CN115684908A - State monitoring method for power GIS equipment, storage medium and electronic device - Google Patents
State monitoring method for power GIS equipment, storage medium and electronic device Download PDFInfo
- Publication number
- CN115684908A CN115684908A CN202211341817.6A CN202211341817A CN115684908A CN 115684908 A CN115684908 A CN 115684908A CN 202211341817 A CN202211341817 A CN 202211341817A CN 115684908 A CN115684908 A CN 115684908A
- Authority
- CN
- China
- Prior art keywords
- signal
- state
- gis equipment
- graph
- equipment
- 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.)
- Pending
Links
Images
Abstract
The invention relates to the technical field of state monitoring of GIS equipment, and discloses a state monitoring method for electric GIS equipment, which comprises the following steps: s1: converting the GIS vibration signal obtained by collection into a map signal; s2: transforming the graph signal from the node domain to a graph frequency domain by a graph Fourier transform technique; s3: encoding the one-dimensional image signal into a two-dimensional image matrix by using an inverse trigonometric function; s4: training and calculating by using a convolutional neural network to realize state identification of the GlS equipment; the method has high calculation efficiency and accurate result, and the obtained two-dimensional map contains rich GIS equipment state information, thereby providing reliable theoretical and data support for state identification and fault diagnosis of the GIS equipment under different working conditions.
Description
Technical Field
The invention relates to the technical field of GIS equipment state monitoring, in particular to a state monitoring method, a storage medium and an electronic device for power GIS equipment.
Background
The gas insulated metal enclosed switch is a reliable device for controlling the power system, and the running state of the gas insulated metal enclosed switch directly influences the safe running of the whole power system. With the rapid development of power grid construction, the number of the GIS devices is greatly increased, so that effective monitoring of the operation state of the GIS devices is necessary to ensure the operation safety of the GIS devices. Most of conventional monitoring methods are based on manual inspection, the working efficiency is not high, and the full-time coverage monitoring of monitoring equipment is difficult to realize. Therefore, it is very important to provide a reliable online monitoring means for power equipment.
Because the structural vibration that GIS arouses in service can characterize power equipment's operating condition to a certain extent, consequently can realize discerning GIS's state through analysis vibration signal. In recent years, monitoring means based on vibration signals have been rapidly developed by virtue of high detection accuracy, simple principle and the like, and have been widely applied to the field of monitoring of the state of power equipment. However, the conventional detection method based on vibration signal analysis has a good recognition effect on simple vibration signals with a single mode, and for "non-periodic" and "non-stationary" signals affected by the actual working environment, the vibration transmission path and the coupling factors of multiple vibration sources, it is generally difficult to achieve effective extraction of early defect features. In view of the above, it is desirable to provide a status monitoring method, a storage medium and an electronic device for a power GIS device.
Disclosure of Invention
The present invention is directed to a method, a storage medium, and an electronic device for monitoring the status of a power GIS device, so as to solve the above-mentioned problems in the background art.
In order to achieve the purpose, the invention provides the following technical scheme: the state monitoring method for the power GIS equipment comprises the following steps:
s1: converting the GIS vibration signal obtained by collection into a map signal;
s2: transforming the graph signal from the node domain to the graph frequency domain by a graph Fourier transform technique;
s3: encoding the one-dimensional image signal into a two-dimensional image matrix by using an inverse trigonometric function;
s4: and (3) realizing the state identification of the GlS equipment by utilizing the training and calculation of a convolutional neural network.
Preferably, the specific method of S1 is: the vibration signal y = { y } of the GIS equipment 1 ,y 2 ,…y n Characterization of the map signal:
y→G＝{V,E}
wherein, V = { V 1 ,V 2 …V N Denotes a set of nodes, which can be used for the representation of different sample points, N denotes the number of vertices; e = { E = 1 ,E 2 …E M Denotes a set of connecting edges, which can be used to make a description of the connections between different nodes, M denotes the number of connecting edges,
further, a map signal vector X can be obtained:
X＝[X 1 ,X 2 …X N ] T
in the formula, X N Representing the vibration signal amplitude corresponding to the nth node, obviously, the order of the vertices determines the order of the map signal amplitudes,
further, the graph signal is transformed from the node domain to the graph frequency domain using a graph fourier transform:
in the formula, X GF (k) Representing a graph Fourier transform signal, k representing the number of nodes, x ki Represents the k < th >The fourier transform basis, "T" denotes transpose.
