparse constraint model of the coil according to the information of the various feature data of the coil; modeling coil faults (data errors) according to various characteristic data information of the coil to obtain a reconstructed fault model of the coil;
according to a reconstructed fault model of the coil, fusing data processed in each mode to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil; and judging the fault reason of the coil according to the size of the E.
And judging the fault reason of the coil according to the size of the E, wherein the fault judging method comprises the following steps: e.g. of the typeijRepresents the ith row and jth column element in E when
Figure BDA0003228955060000021
When the coil is in failure, the coil is judged, delta is the quality coefficient of the coil and represents the quality of the coil, namely the anti-interference capability, and the delta is the ratio of inductive reactance (or capacitive reactance) to equivalent resistance; to further determine which fault caused the fault, further calculations may be performed in each mode
Figure BDA0003228955060000022
And when the number is more than 0, judging that the fault exists.
The working principle is as follows:
based on the problems of low diagnostic performance and low accuracy of the existing intelligent coil fault diagnostic method, the invention designs an intelligent magnetic balance sensor coil fault diagnostic method based on multi-modal characteristic learning of current signals, vibration signals and the like. The multi-modal signal processing method has not been applied to analysis and processing of electrical signals such as current signals, vibration signals, voltage signals, and the like. For these electrical parameter signals, it is difficult for a single mode to provide complete information about noise interference during system operation. The multi-mode fusion mode can integrate the characteristic information from different modes, draw the advantages of different modes and complete the integration of the information. For electrical signals, the difficulty in fusing extracted features of different modes is that most of the feature signals are weak signals, and the feature difference is not obvious. For this case, we amplify the signal characteristics of the electrical signal in its multi-modal processing. The multimodal fusion method of the present invention is shown in FIG. 2. The invention provides a specific method for extracting characteristics before diagnosing a fault coil.
According to the coil fault intelligent diagnosis method based on multi-mode feature learning, the diagnosis performance and accuracy of the coil fault are improved; the method is not limited to the fault diagnosis of the sensor coil, and can be popularized and applied to motor coils, engine coils, transformer magnetic cores and the like.
Further, the various characteristic data information of the coil comprises a current signal and a vibration signal of the coil.
Further, a multi-feature data sparse constraint model of the coil is constructed according to the information of the various feature data of the coil; the multi-feature data sparse constraint model of the coil is expressed as:
Figure BDA0003228955060000023
wherein D is(v)Representing the features of the v-th mode of the coil, D representing a characteristic signal component of m rows and n columnsThe matrix is a matrix of a plurality of matrices,
Figure BDA0003228955060000024
λ1representing a balance parameter to the sparse constraint term; i | · | purple wind1Is the L1 norm, which represents the sparsity constraint; the principle of the sparse representation theory is to convert a non-sparse original signal into sparse coefficients through a dictionary matrix, and the sparse coefficients are used for representing the original signal. So the sparse representation is also called sparse coding. The present step designs sparse representation coefficients in order to reflect the essential characteristics of the sensor coil signal with a smaller amount of data. In short, the sparse representation coefficients contain information of the original signal.
Further, modeling coil faults (data errors) according to various characteristic data information of the coil to obtain a reconstructed fault model of the coil; the reconstructed fault model of the coil is represented as:
Figure BDA0003228955060000031
wherein λ is2Is that
Figure BDA0003228955060000032
The balance parameter of (a); e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000033
E(v)indicating a fault in the coil in the case of the v-th mode signal.
Further, the objective function of the coil fault diagnosis model based on multi-modal feature learning is as follows:
Figure BDA0003228955060000034
wherein, beta(v)Tuning parameters for each modality; d(v)Features representing the v-th mode of the coil, D*A matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000035
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000036
E(v)indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure BDA0003228955060000037
The balance parameter of (a); s.t. represents a constraint, X(v)Indicating that the v-th modal signal was acquired.
