CN106778692A - A kind of cable local discharge signal recognition method and device based on S-transformation - Google Patents
A kind of cable local discharge signal recognition method and device based on S-transformation Download PDFInfo
- Publication number
- CN106778692A CN106778692A CN201710035996.3A CN201710035996A CN106778692A CN 106778692 A CN106778692 A CN 106778692A CN 201710035996 A CN201710035996 A CN 201710035996A CN 106778692 A CN106778692 A CN 106778692A
- Authority
- CN
- China
- Prior art keywords
- discharge signal
- local discharge
- matrix
- value sequence
- cable
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
- G06F2218/16—Classification; Matching by matching signal segments
- G06F2218/18—Classification; Matching by matching signal segments by plotting the signal segments against each other, e.g. analysing scattergrams
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Signal Processing (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The embodiment of the invention discloses a kind of cable local discharge signal recognition method and device based on S-transformation, for solving at present to the positioning in power cable partial discharge signal source generally using waveform time difference method, the technical problem that accuracy of identification is low and required recognition time is long.Present invention method includes:The local discharge signal in the known source to getting carries out S-transformation, obtains multiple time-frequency matrix;Modulus are carried out to multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain the unusual value sequence of modular matrix;Unusual value sequence is divided at least two interval, the Shannon entropys of singular value in each interval and the ratio of the Shannon entropys of unusual value sequence are calculated, and is set up local discharge signal feature samples storehouse using ratio as local discharge signal characteristic vector and is built supporting vector machine model;The characteristic vector of local discharge signal to be identified is input to supporting vector machine model, the source of local discharge signal to be identified is obtained.
Description
Technical field
The present invention relates to cable local discharge on-line monitoring technique field, more particularly to a kind of cable office based on S-transformation
Portion's discharge signal recognition methods and device.
Background technology
With the high speed sustainable development of power system, the loop length of laying of power cable is steadily improved, in city
Extensive use is arrived.However, electric load is growing with voltage class, by Insulation Problems caused by cable local defect to supplying
Electricity quality and social economy etc. constitute greatly threat.In order to monitor the state of insulation of cable and find that it locally lacks in time
Fall into, so as to prevent the generation of cable interruption of service, it is ensured that the reliability of operation of power networks, it is necessary to which cable local defect is examined
Survey.
In cable local discharge on-line monitoring, the local discharge signal for detecting may be from cable body and cable termination
Head, it is also possible to from coupled switch cubicle.Because the shelf depreciation of separate sources endangers different to equipment, criterion
Difference, so being identified important realistic meaning to local discharge signal source.
In terms of local discharge signal identification, signal characteristic abstraction and grader selection are most critical parts.Feature extraction
It is the local discharge signal identification first step, the quality of feature extraction directly influences the effect of identification.At present, local discharge signal
Feature extracting method mainly has statistical nature method and the major class of temporal analysis two.Wherein statistical nature method is directed to shelf depreciation
The phase of signal, and distribution cable is generally three-core cable and totally one ground wire, when there is shelf depreciation in two-phase or three-phase, detection
The phase property of local discharge signal becomes hardly possible.Temporal analysis be directed to high speed acquisition once discharge generation when
Wave character or corresponding transformation results obtained by the pulse of domain carry out pattern-recognition, mainly including Fourier analysis method, small echo
Analytic approach and waveform parameter direct extraction method etc..Pattern recognition classifier device mainly has neural network classifier, minimum distance classification
Device and fuzzy diagnosis grader.Neutral net easily converges on locally optimal solution defect, and precision is not high.
At present, whole power industry generally uses waveform time difference side to the positioning in power cable partial discharge signal source
Method, through be commonly present for transmit the power cable line of the local discharge signal it is long when, accuracy of identification is substantially reduced and required
Recognition time is long.
The content of the invention
A kind of cable local discharge signal recognition method and device based on S-transformation are the embodiment of the invention provides, is solved
Current power industry generally uses waveform time difference method to the positioning in power cable partial discharge signal source, through being commonly present use
When the power cable line for transmitting the local discharge signal is long, accuracy of identification is substantially reduced and required recognition time is long
Technical problem.
A kind of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention, including:
The local discharge signal in known source is obtained, and S-transformation is carried out to local discharge signal, obtain multiple time-frequency matrix;
Modulus are carried out to multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain modular matrix
Unusual value sequence;
Unusual value sequence is divided into according to unusual value sequence at least two interval, calculates singular value in each interval
The ratio of Shannon entropys and the Shannon entropys of unusual value sequence, and using ratio as local discharge signal characteristic vector foundation office
Portion discharge signal feature samples storehouse;
Using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as input, many classification branch are built
Hold vector machine model;
Using local discharge signal as sample, supporting vector machine model is trained, the supporting vector for being trained
Machine model;
The characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training, obtains waiting to know
The source of other local discharge signal.
Alternatively, local discharge signal includes cable body local discharge signal, cable terminal local discharge signal, opens
Close the corona discharge signal of cabinet and the surface-discharge signal of switch cubicle.