Preferably, the specific method of S3 is: the conversion of the one-dimensional graph Fourier transform signal and the two-dimensional matrix is realized by using an inverse trigonometric function:
in the formula, θ represents an angle obtained by an arcsine function, R ij The ith row and jth column values of the two-dimensional matrix are represented, and subscripts "i" and "j" represent the ith and jth nodes, respectively.
Preferably, the specific method of S4 is: the method comprises the following steps of utilizing a VGG16 neural network to realize effective identification of two-dimensional images of graphs obtained by different GIS equipment states, reducing input parameter dimensionality through laminated convolution by the neural network, and further improving the performance of feature extraction, wherein the realization formulas of an activation function ReLU, a pooling step, a full connection step and a SoftMax classification function are respectively as follows:
wherein x is an input; s is down A down-sampling function is represented that is,
the deviation in the scale is shown as a deviation,
the deviation in the horizontal direction is indicated,
denotes the ith cell of the l-1 st layer,
represents the weight from the ith cell to the lower J cell, I represents the number of samples, J represents the number of classification states, g ij Tag code representing class state j corresponding to sample i, E ij Representing a samplei is predicted as the probability of classification state j.
The invention also provides a state monitoring electronic device for the electric GIS equipment, which is used for the state monitoring method for the electric GIS equipment.
Preferably, the data sensing unit is used for acquiring a vibration signal of the target device.
Preferably, the data acquisition unit is used for processing original vibration signal data, and comprises the steps of collecting vibration signals obtained by the data sensing unit, amplifying the vibration signals, acquiring and storing the signals by using the ADC, and transmitting the data to the data processing unit.
Preferably, the data processing unit is used for processing and calculating vibration data, firstly mapping the vibration signal to an image domain, and mapping the vibration signal to the image frequency domain by means of an image fourier transform technology; then, encoding the one-dimensional image Fourier signal into a two-dimensional image matrix by means of an inverse trigonometric function; and finally, realizing intelligent identification of the state of the power GIS equipment by means of a convolutional neural network.
Preferably, the result display unit is used for displaying the image frequency signal coding images of the GIS equipment in different states and outputting state identification and classification accuracy.
The invention also provides a storage medium applied to the state monitoring device for the power GIS equipment, wherein the computer equipment comprises a storage medium, the storage medium stores a computer program, and the computer program is executed by a data processing unit to realize the state monitoring method for the power GIS equipment.
Compared with the prior art, the invention has the beneficial effects that:
1. the state monitoring method for the electric power GIS equipment overcomes the defect that the conventional monitoring method is difficult to obtain rich information of the running state of the GIS equipment, can effectively sense the state change degree of the electric power GIS equipment, realizes time domain-image frequency domain conversion of signals by using an advanced image signal processing technology, finally and intuitively displays the current state of the GIS equipment in an image form by using an inverse trigonometric function, is simple and convenient to realize and high in calculation efficiency, and finally realizes reliable identification of the state of the GIS equipment by using a convolutional neural network on the basis of obtaining a two-dimensional map capable of fully representing the current working condition of the GIS equipment.
Drawings
FIG. 1 is a system diagram of the method of the present invention;
FIG. 2 is a diagram of the implementation steps of the VGG16 neural network of the present invention;
FIG. 3 is a schematic diagram of the framework of the apparatus of the present invention;
FIG. 4 is a schematic diagram of the computer apparatus of the present invention;
FIG. 5 is a schematic diagram of an encoded image of a graph Fourier signal and a graph frequency signal of an exemplary fault of the present invention;
FIG. 6 is a graph of performance during training of the present invention;
FIG. 7 is a confusion matrix diagram according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, the present invention provides a method for monitoring the status of a power GIS device, including the following steps:
s1: converting the GIS vibration signal obtained by collection into a map signal;
s2: transforming the graph signal from the node domain to the graph frequency domain by a graph Fourier transform technique;
s3: encoding the one-dimensional image signal into a two-dimensional image matrix by using an inverse trigonometric function;
s4: and (3) realizing the state identification of the GlS equipment by utilizing the training and calculation of a convolutional neural network.