Further, the objective function of the coil fault diagnosis model based on the multi-modal feature learning is optimized and solved by adopting a Lagrange multiplier method.
In a second aspect, the present invention further provides a coil fault intelligent diagnosis apparatus based on multi-modal feature learning, which is characterized in that the apparatus supports the coil fault intelligent diagnosis method based on multi-modal feature learning, and the apparatus includes:
the device comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring various characteristic data information of a coil, and the various characteristic data information of the coil comprises a current signal and a vibration signal of the coil; each kind of characteristic data is used as data of one modality;
the multi-feature data sparse constraint model building unit is used for building a multi-feature data sparse constraint model of the coil according to the information of various feature data of the coil;
the reconstruction fault model building unit of the coil is used for modeling coil faults (data errors) according to various characteristic data information of the coil to obtain a reconstruction fault model of the coil;
the coil fault diagnosis model building and solving unit is used for fusing data processed in each mode according to a reconstructed fault model of the coil to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil;
the coil fault diagnosis unit is used for judging the fault reason of the coil according to the size of E, wherein the fault reason of the coil is judged according to the size of E, and the fault judgment method comprises the following steps: e.g. of the typeijRepresents the ith row and jth column element in E when
Figure BDA0003228955060000041
When the coil is in failure, the coil is judged, delta is the quality coefficient of the coil and represents the quality of the coil, namely the anti-interference capability, and the delta is the ratio of inductive reactance (or capacitive reactance) to equivalent resistance; to further determine which fault caused the fault, further calculations may be performed in each mode
Figure BDA0003228955060000042
And when the number is more than 0, judging that the fault exists.
Further, the objective function of the coil fault diagnosis model based on multi-modal feature learning is as follows:
Figure BDA0003228955060000043
wherein, beta(v)Tuning parameters for each modality; d(v)Features representing the v-th mode of the coil, D*A matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000044
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000045
E(v)indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure BDA0003228955060000046
The balance parameter of (a); s.t. represents a constraint, X(v)Indicating that the v-th modal signal was acquired.
In a third aspect, the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the intelligent coil fault diagnosis method based on multi-modal feature learning when executing the computer program.
In a fourth aspect, the present invention further provides a computer-readable storage medium, where a computer program is stored, where the computer program, when executed by a processor, implements the method for intelligently diagnosing coil faults based on multi-modal feature learning.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the invention designs an intelligent diagnosis method for the coil fault of the magnetic balance sensor based on multi-modal characteristic learning of current signals, vibration signals and the like, and the invention considers that signals have multiple modes and applies a multi-modal signal processing method to the analysis and processing of electric signals such as current signals, vibration signals, voltage signals and the like; the essential characteristics of the sensor coil signals are reflected by using sparse constraint through smaller data quantity, and the amplitudes of the signals such as current, vibration and the like are increased and harmonic waves (noise) appear after a fault (such as turn-to-turn short circuit) occurs, so that the fault reason can be judged according to the size of the fault E of the coil by modeling the noise of the signals.
2. The invention designs the intelligent diagnosis method for the coil fault of the magnetic balance sensor based on the multi-mode characteristic learning of current signals, vibration signals and the like, thereby improving the diagnosis accuracy of the coil fault; the method is not limited to the fault diagnosis of the sensor coil, and can be popularized and applied to motor coils, engine coils, transformer magnetic cores and the like.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a flow chart of a coil fault intelligent diagnosis method based on multi-modal feature learning according to the invention.
FIG. 2 is a schematic flow diagram of the multimodal fusion method of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