Alternatively, the local discharge signal in known source is obtained, and S-transformation is carried out to local discharge signal, when obtaining multiple
Frequency matrix includes:
The local discharge signal in known source is obtained, and S-transformation is carried out by a pair of local discharge signals of preset formula, obtained
To multiple time-frequency matrix, preset formula one is specially:
Wherein, h (kT) is local discharge signal discrete-time series, and ST passes through for local discharge signal discrete-time series
The multiple time-frequency matrix obtained after S-transformation, T for discrete-time series sampling period, N for discrete-time series length, H be from
The Fourier transform of time series is dissipated, j is imaginary unit, k, n, m=0,1 ..., N-1.
Alternatively, modulus are carried out to multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain mould
The unusual value sequence of matrix includes:
Modulus are carried out to multiple time-frequency matrix ST and obtains modular matrix STA, and modular matrix STA is carried out by preset second formula
Singular value decomposition, obtains the unusual value sequence of modular matrix, and preset second formula is specially:
Wherein, U and V are N × N rank orthogonal matrixes, D=diag (σ1,σ2,…,σN) it is diagonal matrix, its diagonal element
(σ1,σ2,…,σN) it is the singular value of matrix STA, uiAnd viRespectively the i-th row singular value vector of matrix U and V.
Alternatively, unusual value sequence is divided into by least two intervals according to unusual value sequence, is calculated in each interval
The ratio of the Shannon entropys of singular value and the Shannon entropys of unusual value sequence, and using ratio as local discharge signal feature to
Amount sets up local discharge signal feature samples storehouse to be included:
Unusual value sequence is divided into according to unusual value sequence at least two interval, calculates singular value in each interval
The ratio of Shannon entropys and the Shannon entropys of unusual value sequence, and believed as shelf depreciation using ratio by preset 3rd formula
Number characteristic vector sets up local discharge signal feature samples storehouse, and preset 3rd formula is specially:
λ=[E1/E,E2/E,Eq/E…,EQ/E];
Wherein,It is unusual value sequence(σ1, σ2..., σN) Shannon entropys,
For singular value obtains Shannon entropys, σ in the q of minizoneiQ () is the singular value in the q of minizone, Q is minizone number.
Alternatively, the source categories of local discharge signal characteristic vector are marked using two bits.
Alternatively, using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as input, build
Multi-category support vector machines model includes:
With two sorting algorithms by the source of the local discharge signal characteristic vector in local discharge signal feature samples storehouse
Each source categories in classification are combined, and constitute a plurality of sub-classifiers, and with local discharge signal feature samples storehouse
Local discharge signal characteristic vector as input, build multi-category support vector machines model.
Alternatively, using local discharge signal as sample, supporting vector machine model is trained, the branch for being trained
Holding vector machine model includes:
The local discharge signal that separate sources according to local discharge signal selects the separate sources of equal number at random is made
It is training sample, and training sample is input into supporting vector machine model to be trained, the SVMs mould for being trained
Type.
Alternatively, the characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training,
The source for obtaining local discharge signal to be identified includes:
The characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training to be exported
Value, and output valve is contrasted with the binary number mark of the source categories of local discharge signal characteristic vector, obtain waiting to know
The source of other local discharge signal.
A kind of cable local discharge signal identification device based on S-transformation provided in an embodiment of the present invention, it is characterised in that
Including:
Conversion module, for obtaining the local discharge signal in known source, and carries out S-transformation to local discharge signal, obtains
To multiple time-frequency matrix;
Decomposing module, modular matrix is obtained for carrying out modulus to multiple time-frequency matrix, and carries out singular value decomposition to modular matrix,
Obtain the unusual value sequence of modular matrix;
Computing module, for unusual value sequence to be divided into at least two intervals according to unusual value sequence, calculates each area
The ratio of the Shannon entropys of interior singular value and the Shannon entropys of unusual value sequence, and using ratio as local discharge signal
Characteristic vector sets up local discharge signal feature samples storehouse;
Module is built, for using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as defeated
Enter, build multi-category support vector machines model;
Training module, for using local discharge signal as sample, being trained to supporting vector machine model, is trained
Good supporting vector machine model;
Input module, for the characteristic vector of local discharge signal to be identified to be input into the SVMs for training
Model, obtains the source of local discharge signal to be identified.