In this embodiment, the specific method of S1 is: vibrating signal y = { y ] of GIS equipment 1 ,y 2 ,…y n Characterization of the map signal:
y→G＝{V,E}
wherein V = { V = { (V) 1 ,V 2 …V N Denotes a set of nodes, which can be used for the representation of different sample points, N denotes the number of vertices; e = { E = 1 ,E 2 …E M Denotes a set of connecting edges, which can be used to make a description of the connections between different nodes, M denotes the number of connecting edges,
further, a map signal vector X can be obtained:
X＝[X 1 ,X 2 …X N ] T
in the formula, X N Representing the vibration signal amplitude corresponding to the nth node, obviously, the order of the vertices determines the order of the map signal amplitudes,
further, the graph signal is transformed from the node domain to the graph frequency domain using a graph fourier transform:
in the formula, X GF (k) Representing a graph Fourier transform signal, k representing the number of nodes, x ki Denotes the kth graph fourier transform basis and "T" denotes transpose.
In this embodiment, the specific method of S3 is: the conversion between the one-dimensional graph Fourier transform signal and the two-dimensional matrix is realized by using an inverse trigonometric function:
in the formula, θ represents an angle obtained by an arcsine function, R ij The ith row and jth column values of the two-dimensional matrix are represented, and subscripts "i" and "j" represent the ith and jth nodes, respectively.
In this embodiment, the specific method of S4 is: the method comprises the following steps of utilizing a VGG16 neural network to realize effective identification of two-dimensional images of graphs obtained by different GIS equipment states, reducing input parameter dimensionality through laminated convolution by the neural network, and further improving the performance of feature extraction, wherein the realization formulas of an activation function ReLU, a pooling step, a full connection step and a SoftMax classification function are respectively as follows:
wherein x is the input; s is down A down-sampling function is represented that,
the deviation in the scale is shown as a deviation,
the deviation in the horizontal direction is indicated,
represents the ith cell of the l-1 layer,
represents the weight from the ith cell to the lower J cell, I represents the number of samples, J represents the number of classification states, g ij Tag code representing class state j corresponding to sample i, E ij Representing the probability that sample i is predicted as classification state j.
The invention also provides a state monitoring device for the electric GIS equipment, which is used for the state monitoring method for the electric GIS equipment.
In this embodiment, the data sensing unit is configured to collect a vibration signal of the target device.
In this embodiment, the data acquisition unit is configured to process original vibration signal data, and includes collecting a vibration signal obtained by the data sensing unit, performing signal amplification, acquiring by an ADC, storing the signal, and transmitting the data to the data processing unit.
In this embodiment, the data processing unit is configured to process and calculate vibration data, and map the vibration signal to an image domain first, and map the vibration signal to the image domain by using an image fourier transform technique; then, encoding the one-dimensional image Fourier signal into a two-dimensional image matrix by means of an inverse trigonometric function; and finally, realizing intelligent identification of the state of the power GIS equipment by means of a convolutional neural network.
In this embodiment, the result display unit is configured to display the image-frequency signal encoded images of the GIS device in different states and output the state identification and classification accuracy.
The invention also provides a storage medium, which is applied to the state monitoring device for the power GIS equipment, wherein the computer equipment comprises a storage medium, the storage medium stores a computer program, and the computer program realizes the state monitoring method for the power GIS equipment when being executed by a data processing unit.
It should be noted that, in order to highlight the superiority of the method of the present invention, for a certain set of measured data, the set of data includes three types of data of a normal state of the GIS device, a contact loosening fault and a foreign matter fault inside the GIS, and the method of the present invention is adopted to calculate, and the obtained graph fourier signals and graph frequency signal coded images of three typical faults are as shown in fig. 5, it can be seen that when the method of the present invention is used to obtain two-dimensional images of three typical faults, the extreme values of the matrix are distributed only at the lower left corner of the two-dimensional image, which is consistent with the graph fourier signals; when the contact is in a loose state, the matrix extreme value is mainly concentrated in a 1/4 area on the left side of the two-dimensional image; when the internal foreign matter is in the state, the extremum region of the internal foreign matter is distributed in the whole two-dimensional image, and the Fourier signal of the image also has the characteristic of disordered distribution;
the performance curve during the training process is shown in fig. 6, where the curve in fig. 6 (a) represents the change of the classification accuracy, and the curve in fig. 6 (b) represents the corresponding loss function value, and the result shows that the loss function can quickly reach a stable value after 60 iterations. The corresponding classification accuracy rate is stable after iteration is carried out for 60 times, and the classification accuracy rate almost reaches 100%;
the confusion matrix is introduced to perform quantitative evaluation of the classification effect, and as shown in fig. 7, the calculation result of the confusion matrix obtained by the method is obtained. The diagonal lines of the confusion matrix in the figure represent the recognition accuracy for each state respectively using the proposed method. As can be seen from fig. 7, only 4 of the 300 test samples were misclassified, and the classification accuracy of the test sample set was 98.7%. As can be seen from table 1, the state recognition accuracy, recall rate and F1 value of the proposed method are 1, 0.98 and 0.98, respectively. The result shows that the method has good classification performance, and the effectiveness of the method is proved.