The invention designs a magnetic balance sensor coil fault intelligent diagnosis method based on multi-modal characteristic learning of current signals, vibration signals and the like, which considers that signals have various modes, reflects the essential characteristics of the sensor coil signals through smaller data quantity by using sparse constraint, and increases the amplitudes of the signals of current, vibration and the like and generates harmonic waves (noise) according to the signals after faults (such as turn-to-turn short circuit) occur, so that the fault reasons can be judged according to the magnitude of the fault E of the coil by modeling the noise of the signals. The multi-modal signal processing method has not been applied to analysis and processing of electrical signals such as current signals, vibration signals, voltage signals, and the like. For these electrical parameter signals, it is difficult for a single mode to provide complete information about noise interference during system operation. The multi-mode fusion mode can integrate the characteristic information from different modes, draw the advantages of different modes and complete the integration of the information. For electrical signals, the difficulty in fusing extracted features of different modes is that most of the feature signals are weak signals, and the feature difference is not obvious. For this case, we amplify the signal characteristics of the electrical signal in its multi-modal processing. The multimodal fusion method of the present invention is shown in FIG. 2.
Example 1
As shown in figure 1, the invention discloses a coil fault intelligent diagnosis method based on multi-mode feature learning, and provides a specific method for extracting features before diagnosing a fault coil. The method comprises the following steps:
step 1: acquiring various characteristic data information of the coil, wherein the characteristic data information mainly comprises a current signal and a vibration signal of the coil, and certainly can also comprise a voltage signal, a frequency signal, a power signal, a turn-to-turn short circuit signal and the like; each type of feature data is represented as D data of one modality(v)Wherein D is(v)Representing the characteristics of the v-th modality.
Step 2: according to the information of various characteristic data of the coil, a multi-characteristic data sparse constraint model of the coil is constructed, and the original signals of various modes are represented by the least elements; the multi-feature data sparse constraint model of the coil is expressed as:
Figure BDA0003228955060000061
wherein D is(v)Features the coil in the v-th mode, D a matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000062
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; the principle of the sparse representation theory is to convert a non-sparse original signal into sparse coefficients through a dictionary matrix, and the sparse coefficients are used for representing the original signal. So the sparse representation is also called sparse coding. The present step designs sparse representation coefficients in order to reflect the essential characteristics of the sensor coil signal with a smaller amount of data. In short, the sparse representation coefficients contain information of the original signal.
And step 3: modeling coil faults (data errors) according to various characteristic data information of the coil to obtain a reconstructed fault model of the coil; the reconstructed fault model of the coil is represented as:
Figure BDA0003228955060000063
wherein λ is2Is that
Figure BDA0003228955060000064
The balance parameter of (a); e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000065
E(v)indicating a fault in the coil in the case of the v-th mode signal.
And 4, step 4: according to a reconstructed fault model of the coil, fusing data processed in each mode to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil; and judging the fault reason of the coil according to the size of the E. The design theory of the step 4 is that after a fault (such as turn-to-turn short circuit) occurs, according to theoretical research, the amplitudes of current and vibration of the fault are increased and harmonic waves (noise) occur, so that the fault reason can be judged according to the magnitude of E by modeling the noise of the fault.
And judging the fault reason of the coil according to the size of the E, wherein the fault judging method comprises the following steps: e.g. of the typeijRepresents the ith row and jth column element in E when
Figure BDA0003228955060000066
When the coil is in failure, the coil is judged, delta is the quality coefficient of the coil and represents the quality of the coil, namely the anti-interference capability, and the delta is the ratio of inductive reactance (or capacitive reactance) to equivalent resistance; to further determine which fault caused the fault, further calculations may be performed in each mode
Figure BDA0003228955060000067
And when the number is more than 0, judging that the fault exists.
Specifically, the objective function of the coil fault diagnosis model based on multi-modal feature learning is as follows:
Figure BDA0003228955060000068