As can be seen from the above technical solutions, the embodiment of the present invention has advantages below:
A kind of cable local discharge signal recognition method and device based on S-transformation are the embodiment of the invention provides, including:
The local discharge signal in known source is obtained, and S-transformation is carried out to local discharge signal, obtain multiple time-frequency matrix;To multiple time-frequency
Matrix carries out modulus and obtains modular matrix, and carries out singular value decomposition to modular matrix, obtains the unusual value sequence of modular matrix;According to strange
Different value sequence unusual value sequence is divided into it is at least two interval, calculate the Shannon entropys of singular value in each interval with it is strange
The ratio of the Shannon entropys of different value sequence, and it is special to set up local discharge signal using ratio as local discharge signal characteristic vector
Levy Sample Storehouse;Using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as input, many classification are built
Supporting vector machine model;Using local discharge signal as sample, supporting vector machine model is trained, the branch for being trained
Hold vector machine model;The characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training, is obtained
To the source of local discharge signal to be identified, in the embodiment of the present invention after carrying out S-transformation to local discharge signal, obtain
Multiple time-frequency matrix;Modulus are carried out to multiple time-frequency matrix again and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain mould square
The unusual value sequence of battle array;Unusual value sequence is divided into by least two intervals according to unusual value sequence, is calculated in each interval
The ratio of the Shannon entropys of singular value and the Shannon entropys of unusual value sequence, and using ratio as local discharge signal feature to
Amount sets up local discharge signal feature samples storehouse;With the local discharge signal characteristic vector in local discharge signal feature samples storehouse
As input, multi-category support vector machines model is built, as long as final that the characteristic vector of local discharge signal to be identified is defeated
Enter to the supporting vector machine model for training, you can the source of local discharge signal to be identified is got, with identification step
Simply, the characteristics of accuracy of identification is high, recognition speed is fast, solves current power industry to power cable partial discharge signal source
Positioning generally uses waveform time difference method, long for transmitting the power cable line of the local discharge signal through being commonly present
When, accuracy of identification is substantially reduced and the long technical problem of required recognition time.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
The accompanying drawing to be used needed for having technology description is briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, without having to pay creative labor, may be used also
Other accompanying drawings are obtained with according to these accompanying drawings.
Fig. 1 is of a kind of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
The schematic flow sheet of embodiment;
Fig. 2 is cable body local discharge signal oscillogram provided in an embodiment of the present invention;
Fig. 3 is cable terminal local discharge signal oscillogram provided in an embodiment of the present invention;
Fig. 4 is the corona discharge signal of switch cubicle provided in an embodiment of the present invention;
Fig. 5 is the surface-discharge signal waveforms of switch cubicle provided in an embodiment of the present invention;
Fig. 6 is a kind of the another of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
The schematic flow sheet of individual embodiment;
Fig. 7 is a kind of structure of cable local discharge signal identification device based on S-transformation provided in an embodiment of the present invention
Schematic diagram.
Specific embodiment
A kind of cable local discharge signal recognition method and device based on S-transformation are the embodiment of the invention provides, is used for
Current power industry is solved to the positioning in power cable partial discharge signal source generally using waveform time difference method, through being commonly present
When power cable line for transmitting the local discharge signal is long, accuracy of identification is substantially reduced and required recognition time is long
Technical problem.
To enable that goal of the invention of the invention, feature, advantage are more obvious and understandable, below in conjunction with the present invention
Accompanying drawing in embodiment, is clearly and completely described, it is clear that disclosed below to the technical scheme in the embodiment of the present invention
Embodiment be only a part of embodiment of the invention, and not all embodiment.Based on the embodiment in the present invention, this area
All other embodiment that those of ordinary skill is obtained under the premise of creative work is not made, belongs to protection of the present invention
Scope.
Fig. 1 is referred to, a kind of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
One embodiment includes:
101st, the local discharge signal in known source is obtained, and S-transformation is carried out to local discharge signal, obtain multiple time-frequency square
Battle array;
First, the local discharge signal of cable is got, wherein the office of separate sources can be got by corresponding experiment
Portion's discharge signal, the local discharge signal of separate sources includes cable body local discharge signal, cable terminal shelf depreciation
Surface-discharge signal of signal, the corona discharge signal of switch cubicle and switch cubicle etc., as shown in Fig. 2 for cable body is locally put
Electric signal waveform figure;As shown in figure 3, being cable terminal local discharge signal oscillogram;As shown in figure 4, being the electricity of switch cubicle
Corona signal;As shown in figure 5, being the surface-discharge signal waveforms of switch cubicle.Local discharge signal to getting carries out S
Conversion, you can obtain corresponding multiple time-frequency matrix.
102nd, modulus are carried out to multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain mould square
The unusual value sequence of battle array;
In the local discharge signal in the known source of acquisition, and S-transformation is carried out to local discharge signal, obtain multiple time-frequency matrix
Afterwards, modulus are carried out to multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to modular matrix, obtain the unusual of modular matrix
Value sequence.
103rd, unusual value sequence is divided into by least two intervals according to unusual value sequence, calculates unusual in each interval
The ratio of the Shannon entropys of value and the Shannon entropys of unusual value sequence, and built as local discharge signal characteristic vector using ratio
Vertical local discharge signal feature samples storehouse;
Modular matrix is obtained modulus are carried out to multiple time-frequency matrix, and singular value decomposition is carried out to modular matrix, obtain modular matrix
Unusual value sequence after, unusual value sequence is divided into according to unusual value sequence it is at least two interval, such as by unusual value sequence
10 minizones are divided into, the Shannon entropys of singular value in each minizone and the Shannon of whole unusual value sequence is calculated
The ratio of entropy, and local discharge signal feature samples storehouse is set up as local discharge signal characteristic vector using ratio.