Table 1 evaluation parameter table
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. The state monitoring method for the power GIS equipment is characterized by comprising the following steps of:
s1: converting the GIS vibration signal acquired into an image signal;
s2: transforming the graph signal from the node domain to the graph frequency domain by a graph Fourier transform technique;
s3: encoding the one-dimensional image signal into a two-dimensional image matrix by using an inverse trigonometric function;
s4: and (3) realizing the state identification of the GlS equipment by utilizing convolutional neural network training and calculation.
2. State monitoring for electric GIS devices according to claim 1The measuring method is characterized in that the specific method of S1 is as follows: vibrating signal y = { y ] of GIS equipment 1 ,y 2 ,…y n Characterization of the map signal:
y→G＝{V,E}
wherein V = { V = { (V) 1 ,V 2 …V N Denotes a set of nodes, which can be used for the representation of different sample points, N denotes the number of vertices; e = { E = 1 ,E 2 …E M Denotes a set of connected edges, which can be used to describe the connections between different nodes, M denotes the number of connected edges,
further, a map signal vector X can be obtained:
X＝[X 1 ,X 2 …X N ] T
in the formula, X N Representing the vibration signal amplitude corresponding to the nth node, obviously, the order of the vertices determines the order of the map signal amplitudes,
further, the graph signal is transformed from the node domain to the graph frequency domain using a graph fourier transform:
in the formula, X GF (k) Representing a graph Fourier transform signal, k representing the number of nodes, x ki Denotes the kth graph fourier transform basis and "T" denotes transpose.
3. The state monitoring method for the power GIS device according to claim 1, characterized in that the specific method of S3 is as follows: the conversion of the one-dimensional graph Fourier transform signal and the two-dimensional matrix is realized by using an inverse trigonometric function:
in the formula, θ represents an angle obtained by an arcsine function, R ij The i-th row and j-th column values of the two-dimensional matrix are expressed, and the subscripts "i" and "j" are respectivelyRepresenting the ith and jth nodes.
4. The state monitoring method for the power GIS equipment according to claim 1, characterized in that the specific method of S4 is as follows: the method comprises the following steps of utilizing a VGG16 neural network to realize effective identification of two-dimensional images of graphs obtained by different GIS equipment states, reducing input parameter dimensionality through laminated convolution by the neural network, and further improving the performance of feature extraction, wherein the realization formulas of an activation function ReLU, a pooling step, a full connection step and a SoftMax classification function are respectively as follows:
wherein x is an input; s down A down-sampling function is represented that is,
the deviation in the scale is shown as a deviation,
the deviation in the horizontal direction is indicated,
denotes the ith cell of the l-1 st layer,
represents the weight from the ith cell to the lower J cell, I represents the number of samples, J represents the number of classification states, g ij Tag code representing class state j corresponding to sample i, E ij Representing the probability that sample i is predicted as classification state j.
5. The state monitoring electronic device for the power GIS equipment is used for the state monitoring method for the power GIS equipment according to any one of claims 1 to 5, and is characterized by comprising a data sensing unit, a data acquisition unit and computer equipment, wherein the computer equipment comprises a data processing unit and a result display unit.
6. The electronic device for monitoring the state of the power GIS equipment according to claim 5, wherein the data sensing unit is used for collecting vibration signals of target equipment.
7. The electronic device for monitoring the state of the power GIS equipment according to claim 5, wherein the data acquisition unit is used for processing raw vibration signal data, and comprises collecting vibration signals obtained by the data sensing unit, carrying out signal amplification, ADC acquisition and signal storage, and transmitting data to the data processing unit.
8. The condition monitoring electronic device for power GIS equipment as claimed in claim 5, characterized in that the data processing unit is used to process and calculate vibration data, firstly mapping the vibration signal to the map domain and mapping it to the map frequency domain by means of map Fourier transform technique; then, encoding the one-dimensional image Fourier signal into a two-dimensional image matrix by means of an inverse trigonometric function; and finally, realizing intelligent identification of the state of the power GIS equipment by means of a convolutional neural network.
9. The electronic device for monitoring the state of the power GIS equipment according to claim 5, wherein the result display unit is used for displaying the image frequency signal coding images of different states of the GIS equipment and the output of state identification and classification accuracy.
10. A storage medium applied to the state monitoring apparatus for a power GIS device according to any one of claims 5 to 9, wherein the computer device comprises a storage medium storing a computer program, and the computer program is executed by a data processing unit to implement the state monitoring method for the power GIS device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211341817.6A CN115684908A (en) | 2022-10-28 | 2022-10-28 | State monitoring method for power GIS equipment, storage medium and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211341817.6A CN115684908A (en) | 2022-10-28 | 2022-10-28 | State monitoring method for power GIS equipment, storage medium and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115684908A true CN115684908A (en) | 2023-02-03 |
Family
ID=85046776
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211341817.6A Pending CN115684908A (en) | 2022-10-28 | 2022-10-28 | State monitoring method for power GIS equipment, storage medium and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115684908A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115982577A (en) * | 2023-03-20 | 2023-04-18 | 山东华网合众信息技术有限公司 | Intelligent electricity consumption real-time monitoring method and system |
-
2022
- 2022-10-28 CN CN202211341817.6A patent/CN115684908A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115982577A (en) * | 2023-03-20 | 2023-04-18 | 山东华网合众信息技术有限公司 | Intelligent electricity consumption real-time monitoring method and system |
| CN115982577B (en) * | 2023-03-20 | 2023-09-08 | 山东华网合众信息技术有限公司 | Intelligent electricity utilization real-time monitoring method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109649432B (en) | System and method for monitoring integrity of steel rail of cloud platform based on guided wave technology | |
| CN115170000B (en) | Remote monitoring method and system based on electric energy meter communication module | |
| CN103267907A (en) | Method for identifying modal parameters of transformer coil | |
| CN110705456A (en) | Micro motor abnormity detection method based on transfer learning | |
| CN103439631A (en) | Method and system for detecting corrosion state of grounding grid | |
| CN105572501A (en) | Power quality disturbance identification method based on SST conversion and LS-SVM | |
| CN115684908A (en) | State monitoring method for power GIS equipment, storage medium and electronic device | |
| CN114414955A (en) | Method and system for monitoring and evaluating state of switch cabinet | |
| CN114330110A (en) | Model training method, photovoltaic power generation power prediction method and system | |
| CN105203937A (en) | Internal discharge mode recognition method and fault diagnosis system for transformer | |
| CN114492146B (en) | Bolt group loosening positioning and quantitative analysis method and system based on transfer learning | |
| CN114970610A (en) | Power transformer state identification method and device based on gram angular field enhancement | |
| CN112034312A (en) | Power equipment insulation defect mode identification method | |
| CN111638047A (en) | GIS equipment mechanical fault diagnosis system and method | |
| CN112560888B (en) | Method and system for evaluating short circuit resistance of transformer winding based on octyl geometric transformation | |
| CN115656926A (en) | Positioning identification method and device for monitoring abnormal sound of electrical equipment | |
| CN115310499A (en) | Industrial equipment fault diagnosis system and method based on data fusion | |
| CN112816122B (en) | Bolt tightness degree monitoring method based on deep learning and piezoelectric active sensing | |
| CN113156278A (en) | CNN network-based GIS partial discharge positioning method | |
| CN108761258A (en) | Transformer short period overload capability assessment system based on artificial intelligence and big data technology | |
| CN114838923A (en) | Fault diagnosis model establishing method and fault diagnosis method for on-load tap-changer | |
| CN112231624A (en) | Real-time evaluation system for short-circuit resistance of multi-transformer winding based on Internet of things | |
| CN110472801B (en) | Electromagnetic environment assessment method and system for direct-current transmission line | |
| CN117335912B (en) | Spectrum intelligent sensing method and system | |
| CN116660678A (en) | Evaluation system for power transmission network fault distance measurement and fault |
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 |