wherein, beta(v)Tuning parameters for each modality; d(v)Features the coil in the v-th mode, D a matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000071
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000072
E(v)indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure BDA0003228955060000073
The balance parameter of (1).
Equation (3) can be expressed as lagrange multiplier:
Figure BDA0003228955060000074
wherein, Λ1,Λ2Is the lagrange multiplier and μ is the penalty factor for the lagrange term.
The solution process for company (4) is as follows:
I. update D(v)With respect to D(v)The following sub-problems:
Figure BDA0003228955060000075
wherein the content of the first and second substances,
Figure BDA0003228955060000076
can yield D(v)＝UΣVTWherein
Figure BDA0003228955060000077
ε is H(v)Of singular values of (A) and (B), U and V being represented by H(v)And singular value vectors after singular value decomposition.
Update E(v)With respect to E(v)The following sub-problems:
Figure BDA0003228955060000078
it can be seen that,
Figure BDA0003228955060000079
update S(v)With respect to S(v)The following sub-problems:
Figure BDA0003228955060000081
it can be seen that,
Figure BDA0003228955060000082
update D*With respect to D*The following sub-problems:
Figure BDA0003228955060000083
it can be seen that,
Figure BDA0003228955060000084
to sum up, E is obtained.
According to the coil fault intelligent diagnosis method based on multi-mode feature learning, the diagnosis performance and accuracy of the coil fault are improved; the method is not limited to the fault diagnosis of the sensor coil, and can be popularized and applied to motor coils, engine coils, transformer magnetic cores and the like.
Example 2
As shown in fig. 1 to 2, the present embodiment differs from embodiment 1 in that the present embodiment provides a coil fault intelligent diagnosis apparatus based on multi-modal feature learning, which supports the coil fault intelligent diagnosis method based on multi-modal feature learning described in embodiment 1, and the apparatus includes:
the device comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring various characteristic data information of a coil, and the various characteristic data information of the coil comprises a current signal and a vibration signal of the coil; each kind of characteristic data is used as data of one modality;
the multi-feature data sparse constraint model building unit is used for building a multi-feature data sparse constraint model of the coil according to the information of various feature data of the coil;
the reconstruction fault model building unit of the coil is used for modeling coil faults (data errors) according to various characteristic data information of the coil to obtain a reconstruction fault model of the coil;
the coil fault diagnosis model building and solving unit is used for fusing data processed in each mode according to a reconstructed fault model of the coil to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil;
the coil fault diagnosis unit is used for judging the fault reason of the coil according to the size of E, wherein the fault reason of the coil is judged according to the size of E, and the fault judgment method comprises the following steps: e.g. of the typeijRepresents the ith row and jth column element in E when
Figure BDA0003228955060000091
If so, the coil is judged to be faulty, delta is the quality coefficient of the coil and represents a lineThe circle quality, namely the anti-interference capability, is obtained by taking the ratio of the inductive reactance (or capacitive reactance) to the equivalent resistance; to further determine which fault caused the fault, further calculations may be performed in each mode
Figure BDA0003228955060000092
And when the number is more than 0, judging that the fault exists.
To further illustrate the present embodiment, the multi-feature data sparse constraint model of the coil is represented as:
Figure BDA0003228955060000093
wherein D is(v)Features the coil in the v-th mode, D a matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000094
λ1representing a balance parameter to the sparse constraint term; i | · | purple wind1Is the L1 norm, which represents the sparsity constraint.
To further illustrate the present embodiment, the reconstructed fault model of the coil is expressed as:
Figure BDA0003228955060000095
wherein λ is2Is that
Figure BDA0003228955060000096
The balance parameter of (a); e denotes an error matrix of m rows and n columns,
Figure BDA0003228955060000097
E(v)indicating a fault in the coil in the case of the v-th mode signal.
To further illustrate the present embodiment, the objective function of the coil fault diagnosis model based on multi-modal feature learning is:
Figure BDA0003228955060000098
wherein, beta(v)Tuning parameters for each modality; d(v)Features the coil in the v-th mode, D a matrix of m rows and n columns of characteristic signals,
Figure BDA0003228955060000099
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e(v)Indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure BDA00032289550600000910
The balance parameter of (1).
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (10)