It should be noted that when local discharge signal feature samples storehouse is set up, can be marked using two bits
The source categories of local discharge signal characteristic vector, that is, mark the separate sources of local discharge signal, such as cable local discharge letter
Number, the surface-discharge signal in corona discharge signal and switch cubicle in cable terminal local discharge signal, switch cubicle can divide
Not Biao Ji for:(+1 ,+1), (+1, -1), (- 1 ,+1), (- 1, -1).
104th, using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as input, many points are built
Class support vector machines model;
Unusual value sequence is being divided into by least two intervals according to unusual value sequence, the singular value in each interval is being calculated
Shannon entropys and the Shannon entropys of unusual value sequence ratio, and set up as local discharge signal characteristic vector using ratio
After local discharge signal feature samples storehouse, made with the local discharge signal characteristic vector in local discharge signal feature samples storehouse
It is input, builds multi-category support vector machines model.
105th, using local discharge signal as sample, supporting vector machine model is trained, the support for being trained
Vector machine model;
After multi-category support vector machines model construction is finished, by getting the local discharge signal of separate sources and inciting somebody to action
The local discharge signal of the separate sources for getting is trained as sample in input to supporting vector machine model, can obtain
To the supporting vector machine model for training.
106th, the characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training, is obtained
The source of local discharge signal to be identified.
Finally, when needing to be identified the source of the local discharge signal of cable, shelf depreciation to be identified is believed
Number the characteristic vector for obtaining local discharge signal is calculated, and the characteristic vector of local discharge signal is input to the support for training
Vector machine model, the output valve according to supporting vector machine model is the source of the local discharge signal that can obtain to be identified.
It is more than to a kind of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
The detailed description of individual embodiment, below by a kind of cable local discharge signal based on S-transformation provided in an embodiment of the present invention
Another embodiment of recognition methods is described in detail.
Fig. 6 is referred to, a kind of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
Another embodiment includes:
201st, the local discharge signal in known source is obtained, and S changes is carried out by a pair of local discharge signals of preset formula
Change, obtain multiple time-frequency matrix;
First, the local discharge signal of cable is got, wherein the office of separate sources can be got by corresponding experiment
Portion's discharge signal, the local discharge signal of separate sources includes cable body local discharge signal, cable terminal shelf depreciation
Surface-discharge signal of signal, the corona discharge signal of switch cubicle and switch cubicle etc., as shown in Fig. 2 for cable body is locally put
Electric signal waveform figure;As shown in figure 3, being cable terminal local discharge signal oscillogram;As shown in figure 4, being the electricity of switch cubicle
Corona signal;As shown in figure 5, being the surface-discharge signal waveforms of switch cubicle.Local discharge signal to getting passes through
Preset formula one carries out S-transformation, you can obtain corresponding multiple time-frequency matrix, and preset formula one is specially:
Wherein, h (kT) is local discharge signal discrete-time series, and ST passes through for local discharge signal discrete-time series
The multiple time-frequency matrix obtained after S-transformation, T for discrete-time series sampling period, N for discrete-time series length, H be from
The Fourier transform of time series is dissipated, j is imaginary unit, k, n, m=0,1 ..., N-1.
202nd, modulus are carried out to multiple time-frequency matrix ST and obtains modular matrix STA, and by preset second formula to modular matrix STA
Singular value decomposition is carried out, the unusual value sequence of modular matrix is obtained;
In the local discharge signal in the known source of acquisition, and S-transformation is carried out by a pair of local discharge signals of preset formula,
Obtain after multiple time-frequency matrix, modulus are carried out to multiple time-frequency matrix ST and obtains modular matrix STA, and by preset second formula to mould
Matrix STA carries out singular value decomposition, obtains the unusual value sequence of modular matrix, and preset second formula is specially:
Wherein, U and V are N × N rank orthogonal matrixes, D=diag (σ1,σ2,…,σN) it is diagonal matrix, its diagonal element
(σ1,σ2,…,σN) it is the singular value of matrix STA, uiAnd viRespectively the i-th row singular value vector of matrix U and V.
203rd, unusual value sequence is divided into by least two intervals according to unusual value sequence, calculates unusual in each interval
The ratio of the Shannon entropys of value and the Shannon entropys of unusual value sequence, and put as local using ratio by preset 3rd formula
Signal characteristics vector sets up local discharge signal feature samples storehouse;
Modular matrix STA is obtained modulus are carried out to multiple time-frequency matrix ST, and modular matrix STA is entered by preset second formula
Row singular value decomposition, obtains after the unusual value sequence of modular matrix, is divided at least unusual value sequence according to unusual value sequence
Two intervals, calculate the Shannon entropys of singular value in each interval and the ratio of the Shannon entropys of unusual value sequence, and lead to
Cross preset 3rd formula and local discharge signal feature samples storehouse, preset are set up as local discharge signal characteristic vector using ratio
Three formula are specially:
λ=[E1/E,E2/E,Eq/E…,EQ/E];
Wherein,It is unusual value sequence(σ1, σ2..., σN) Shannon entropys,
For singular value obtains Shannon entropys, σ in the q of minizoneiQ () is the singular value in the q of minizone, Q is minizone number.