1. A coil fault intelligent diagnosis method based on multi-mode feature learning is characterized by comprising the following steps:
acquiring various characteristic data information of a coil, wherein each characteristic data is used as data of one mode;
constructing a multi-feature data sparse constraint model of the coil according to the information of the various feature data of the coil; modeling the coil fault according to various characteristic data information of the coil to obtain a reconstructed fault model of the coil;
according to a reconstructed fault model of the coil, fusing data processed in each mode to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil; and judging the fault reason of the coil according to the size of the E.
2. The intelligent coil fault diagnosis method based on multi-modal feature learning of claim 1, wherein the plurality of feature data information of the coil comprises a current signal and a vibration signal of the coil.
3. The intelligent coil fault diagnosis method based on multi-modal feature learning according to claim 1, characterized in that a multi-feature data sparse constraint model of the coil is constructed according to various feature data information of the coil; the multi-feature data sparse constraint model of the coil is expressed as:
Figure FDA0003228955050000011
wherein D is(v)Features the coil in the v-th mode, D a matrix of m rows and n columns of characteristic signals,
Figure FDA0003228955050000012
λ1representing a balance parameter to the sparse constraint term; i | · | purple wind1Is the L1 norm, which represents the sparsity constraint.
4. The intelligent coil fault diagnosis method based on multi-modal feature learning according to claim 1, characterized in that modeling is performed on coil faults according to various feature data information of the coil to obtain a reconstructed fault model of the coil; the reconstructed fault model of the coil is represented as:
Figure FDA0003228955050000013
wherein λ is2Is that
Figure FDA0003228955050000014
The balance parameter of (a); e denotes an error matrix of m rows and n columns,
Figure FDA0003228955050000015
E(v)indicating a fault in the coil in the case of the v-th mode signal.
5. The intelligent coil fault diagnosis method based on multi-modal feature learning of claim 1, wherein the objective function of the coil fault diagnosis model based on multi-modal feature learning is as follows:
Figure FDA0003228955050000016
wherein, beta(v)Tuning parameters for each modality; d(v)Features representing the v-th mode of the coil, D*A matrix of m rows and n columns of characteristic signals,
Figure FDA0003228955050000021
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e denotes an error matrix of m rows and n columns,
Figure FDA0003228955050000022
E(v)indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure FDA0003228955050000023
The balance parameter of (a); s.t. represents a constraint, X(v)Indicating that the v-th modal signal was acquired.
6. The method as claimed in claim 5, wherein the objective function of the coil fault diagnosis model based on multi-modal feature learning is optimized by using Lagrangian multiplier method.
7. An intelligent diagnosis device for coil fault based on multi-modal feature learning, which is characterized in that the device supports an intelligent diagnosis method for coil fault based on multi-modal feature learning according to any one of claims 1 to 6, and the device comprises:
the device comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring various characteristic data information of a coil, and the various characteristic data information of the coil comprises a current signal and a vibration signal of the coil; each kind of characteristic data is used as data of one modality;
the multi-feature data sparse constraint model building unit is used for building a multi-feature data sparse constraint model of the coil according to the information of various feature data of the coil;
the reconstructed fault model building unit of the coil is used for modeling the coil fault according to the various characteristic data information of the coil to obtain a reconstructed fault model of the coil;
the coil fault diagnosis model building and solving unit is used for fusing data processed in each mode according to a reconstructed fault model of the coil to obtain a coil fault diagnosis model target function based on multi-mode feature learning; solving the coil fault diagnosis model target function based on multi-modal feature learning to obtain a fault E of the coil under a certain modal signal of the coil;
and the coil fault diagnosis unit is used for judging the fault reason of the coil according to the size of the E.
8. The intelligent coil fault diagnosis device based on multi-modal feature learning as claimed in claim 7, wherein the objective function of the coil fault diagnosis model based on multi-modal feature learning is as follows:
Figure FDA0003228955050000024
wherein, beta(v)Tuning parameters for each modality; d(v)Features representing the v-th mode of the coil, D*A matrix of m rows and n columns of characteristic signals,
Figure FDA0003228955050000025
λ1representing the balance parameter to the sparse constraint term, | · | | non-woven1Is the L1 norm, which represents the sparsity constraint; e denotes an error matrix of m rows and n columns,
Figure FDA0003228955050000026
E(v)indicating the existence of a fault in the coil in the case of the v-th modal signal; lambda [ alpha ]2Is that
Figure FDA0003228955050000027
The balance parameter of (a); s.t. represents a constraint, X(v)Indicating that the v-th modal signal was acquired.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements a method for intelligent diagnosis of coil faults based on multi-modal feature learning according to any one of claims 1 to 6 when executing the computer program.
10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements a method for intelligent diagnosis of coil fault based on multi-modal feature learning according to any one of claims 1 to 6.
CN202110980867.8A 2021-08-25 2021-08-25 Coil fault intelligent diagnosis method and device based on multi-mode feature learning Active CN113656977B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110980867.8A CN113656977B (en) 2021-08-25 2021-08-25 Coil fault intelligent diagnosis method and device based on multi-mode feature learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110980867.8A CN113656977B (en) 2021-08-25 2021-08-25 Coil fault intelligent diagnosis method and device based on multi-mode feature learning

Publications (2)

Publication Number Publication Date
CN113656977A true CN113656977A (en) 2021-11-16
CN113656977B CN113656977B (en) 2023-07-07

Family

ID=78481970

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110980867.8A Active CN113656977B (en) 2021-08-25 2021-08-25 Coil fault intelligent diagnosis method and device based on multi-mode feature learning

Country Status (1)

Country Link
CN (1) CN113656977B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114359660A (en) * 2021-12-20 2022-04-15 合肥工业大学 Multi-modal target detection method and system suitable for modal intensity change

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101340121A (en) * 2008-08-28 2009-01-07 湖北省电力试验研究院 Transient simulation simplified model for internal fault of generator, modeling method and application thereof
US20090091289A1 (en) * 2007-10-08 2009-04-09 University Of Victoria Innovation And Development Corporation Stator inter-turn fault detection of synchronous machines
CN104181460A (en) * 2014-07-07 2014-12-03 沈阳工业大学 Multi-information-fusion-based fault diagnosis method for vacuum circuit breaker of permanent magnetic mechanism
CN106441897A (en) * 2016-10-27 2017-02-22 石家庄铁道大学 Morphological component bearing failure diagnosis method based on dictionary study
KR20190072165A (en) * 2017-12-15 2019-06-25 주식회사 에이스이앤티 Fault diagnosis system of motor
US10443530B1 (en) * 2018-05-22 2019-10-15 Gm Global Technology Operations Llc. System with solenoid assembly and method of fault diagnosis and isolation
US20200386811A1 (en) * 2019-06-05 2020-12-10 Wuhan University Power electronic circuit fault diagnosis method based on extremely randomized trees and stacked sparse auto-encoder algorithm
CN112132102A (en) * 2020-09-28 2020-12-25 天津工业大学 Intelligent fault diagnosis method combining deep neural network with artificial bee colony optimization
CN112147494A (en) * 2020-07-10 2020-12-29 国网河北省电力有限公司雄安新区供电公司 Mechanical fault detection method for high-voltage vacuum circuit breaker
CN112345213A (en) * 2020-09-18 2021-02-09 华能河南中原燃气发电有限公司 Low-voltage direct-current circuit breaker mechanical fault diagnosis method
CN112926728A (en) * 2021-03-22 2021-06-08 南京理工大学 Small sample turn-to-turn short circuit fault diagnosis method for permanent magnet synchronous motor