It should be noted that when local discharge signal feature samples storehouse is set up, can be marked using two bits
The source categories of local discharge signal characteristic vector, that is, mark the separate sources of local discharge signal, such as cable local discharge letter
Number, the surface-discharge signal in corona discharge signal and switch cubicle in cable terminal local discharge signal, switch cubicle can divide
Not Biao Ji for:(+1 ,+1), (+1, -1), (- 1 ,+1), (- 1, -1).
204th, with two sorting algorithms by the local discharge signal characteristic vector in local discharge signal feature samples storehouse
Each source categories in source categories are combined, and constitute a plurality of sub-classifiers, and with local discharge signal feature samples
Local discharge signal characteristic vector in storehouse builds multi-category support vector machines model as input;
Unusual value sequence is being divided into by least two intervals according to unusual value sequence, the singular value in each interval is being calculated
Shannon entropys and the Shannon entropys of unusual value sequence ratio, and by preset 3rd formula using ratio as shelf depreciation
Signal characteristic vector is set up after local discharge signal feature samples storehouse, with two sorting algorithms by local discharge signal feature sample
Each source categories in the source categories of the local discharge signal characteristic vector of Ben Kunei are combined, and constitute plural height point
Class device, and using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as input, build many classification branch
Hold vector machine model.Because common supporting vector machine model is two category support vector machines models, and the part of required identification
Discharge signal source categories have four kinds, it is therefore desirable to by using two sorting algorithms by local discharge signal feature samples storehouse
Each source categories in the source categories of local discharge signal characteristic vector are combined, and may be constructed 2 sub-classifiers
SVM1 and SVM2, many classification are expanded to by the classification of SVMs two.
If with A, B, C, D represent cable body local discharge signal, cable terminal local discharge signal, open respectively
The 4 kinds of discharge signals of surface-discharge signal in the corona discharge signal and switch cubicle in cabinet are closed, then SVM1 and SVM2 output results
It is as shown in table 1 with discharge signal type corresponding relation.
Table 1
205th, the separate sources according to local discharge signal selects the shelf depreciation letter of the separate sources of equal number at random
Number as training sample, and training sample is input to supporting vector machine model it is trained, the supporting vector for being trained
Machine model;
After multi-category support vector machines model construction is finished, chosen at random by the separate sources according to local discharge signal
Select the local discharge signal of separate sources of equal number as training sample, i.e., selected at random from four kinds of local discharge signals
The local discharge signal of equal number, such as 50 cable body local discharge signals, 50 cable terminal local discharge signals,
The surface-discharge signal of 50 corona discharge signals of switch cubicle, 50 switch cubicles, and training sample is input to supporting vector
Machine model is trained, the supporting vector machine model for being trained.
206th, the characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training to obtain
Output valve, and output valve is contrasted with the binary number mark of the source categories of local discharge signal characteristic vector, obtain
The source of local discharge signal to be identified.
Finally, when needing to be identified the source of the local discharge signal of cable, shelf depreciation to be identified is believed
Number the characteristic vector for obtaining local discharge signal is calculated, and the characteristic vector of local discharge signal to be identified is input to training
Good supporting vector machine model obtains output valve, and output valve is entered with the two of the source categories of local discharge signal characteristic vector
Number mark processed is contrasted, and obtains the source of local discharge signal to be identified, i.e., as shown in table 1, output valve is (+1 ,+1),
(+1, -1), (- 1 ,+1), (- 1, -1) corresponds to cable local discharge signal, cable terminal local discharge signal, switch cubicle respectively
In corona discharge signal, switch cubicle in surface-discharge signal.
It is more than to a kind of the another of cable local discharge signal recognition method based on S-transformation provided in an embodiment of the present invention
The detailed description of one embodiment, below by a kind of cable local discharge letter based on S-transformation provided in an embodiment of the present invention
Number identifying device is described in detail.
Refer to Fig. 7, a kind of cable local discharge signal identification device bag based on S-transformation provided in an embodiment of the present invention
Include:
Conversion module 301, for obtaining the local discharge signal in known source, and carries out S-transformation to local discharge signal,
Obtain multiple time-frequency matrix;
Decomposing module 302, modular matrix is obtained for carrying out modulus to multiple time-frequency matrix, and singular value point is carried out to modular matrix
Solution, obtains the unusual value sequence of modular matrix;
Computing module 303, for unusual value sequence to be divided into at least two intervals according to unusual value sequence, calculates each
The ratio of the Shannon entropys of the singular value in interval and the Shannon entropys of unusual value sequence, and believed as shelf depreciation using ratio
Number characteristic vector sets up local discharge signal feature samples storehouse;
Build module 304, for using the local discharge signal characteristic vector in local discharge signal feature samples storehouse as
Input, builds multi-category support vector machines model;
Training module 305, for using local discharge signal as sample, being trained to supporting vector machine model, obtains
The supporting vector machine model for training;
Input module 306, for by the characteristic vector of local discharge signal to be identified be input to the support for training to
Amount machine model, obtains the source of local discharge signal to be identified.
It is apparent to those skilled in the art that, for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, may be referred to the corresponding process in preceding method embodiment, will not be repeated here.
In several embodiments provided herein, it should be understood that disclosed system, apparatus and method can be with
Realize by another way.For example, device embodiment described above is only schematical, for example, the unit
Divide, only a kind of division of logic function there can be other dividing mode when actually realizing, for example multiple units or component
Can combine or be desirably integrated into another system, or some features can be ignored, or do not perform.It is another, it is shown or
The coupling each other for discussing or direct-coupling or communication connection can be the indirect couplings of device or unit by some interfaces
Close or communicate to connect, can be electrical, mechanical or other forms.
The unit that is illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit
The part for showing can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can be according to the actual needs selected to realize the mesh of this embodiment scheme
's.
In addition, during each functional unit in each embodiment of the invention can be integrated in a processing unit, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a unit.Above-mentioned integrated list
Unit can both be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If the integrated unit is to realize in the form of SFU software functional unit and as independent production marketing or use
When, can store in a computer read/write memory medium.Based on such understanding, technical scheme is substantially
The part for being contributed to prior art in other words or all or part of the technical scheme can be in the form of software products
Embody, the computer software product is stored in a storage medium, including some instructions are used to so that a computer
Equipment (can be personal computer, server, or network equipment etc.) performs the complete of each embodiment methods described of the invention
Portion or part steps.And foregoing storage medium includes:USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can store journey
The medium of sequence code.
The above, the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although with reference to preceding
Embodiment is stated to be described in detail the present invention, it will be understood by those within the art that:It still can be to preceding
State the technical scheme described in each embodiment to modify, or equivalent is carried out to which part technical characteristic;And these
Modification is replaced, and does not make the spirit and scope of the essence disengaging various embodiments of the present invention technical scheme of appropriate technical solution.
Claims (10)
1. a kind of cable local discharge signal recognition method based on S-transformation, it is characterised in that including:
The local discharge signal in known source is obtained, and S-transformation is carried out to the local discharge signal, obtain multiple time-frequency matrix;
Modulus are carried out to the multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to the modular matrix, obtain described
The unusual value sequence of modular matrix;
The unusual value sequence is divided into by least two intervals according to the unusual value sequence, calculates unusual in each interval
The ratio of the Shannon entropys of value and the Shannon entropys of the unusual value sequence, and it is special as local discharge signal using the ratio
Levy vector and set up local discharge signal feature samples storehouse;
Using the local discharge signal characteristic vector in the local discharge signal feature samples storehouse as input, many classification branch are built
Hold vector machine model;
Using the local discharge signal as sample, the supporting vector machine model is trained, the support for being trained
Vector machine model;
The characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training, obtains waiting to know
The source of other local discharge signal.
2. the cable local discharge signal recognition method based on S-transformation according to claim 1, it is characterised in that described
Local discharge signal includes cable body local discharge signal, cable terminal local discharge signal, the corona discharge of switch cubicle
The surface-discharge signal of signal and switch cubicle.
3. the cable local discharge signal recognition method based on S-transformation according to claim 1, it is characterised in that described
The local discharge signal in known source is obtained, and S-transformation is carried out to the local discharge signal, obtaining multiple time-frequency matrix includes:
Obtain the local discharge signal in known source, and S-transformation is carried out by local discharge signal described in a pair of preset formula, obtain
To multiple time-frequency matrix, the preset formula one is specially:
Wherein, h (kT) is local discharge signal discrete-time series, and ST is that local discharge signal discrete-time series becomes by S
The multiple time-frequency matrix obtained after changing, T is the sampling period of discrete-time series, and N is the length of discrete-time series, and H is discrete
The Fourier transform of time series, j is imaginary unit, k, n, m=0,1 ..., N-1.
4. the cable local discharge signal recognition method based on S-transformation according to claim 3, it is characterised in that described
Modulus are carried out to the multiple time-frequency matrix and obtains modular matrix, and singular value decomposition is carried out to the modular matrix, obtain the mould square
The unusual value sequence of battle array includes:
Modulus are carried out to the multiple time-frequency matrix ST and obtains modular matrix STA, and by preset second formula to the modular matrix STA
Singular value decomposition is carried out, the unusual value sequence of the modular matrix is obtained, preset second formula is specially:
Wherein, U and V are N × N rank orthogonal matrixes, D=diag (σ1,σ2,…,σN) it is diagonal matrix, its diagonal element (σ1,
σ2,…,σN) it is the singular value of matrix STA, uiAnd viRespectively the i-th row singular value vector of matrix U and V.
5. the cable local discharge signal recognition method based on S-transformation according to claim 4, it is characterised in that described
The unusual value sequence is divided into according to the unusual value sequence at least two interval, calculates singular value in each interval
The ratio of Shannon entropys and the Shannon entropys of the unusual value sequence, and using the ratio as local discharge signal feature to
Amount sets up local discharge signal feature samples storehouse to be included:
The unusual value sequence is divided into by least two intervals according to the unusual value sequence, calculates unusual in each interval
The ratio of the Shannon entropys of value and the Shannon entropys of the unusual value sequence, and made with the ratio by preset 3rd formula
For local discharge signal characteristic vector sets up local discharge signal feature samples storehouse, preset 3rd formula is specially:
λ=[E1/E,E2/E,Eq/E…,EQ/E];
Singular value, Q be minizone number.
6. the cable local discharge signal recognition method based on S-transformation according to claim 5, it is characterised in that use
Two bits mark the source categories of the local discharge signal characteristic vector.
7. the cable local discharge signal recognition method based on S-transformation according to claim 6, it is characterised in that described
Using the local discharge signal characteristic vector in the local discharge signal feature samples storehouse as input, build many classification support to
Amount machine model includes:
With two sorting algorithms by the source of the local discharge signal characteristic vector in the local discharge signal feature samples storehouse
Each source categories in classification are combined, and constitute a plurality of sub-classifiers, and with the local discharge signal feature samples
Local discharge signal characteristic vector in storehouse builds multi-category support vector machines model as input.
8. the cable local discharge signal recognition method based on S-transformation according to claim 7, it is characterised in that described
Using the local discharge signal as sample, the supporting vector machine model is trained, the supporting vector for being trained
Machine model includes:
The local discharge signal that separate sources according to the local discharge signal selects the separate sources of equal number at random is made
It is training sample, and the training sample is input into the supporting vector machine model to be trained, the support for being trained
Vector machine model.
9. the cable local discharge signal recognition method based on S-transformation according to claim 8, it is characterised in that described
The characteristic vector of local discharge signal to be identified is input to the supporting vector machine model for training, obtains to be identified
The source of local discharge signal includes:
The characteristic vector of local discharge signal to be identified is input into the supporting vector machine model for training to be exported
Value, and the output valve is contrasted with the binary number mark of the source categories of the local discharge signal characteristic vector,
Obtain the source of local discharge signal to be identified.
10. a kind of cable local discharge signal identification device based on S-transformation, it is characterised in that including:
Conversion module, for obtaining the local discharge signal in known source, and carries out S-transformation to the local discharge signal, obtains
To multiple time-frequency matrix;
Decomposing module, obtains modular matrix, and carry out singular value to the modular matrix for carrying out modulus to the multiple time-frequency matrix
Decompose, obtain the unusual value sequence of the modular matrix;
Computing module, for the unusual value sequence to be divided into at least two intervals according to the unusual value sequence, calculates every
It is individual interval in singular value Shannon entropys and the Shannon entropys of the unusual value sequence ratio, and using the ratio as
Local discharge signal characteristic vector sets up local discharge signal feature samples storehouse;
Module is built, for using the local discharge signal characteristic vector in the local discharge signal feature samples storehouse as defeated
Enter, build multi-category support vector machines model;
Training module, for using the local discharge signal as sample, being trained to the supporting vector machine model, obtains
The supporting vector machine model for training;
Input module, for the characteristic vector of local discharge signal to be identified to be input into the SVMs for training
Model, obtains the source of local discharge signal to be identified.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710035996.3A CN106778692B (en) | 2017-01-17 | 2017-01-17 | Cable partial discharge signal identification method and device based on S transformation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710035996.3A CN106778692B (en) | 2017-01-17 | 2017-01-17 | Cable partial discharge signal identification method and device based on S transformation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106778692A true CN106778692A (en) | 2017-05-31 |
CN106778692B CN106778692B (en) | 2020-07-03 |
Family
ID=58944005
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710035996.3A Active CN106778692B (en) | 2017-01-17 | 2017-01-17 | Cable partial discharge signal identification method and device based on S transformation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106778692B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107255779A (en) * | 2017-06-15 | 2017-10-17 | 国网河南省电力公司济源供电公司 | A kind of high-tension switch cabinet Analysis of Partial Discharge method |
CN107862320A (en) * | 2017-11-28 | 2018-03-30 | 广东电网有限责任公司珠海供电局 | A kind of porcelain shell for cable terminal Infrared Image Features vector extracting method |
CN114325270A (en) * | 2021-12-30 | 2022-04-12 | 广东电网有限责任公司江门供电局 | Switch cabinet partial discharge monitoring method and device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103077402A (en) * | 2012-12-28 | 2013-05-01 | 湖北省电力公司电力科学研究院 | Transformer partial-discharging mode recognition method based on singular value decomposition algorithm |
CN103076547A (en) * | 2013-01-24 | 2013-05-01 | 安徽省电力公司亳州供电公司 | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines |
CN103091612A (en) * | 2013-01-16 | 2013-05-08 | 重庆市电力公司电力科学研究院 | Separation and recognition algorithm for transformer oiled paper insulation multiple partial discharging source signals |
CN106250904A (en) * | 2016-05-18 | 2016-12-21 | 国网新疆电力公司电力科学研究院 | Based on Power Disturbance analyser and the sorting technique of improving S-transformation |
-
2017
- 2017-01-17 CN CN201710035996.3A patent/CN106778692B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103077402A (en) * | 2012-12-28 | 2013-05-01 | 湖北省电力公司电力科学研究院 | Transformer partial-discharging mode recognition method based on singular value decomposition algorithm |
CN103091612A (en) * | 2013-01-16 | 2013-05-08 | 重庆市电力公司电力科学研究院 | Separation and recognition algorithm for transformer oiled paper insulation multiple partial discharging source signals |
CN103076547A (en) * | 2013-01-24 | 2013-05-01 | 安徽省电力公司亳州供电公司 | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines |
CN106250904A (en) * | 2016-05-18 | 2016-12-21 | 国网新疆电力公司电力科学研究院 | Based on Power Disturbance analyser and the sorting technique of improving S-transformation |
Non-Patent Citations (3)
Title |
---|
DANGDANG DAI等: "《Feature extraction of GIS partial discharge signal based on S-transform and singular value decomposition》", 《IET DIGITAL LIBRARY》 * |
张宇辉等: "《频率切片小波变换在局部放电信号分析中的应用》", 《高电压技术》 * |
郭灿新等: "《S变换在电力电缆局部放电信号时频分析中的应用》", 《电工技术学报》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107255779A (en) * | 2017-06-15 | 2017-10-17 | 国网河南省电力公司济源供电公司 | A kind of high-tension switch cabinet Analysis of Partial Discharge method |
CN107862320A (en) * | 2017-11-28 | 2018-03-30 | 广东电网有限责任公司珠海供电局 | A kind of porcelain shell for cable terminal Infrared Image Features vector extracting method |
CN114325270A (en) * | 2021-12-30 | 2022-04-12 | 广东电网有限责任公司江门供电局 | Switch cabinet partial discharge monitoring method and device |
CN114325270B (en) * | 2021-12-30 | 2023-07-04 | 广东电网有限责任公司江门供电局 | Switch cabinet partial discharge monitoring method and device |
Also Published As
Publication number | Publication date |
---|---|
CN106778692B (en) | 2020-07-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ekici et al. | Energy and entropy-based feature extraction for locating fault on transmission lines by using neural network and wavelet packet decomposition | |
CN109635928B (en) | Voltage sag reason identification method based on deep learning model fusion | |
Peng et al. | Random forest based optimal feature selection for partial discharge pattern recognition in HV cables | |
CN106443380B (en) | A kind of distribution cable local discharge signal recognition methods and device | |
CN103076547B (en) | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines | |
CN112041693A (en) | Power distribution network fault positioning system based on mixed wave recording | |
CN102508076B (en) | Fault diagnosis device and method based on multi-agent system and wavelet analysis | |
CN109977920A (en) | Fault Diagnosis of Hydro-generator Set method based on time-frequency spectrum and convolutional neural networks | |
CN109933881A (en) | A kind of Fault Diagnosis of Power Electronic Circuits method based on optimization deepness belief network | |
CN106599777A (en) | Cable partial discharge signal identification method based on energy percentage | |
CN101975910A (en) | Intelligent fault classification and location method for ultra-high voltage direct current transmission line | |
CN106778923A (en) | A kind of Power Quality Disturbance sorting technique and device | |
CN106778692A (en) | A kind of cable local discharge signal recognition method and device based on S-transformation | |
CN102279358A (en) | MCSKPCA based neural network fault diagnosis method for analog circuits | |
CN108154223B (en) | Power distribution network working condition wave recording classification method based on network topology and long time sequence information | |
CN102005755B (en) | Extra-high voltage direct current transmission line boundary element forming method based on support vector machine | |
CN110672905A (en) | CNN-based self-supervision voltage sag source identification method | |
CN103440497B (en) | A kind of GIS insulation defect shelf depreciation collection of illustrative plates mode identification method | |
CN106203531A (en) | A kind of transmission line fault sorting technique based on the sparse autocoder of convolution | |
CN109142851A (en) | A kind of novel power distribution network internal overvoltage recognition methods | |
CN104966161A (en) | Electric energy quality recording data calculating analysis method based on Gaussian mixture model | |
CN109212392A (en) | Recognition methods, system and the Caffe convolutional neural networks of direct current cables shelf depreciation defect failure | |
CN106707118B (en) | Partial Discharge Pattern Recognition Method and device | |
CN106950475A (en) | A kind of local discharge signal extracting method and device based on wavelet transformation | |
CN117434384A (en) | Power distribution network insulation fault identification method and related device |
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 |