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090091289A1 (en) * 2007-10-08 2009-04-09 University Of Victoria Innovation And Development Corporation Stator inter-turn fault detection of synchronous machines
CN101340121A (en) * 2008-08-28 2009-01-07 湖北省电力试验研究院 Transient simulation simplified model for internal fault of generator, modeling method and application thereof
CN104181460A (en) * 2014-07-07 2014-12-03 沈阳工业大学 Multi-information-fusion-based fault diagnosis method for vacuum circuit breaker of permanent magnetic mechanism
CN106441897A (en) * 2016-10-27 2017-02-22 石家庄铁道大学 Morphological component bearing failure diagnosis method based on dictionary study
KR20190072165A (en) * 2017-12-15 2019-06-25 주식회사 에이스이앤티 Fault diagnosis system of motor
US10443530B1 (en) * 2018-05-22 2019-10-15 Gm Global Technology Operations Llc. System with solenoid assembly and method of fault diagnosis and isolation
US20200386811A1 (en) * 2019-06-05 2020-12-10 Wuhan University Power electronic circuit fault diagnosis method based on extremely randomized trees and stacked sparse auto-encoder algorithm
CN112147494A (en) * 2020-07-10 2020-12-29 国网河北省电力有限公司雄安新区供电公司 Mechanical fault detection method for high-voltage vacuum circuit breaker
CN112345213A (en) * 2020-09-18 2021-02-09 华能河南中原燃气发电有限公司 Low-voltage direct-current circuit breaker mechanical fault diagnosis method
CN112132102A (en) * 2020-09-28 2020-12-25 天津工业大学 Intelligent fault diagnosis method combining deep neural network with artificial bee colony optimization
CN112926728A (en) * 2021-03-22 2021-06-08 南京理工大学 Small sample turn-to-turn short circuit fault diagnosis method for permanent magnet synchronous motor

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
孙曙光;李勤;杜太行;崔景瑞;王景芹;: "基于一维卷积神经网络的低压万能式断路器附件故障诊断", 电工技术学报, no. 12, pages 2562 - 2573 *
肖怀硕: "基于振动信号的断路器故障诊断与状态评估", 中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑, pages 042 - 1157 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114359660A (en) * 2021-12-20 2022-04-15 合肥工业大学 Multi-modal target detection method and system suitable for modal intensity change
CN114359660B (en) * 2021-12-20 2022-08-26 合肥工业大学 Multi-modal target detection method and system suitable for modal intensity change

Also Published As

Publication number Publication date
CN113656977B (en) 2023-07-07

Similar Documents

Publication Publication Date Title
CN112766342A (en) Abnormity detection method for electrical equipment
CN113656977A (en) Coil fault intelligent diagnosis method and device based on multi-mode feature learning
CN112926728B (en) Small sample turn-to-turn short circuit fault diagnosis method for permanent magnet synchronous motor
CN111830409A (en) Motor thermal fault diagnosis method and device based on deep neural network
CN110987396B (en) Intelligent fault diagnosis and service life prediction method for coal mining machine rocker arm
CN116338454A (en) New energy automobile electrical property detection method and device
CN111946695B (en) Actuator fault diagnosis method combining multi-channel residual error and deep learning
CN110717287B (en) Spatial steel structure support rigidity identification method based on temperature strain
CN116776747A (en) IGBT module service life calculation method and system
CN110018417B (en) Motor fault diagnosis method, system and medium based on radial stray magnetic flux detection
CN116415197A (en) Underground pipe gallery abnormality detection network and method based on attention mechanism
CN114113837B (en) Transformer live detection method and system based on acoustic characteristics
CN115358294A (en) Micro fault detection method for high-speed train traction system
CN115329815A (en) Bearing fault diagnosis method and system for optimizing SVM (support vector machine) model parameters through bubble entropy and AOA (automated optical inspection)
CN113761103A (en) Batch data processing method and device and electronic equipment
CN112098066A (en) High-voltage shunt reactor fault diagnosis method and system based on gate control circulation unit
CN116449204B (en) Fault detection method for opposed-piston magnetic force linear generator and related device
CN114781760B (en) Fault prediction method based on big data
CN118258606A (en) Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment
CN114487819A (en) Dual-stator permanent magnet linear motor demagnetization fault diagnosis system and method
CN114723078B (en) Fault diagnosis method, device, terminal equipment and storage medium
CN117369425B (en) Automobile instrument assembly fault diagnosis method, system, storage medium and computer
CN109426241B (en) Method and device for recording fault data of wind generating set
CN114322917B (en) Transformer winding detection method and device, electronic equipment and storage medium
US20240176318A1 (en) Computer-implemented method and device for predicting a state of a technical system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant