CN110363130A - Voltage sag source discrimination method and device for identifying based on variation mode decomposition - Google Patents
Voltage sag source discrimination method and device for identifying based on variation mode decomposition Download PDFInfo
- Publication number
- CN110363130A CN110363130A CN201910610664.2A CN201910610664A CN110363130A CN 110363130 A CN110363130 A CN 110363130A CN 201910610664 A CN201910610664 A CN 201910610664A CN 110363130 A CN110363130 A CN 110363130A
- Authority
- CN
- China
- Prior art keywords
- modal components
- temporarily
- initial time
- waveform
- variation mode
- 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
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2134—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on separation criteria, e.g. independent component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Algebra (AREA)
- Operations Research (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention discloses voltage sag source discrimination method and device for identifying based on variation mode decomposition, discrimination method temporarily drops signal the following steps are included: obtaining primary voltage from electric energy quality monitoring wave recording device；Signal is temporarily dropped to primary voltage, and one original time-domain signal of acquisition is normalized；Variation mode decomposition is carried out to original time-domain signal, obtains K modal components；The modal components IMF N that can represent different temporarily drop sources simultaneously is filtered out from K modal components；It is recognized according to the feature that waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N corresponds to the exclusive modal components in all kinds of temporary drop sources.
Description
Technical field
Grid power quality of the present invention detects explanation field, and in particular to the voltage sag source based on variation mode decomposition is distinguished
Knowledge method.
Background technique
Voltage effective value is reduced between 0.1~0.9p.u, holds under the conditions of voltage dip (voltage sag) refers to power frequency
The continuous time is the short duration voltage variations phenomenon of 0.5 cycle to 1 minute.Voltage dip is inevitable in system normal course of operation
Short time disturbance phenomenon, higher level's power transmission network, distant place distribution and local distribution failure, transformer excitation, large-size machine starting
And switching of load etc. is simultaneity factor impedance, fault impedance, transformer parameter, guarantor the main reason for leading to voltage dip
The installation and parameter setting etc. of shield and relay protection, also have a major impact voltage dip.
But as computer technology, automatic technology, power electronic technique and microelectric technique etc. are in the wide of every profession and trade
General application, largely electric system is promoted in every profession and trade and accessed to the industrial process based on new and high technology and equipment, these
To electrical energy power quality disturbance, especially voltage dip short time disturbance is very sensitive for industrial process and equipment.When industrial process (or is set
It is standby) when interrupting (or failure) because of voltage dip, it can be very huge to being lost caused by industrial user and society.According to statistics, Europe
The number that voltage dip occurs every year for continent is about tens of times to 1,000 times, what 25 state, European Union bore by power quality problem every year
For economic loss up to 150,000,000,000 Euros, being damaged serious is mostly industry and services.And the U.S. every year because of voltage Sag Disturbance caused by
Direct economic loss be up to hundred million dollars of 120-260.So that voltage dip becomes current power consumer, society and academia most
The power quality problem be concerned about and paid close attention to.
From the point of view of the currently used voltage dip control measures, mainly eliminated from interference source, voltage dip route of transmission
Comprehensively considered in terms of the Ability of Resisting Disturbance three of inhibition and raising equipment.
But voltage dip is administered, inhibit and eliminate temporary drop source to be the pass for administering and improving Problem of Voltage Temporary-Drop from source
Key, and this is needed premised on how accurately recognizing to voltage sag source.Therefore, it is necessary to look for a kind of reliable identification
Accurate method accurately identifies to make to the type of voltage dip, could reasonably control from source and improve voltage dip and ask
Topic.
Summary of the invention
The purpose of the present invention is to provide the voltage sag source discrimination method based on variation mode decomposition, this method can be to short
Voltage dip caused by road failure, transformer switching, electric motor starting etc. is accurately recognized.This method is simple, calculating speed
Fastly, there is good noiseproof feature.
The specific technical proposal of the invention is:
Voltage sag source discrimination method based on variation mode decomposition, comprising the following steps:
Primary voltage is obtained from electric energy quality monitoring wave recording device temporarily drops signal；
Signal is temporarily dropped to primary voltage, and one original time-domain signal of acquisition is normalized；
Variation mode decomposition is carried out to original time-domain signal, obtains K modal components；
The modal components IMF N that can represent different temporarily drop sources simultaneously is filtered out from K modal components；
According to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The detailed process that the feature of the exclusive modal components in corresponding all kinds of temporary drop sources is recognized are as follows:
Variation mode decomposition is carried out to original time-domain signal, obtains the detailed process of K modal components are as follows: decomposes construction
K centre frequency is ω outkMode function ukIt and is ω to K centre frequencykMode function ukResolving is carried out to handle
To K modal components, wherein K is pre-set decomposition scale, also referred to as decomposition mode number.
The detailed process of K modal components is obtained using variation Mode Decomposition are as follows: including construction variational problem and resolving
Variational problem
Constructing variational problem includes:
First constructing K same centre frequency to original time-domain signal decomposition is ωkMode function uk；
Again by each mode function ukIt converts to obtain its analytic signal by Hilbert；
Analytic signal and centre frequency ω will be estimated againkIt is mixed, in each mode function ukThe sum of be equal to original time domain
Under the constraint condition of signal f (t), restrictive variational problem equation is obtained；
Resolving variational problem includes:
It will be constrained for restrictive variational problem equation using Lagrange multiplier operator λ (t) and secondary penalty factor α
Property variational problem it is equations turned for without restrictive variational problem equation；
The saddle point without restrictive variational problem equation is sought using alternating direction orange algorithm, wherein when modal components are full
When sufficient iteration stopping condition, stop updating, the modal components after output decomposition.
Original time-domain signal are as follows: y=f (t)；
Mode function ukIt is defined as an AM/FM amplitude modulation/frequency modulation signal, mode function ukAre as follows:
uk(t)=Ak(t)cosΦk(t)； (1)
In formula: phase Φ 'kIt (t) >=0 is a non-decreasing function, i.e. Φ 'k(t)≥0；AkIt (t) >=0 is envelope line function；
The A in pace of changek(t) and instantaneous frequency ωk(t)=Φ 'k(t) compare Φk(t) slow；
Restrictive variational problem equation are as follows:
Wherein:To seek local derviation to t；δ (t) is impulse function；For center frequency index；
Without restrictive variational problem equation are as follows:
Wherein: α is secondary penalty factor；λ (t) is Lagrange multiplier operator.
The specific implementation step of the saddle point without restrictive variational problem equation is sought using alternating direction orange algorithm are as follows:
A, it initializes
B, circulation n=n+1 is executed；
C, it when to all ω > 0, updatesThat is:
D, ω is updatedk；That is:
E, λ is updated；That is:
F, b~e is repeated, it is out of service until meeting formula (7), K modal components are just obtained, wherein ε is given differentiation essence
Degree；
Wherein for the parameter setting during variation mode decomposition are as follows: sample frequency 48000, sampling number are
1800, α=2000, mode number K=6 is decomposed in noise margin position 0.3, and no direct current component selects equality initialization.
The modal components that can represent different temporarily drop sources simultaneously are filtered out from K modal components method particularly includes:
1 temporarily drop apparent modal components IMF N of initial time T boundary, temporarily drop starting are filtered out from K modal components
The apparent modal components IMF N of moment boundary refers to that it can identify apparent rise for the voltage sag source of all classes
Begin moment T.
According to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The detailed process that the feature of the exclusive modal components in corresponding all kinds of temporary drop sources is recognized are as follows:
If waveform of the exclusive modal components in all kinds of temporary drop sources after temporarily drop initial time T all can be significantly mutated,
During temporarily drop continues, waveform is unchanged, and with time overlapping of axles, then regarding voltage sag source is short circuit class failure；
If the exclusive modal components in all kinds of temporary drop sources are in the waveform after temporarily drop initial time T since temporarily drop initial time
Apparent irregular variation will occur, and one of modal components waveform burr is obvious, then regards voltage sag source as transformation
Device switching；
If the exclusive modal components in all kinds of temporary drop sources the waveform after temporarily drop initial time T be after temporary drop starts and when
Between the straight line, unchanged that coincides of axis, and one of modal components have slight variation, but tend to be steady in general, then
Voltage sag source is regarded as electric motor starting.
General design idea of the invention is: first temporarily dropping signal to primary voltage and is pre-processed accordingly, then to it
The processing of variation mode decomposition is carried out, K modal components is obtained, a certain moduli state component is looked in K modal components, and
The certain moduli state component refers to the temporarily drop apparent modal components IMF N of initial time T boundary, and modal components IMF N is
Relative to that for all types of voltage dip signals, can find one, the apparent mould of initial time T boundary temporarily drops
Then state component IMF N judges waveform after temporary drop initial time T of other modal components in the certain moduli state component
Specific identification feature, therefore the exclusive modal components in all kinds of temporary drop sources can be obtained from K modal components, it finally identifies all kinds of
Under temporarily the exclusive modal components in drop source waveform after the temporarily drop initial time T corresponds to different temporary drop Source Types, the special knowledge of waveform
The matching relationship of other feature determines that the type of source signal temporarily drops in primary voltage.
For example, when it is 6 that K, which is arranged, know primary voltage temporarily drop signal be respectively short-circuit class failure, transformer switching,
In the case where electric motor starting type, we are handled by above-mentioned variation mode decomposition has been respectively obtained under short-circuit class failure correspondence
6 under 6 modal components (referring to Fig. 3), electric motor starting correspondence under 6 modal components (referring to Fig. 2), transformer switchings correspondences
A modal components (refer to Fig. 4), the modal components in observation chart 2, Fig. 3, Fig. 4 we it can be found that wherein in Fig. 2, Fig. 3, Fig. 4
As soon as I MF6 modal components have an obvious temporarily drop initial time T, therefore above-mentioned certain moduli state component refers to I MF6
Number modal components.
Later, we are again according to the temporary drop initial time T of I MF6 modal components found, to observe I MF3 mould
The waveform of state component, I MF4 modal components after temporarily drop initial time T；
It has been observed that
For short-circuit class failure (Fig. 2), temporarily drop initial time T after waveform, IMF3 and IMF4 can be obvious
Mutation, during temporary drop continues, waveform is unchanged, and with time overlapping of axles；
For transformer switching (Fig. 3), waveform after temporarily drop initial time T, IMF3 and IMF4 are from temporary drop
Moment beginning starts that apparent irregular variation can all occur, and the waveform burr of IMF4 component is obvious；
For electric motor starting (Fig. 4), waveform after temporarily drop initial time T, IMF3 be after temporary drop starts with
The straight line that time shaft coincides, unchanged, IMF4 has slight variation, but tends to be steady in general.
By above-mentioned rule, we can be found that for the voltage dip of above-mentioned 3 seed type, IMF3 and IMF4 exist
Temporarily the waveform after drop initial time T can all have above-mentioned respective special characteristic, i.e. it is only that IMF3 and IMF4 represent all kinds of temporary drop sources
Some modal components, therefore after we can preset features described above, then by the waveform analysis after temporary drop initial time T
It goes to match above-mentioned predetermined characteristic, then reversely obtains the temporary drop Source Type of unknown temporary drop signal.
Although the present invention is provided only when K is 6, above-mentioned technical proposal is able to achieve, and can be sought by the above method
Look under more type cases, K value is more often realized, therefore the present invention provides a confirmatory theory of above-mentioned K=6
It is bright, it does not represent and is only able to achieve in K=6, it is all that completed discrimination method is conceived based on foregoing invention, and according to the party
Method device obtained, should all be included in the scope of the present invention.
Based on the above principles, the present invention first proposes a kind of voltage sag source device for identifying, the voltage sag source device for identifying
Internal directly packaged variation mode decomposition model, and K=6 is set, certainly, based on inventive concept of the invention, K is set
It is set to other value, also should be regarded as equivalent.
Voltage sag source device for identifying based on variation mode decomposition,
Include:
Obtain the input unit that signal temporarily drops in primary voltage；
The construction device of variation mode decomposition model is constructed, it is pre-set when constructing variation mode decomposition model
Decomposition scale K value is encapsulated as 6；
The processing output device of 6 modal components obtained is handled by variation mode decomposition；
Obtain I MF6 modal components and I MF3 modal components, I MF4 modal components in 6 modal components
Acquisition device；
Extract the extraction dress of the temporary drop initial time T in the I MF6 modal components that can represent different temporarily drop sources simultaneously
It sets；
Analyze I MF3 modal components, I MF4 modal components temporarily drop initial time T after waveform state and
Export the device of fault type.
Exporting the temporary device for dropping Source Type includes:
The extraction dress of waveform after extracting I MF3 modal components, temporarily dropping initial time T in I MF4 modal components
It sets；
If the waveform of I MF3 modal components, I MF4 modal components after temporarily drop initial time T is all obviously dashed forward
When change, mark the output device exported temporarily to drop caused by short-circuit class failure；
If the waveform of I MF3 modal components, I MF4 modal components after temporarily drop initial time T all occurs bright
Show it is irregular variation but I MF4 modal components temporarily drop initial time T after waveform occur burr it is obvious when, then mark
Output is the output device temporarily dropped caused by transformer switching；
If I MF3 modal components are and time shaft in waveform of the temporarily drop initial time T after temporarily drop initial time T
The straight line and I MF4 modal components to coincide has slightly in waveform of the temporarily drop initial time T after temporarily drop initial time T
Variation when, then marking output is the output device that temporarily drops caused by electric motor starting.
Compared with prior art, the present invention having the following advantages and benefits:
1, the accurate recognition of voltage sag source is realized
In the case where electric network composition is fixed, the problem of for temporarily dropping source identification difficulty, this method can be directly to original
Voltage dip signal carries out the processing of variation mode decomposition, then passes through the otherness of IMF component (modal components), accurately picks out
It whether is voltage dip caused by short-circuit, transformer throwing or electric motor starting；And algorithm is simple, and visual result is reliable.
2, to inhibit and eliminating temporarily drop source from source, base has been established to reach effectively to administer and improve Problem of Voltage Temporary-Drop
Plinth:
Improvement and improvement problem for voltage dip, most effective way are exactly that voltage is administered or inhibited from source
The generation temporarily dropped, and this depends on the accurate recognition to voltage sag source.Such as: if short-circuit class failure causes temporarily to drop in system
Number is more, then can improve electric network composition and optimization relay protection etc..
3, the coordination dispute between sensitive equipment, power supply department and user is selected to provide foundation for user:
Problem of Voltage Temporary-Drop mainly makes the sensitive equipment in system and user fail, to influence system and industrial production
The normal operation of process, and different temporarily drop sources are different to the influence degree of sensitive equipment.Such as the starting or transformation of usual motor
It is three-phase voltage sag caused by the switching of device, and the duration is relatively longer, it is possible to cause three-phase sensitive equipment to fail, because
This can choose the temporarily drop better equipment of immunity.
For industrial user, the big electric motor starting in usual factory can cause voltage dip, it is possible to cause in factory
Other sensitive equipments or the sensitive equipment failure for closing on factory, to cause economic loss to user.Cause user and electric power
Supplier will generate dispute because of the responsible problem of economic loss.It therefore can be to coordination dispute to the accurate recognition in temporary drop source
With the offer foundation is provided.
Detailed description of the invention
Attached drawing described herein is used to provide to further understand the embodiment of the present invention, constitutes one of the application
Point, do not constitute the restriction to the embodiment of the present invention.In the accompanying drawings:
Fig. 1 is process flow diagram of the present invention.
Fig. 2 is 6 modal components decomposed under the conditions of short-circuit class fault-signal.
Fig. 3 is 6 modal components decomposing under the conditions of transformer switching signal.
Fig. 4 is 6 modal components decomposed under electric motor starting signal conditioning.
Fig. 5 has built three kinds of circuit structures and parameter setting for temporarily dropping source model.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention is made below with reference to embodiment
Further to be described in detail, exemplary embodiment of the invention and its explanation for explaining only the invention, are not intended as to this
The restriction of invention.
Embodiment one
Firstly, the technical problems to be solved by the invention are to identify to voltage sag source, the technology of the present invention is being understood
Before logic, the technology for solving voltage dip identifing source to existing this field is needed centainly to be understood.
It is existing that voltage sag source recognize mainly using signal processing, deep learning and both side combined
Method.
(1) method of signal processing:
Voltage dip signal is obtained module time-frequency matrixes and is carried out change of scale using S-transformation to obtain global template, from this
It is chosen in global template most can reflect that such temporarily drops the specific part in source as unique standard form, not eventually by solution
Know and drops the similarity degree of source signal part template and standard form temporarily to achieve the purpose that identification temporarily drop Source Type.
Hilbert-Huang transform (Hilbert-Huang Transform, HHT) and small wave energy are used to voltage dip signal
The method that amount spectrum combines identifies that carrying out empirical mode decomposition to temporary drop signal first is multiple natural modes to temporary drop source
State function (Intrinsic Mode Function, IMF), then each IMF decomposes simultaneously calculator energy spectrum into multi-level Wavelet Transform packet,
And then Hilbert-Huang spectrogram is obtained, by comparing the spy for being included in the Hilbert-Huang spectrogram in different temporarily drop sources
The difference of sign amount recognizes temporary drop source.
(2) method of deep learning
Voltage dip signal is extracted using convolutional neural networks (Convolutional Neural Networks, CNN)
Temporal aspect and space characteristics, and be used for purification high dimensional feature and play the full articulamentum depth confidence network of classification
The deep learning Model Fusion of (Deep Belief Network) replacement carries out the Classification and Identification in single and compound temporary drop source.
(3) method that signal processing and deep learning combine
To voltage dip signal use wavelet transformation (wavelet transform, WT), i.e., db6 small echo temporarily drop signal into
After 5 Scale Decomposition of row, its wavelet coefficient entropy and Energy Spectrum Entropy totally 6 feature composition input vectors are extracted, and be entered into probability
It is trained in neural network (probabilistic neural network, PNN), trained model is used for temporarily
Identifing source drops.
Prior art disadvantage:
(1) computationally intensive, it is difficult to apply in practice using S-transformation method because its time window is fixed；Using HHT and small
The problems such as method that wave packet energy spectrum combines, can encounter HHT there are modal overlap, the Decomposition order of small echo；(2) CNN is used
In the method that DBN is blended, there are the number of plies of the relevant parameter of mathematical model, network structure setting difficulties, and are used for actual electric network
When the problems such as needing the measured signal waveforms of a large amount of known temporarily drop Source Types to be trained；(3) side combined using WT with PNN
Method has certain dependence for the selection of wavelet basis and the determination of Decomposition order, and the determination of the structure and parameter of PNN is to temporary
The recognition accuracy in drop source has large effect, and a large amount of known sample data is needed to be trained model.
In actual application, since the frequency that voltage dip occurs usually will not be very high, and the electricity in actual electric network
Energy quality monitoring device is mounted on mostly in the power grid of 35kV and above, is difficult to monitor that such as big electric motor starting draws
The voltage dip risen.Therefore, it is impossible to which the measured data for obtaining a large amount of known temporarily drop type is come to using deep learning method
The training of model progress model.
To be applied in actual electric system, it is necessary to which method therefor noise resisting ability is strong, is adapted to complexity
Network structure, calculating speed is fast, required known sample is few, the setting without carrying out largely trained and model parameter, identification
Accuracy rate is high.Therefore, it finds the excellent signal processing method of one kind and the core that accurately identification is this patent is carried out to temporary drop source
Content and key technology.
In order to understand inventive concept of the invention, we first go to analyze design of the invention from positive angle, of the invention
Design basis be in the case where known three kinds temporarily drop Source Types, then to the processing of its variation mode decomposition, obtain K (K is default=
6) a modal components.
As shown in figure 5, having built three kinds of temporarily drop source models, including short trouble respectively in PSCAD/EMTDC environment
(a), transformer switching (b) and induction conductivity start (c), which is used for the verifying of the technical solution of the invention patent.
Its circuit structure and parameter setting are as shown in Figure 5.
On the basis of above-mentioned parameter and variation mode decomposition, 6 modal components under short-circuit class failure corresponds to have been obtained
6 modal components under 6 modal components (referring to Fig. 3) under (referring to Fig. 2), transformer switching are corresponding, electric motor starting are corresponding
(referring to Fig. 4).After obtaining above-mentioned data, modal components in our observation charts 2, Fig. 3, Fig. 4 we can be found that: wherein
Only I MF6 modal components have an obvious temporarily drop initial time in different temporarily drop Source Types in Fig. 2, Fig. 3, Fig. 4
T, therefore I MF6 modal components are exactly a certain moduli state component.
Then under this basis, judge temporary drop initial time T of other modal components in the certain moduli state component it
The specific identification feature of waveform afterwards, it has been found that it presents following phenomenon:
For short-circuit class failure (Fig. 2), temporarily drop initial time T after waveform, IMF3 and IMF4 can be obvious
Mutation, during temporary drop continues, waveform is unchanged, and with time overlapping of axles；
For transformer switching (Fig. 3), waveform after temporarily drop initial time T, IMF3 and IMF4 are from temporary drop
Moment beginning starts that apparent irregular variation can all occur, and the waveform burr of IMF4 component is obvious；
For electric motor starting (Fig. 4), waveform after temporarily drop initial time T, IMF3 be after temporary drop starts with
The straight line that time shaft coincides, unchanged, IMF4 has slight variation, but tends to be steady in general.
Therefore, we it can be concluded that, after obtaining temporarily dropping initial time, can in conjunction with IMF3 and IMF4 in initial time
The notable difference of waveform variation later carries out the identification in temporarily drop source.In summary, a temporarily drop initial time T can found
After the apparent modal components IMF N of boundary, temporary drop starting of other modal components in the certain moduli state component is then judged
The specific identification feature of waveform after moment T, to correspond to different waveform specific identification features by different temporarily drop Source Types
Matching relationship determine that the type of signal temporarily drops in primary voltage.
Therefore, by the experiment, the method that we can be utilized variation mode decomposition carrys out the class to voltage sag source
Type is that science is feasible into identification.Therefore available following methods:
Voltage sag source discrimination method based on variation mode decomposition, comprising the following steps:
Primary voltage is obtained from electric energy quality monitoring wave recording device temporarily drops signal；
Signal is temporarily dropped to primary voltage, and one original time-domain signal of acquisition is normalized；
Variation mode decomposition is carried out to original time-domain signal, obtains K modal components；
The modal components IMF N that can represent different temporarily drop sources simultaneously is filtered out from K modal components；
According to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The feature of the exclusive modal components in corresponding all kinds of temporary drop sources is recognized.
Variation mode decomposition is carried out to original time-domain signal, obtains the detailed process of K modal components are as follows: decomposes construction
K centre frequency is ω outkMode function ukIt and is ω to K centre frequencykMode function ukResolving is carried out to handle
To K modal components, wherein K is pre-set decomposition scale, also referred to as decomposition mode number.
The detailed process of K modal components is obtained using variation Mode Decomposition are as follows: including construction variational problem and resolving
Variational problem
Constructing variational problem includes:
First constructing K same centre frequency to original time-domain signal decomposition is ωkMode function uk；
Again by each mode function ukIt converts to obtain its analytic signal by Hilbert；
Analytic signal and centre frequency ω will be estimated againkIt is mixed, in each mode function ukThe sum of be equal to original time domain
Under the constraint condition of signal f (t), restrictive variational problem equation is obtained；
Resolving variational problem includes:
It will be constrained for restrictive variational problem equation using Lagrange multiplier operator λ (t) and secondary penalty factor α
Property variational problem it is equations turned for without restrictive variational problem equation；
The saddle point without restrictive variational problem equation is sought using alternating direction orange algorithm, wherein
When modal components meet iteration stopping condition, stop updating, the modal components after output decomposition.
Original time-domain signal are as follows: y=f (t)；
Mode function ukIt is defined as an AM/FM amplitude modulation/frequency modulation signal, mode function ukAre as follows:
uk(t)=Ak(t)cosΦk(t)； (1)
In formula: phase Φ 'kIt (t) >=0 is a non-decreasing function, i.e. Φ 'k(t)≥0；AkIt (t) >=0 is envelope line function；
The A in pace of changek(t) and instantaneous frequency ωk(t)=Φ 'k(t) compare Φk(t) slow；
Restrictive variational problem equation are as follows:
Wherein:To seek local derviation to t；δ (t) is impulse function；For center frequency index.
Without restrictive variational problem equation are as follows:
Wherein: α is secondary penalty factor；λ (t) is Lagrange multiplier operator.
The specific implementation step of the saddle point without restrictive variational problem equation is sought using alternating direction orange algorithm are as follows:
A, it initializes
B, circulation n=n+1 is executed；
C, it when to all ω > 0, updatesThat is:
D, ω is updatedk；That is:
E, λ is updated；That is:
F, b~e is repeated, it is out of service until meeting formula (7), K modal components are just obtained, wherein ε is given differentiation essence
Degree；
Wherein for the parameter setting during variation mode decomposition are as follows: sample frequency 48000, sampling number are
1800, α=2000, mode number K=6 is decomposed in noise margin position 0.3, and no direct current component selects equality initialization.
The modal components that can represent different temporarily drop sources simultaneously are filtered out from K modal components method particularly includes:
1 temporarily drop apparent modal components IMF N of initial time T boundary, temporarily drop starting are filtered out from K modal components
The apparent modal components IMF N of moment boundary refers to that it can be identified significantly for all types of voltage sag sources
Initial time T.
According to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The detailed process that the feature of the exclusive modal components in corresponding all kinds of temporary drop sources is recognized are as follows:
If waveform of the exclusive modal components in all kinds of temporary drop sources after temporarily drop initial time T all can be significantly mutated,
During temporarily drop continues, waveform is unchanged, and with time overlapping of axles, then regarding voltage sag source is short circuit class failure；
If the exclusive modal components in all kinds of temporary drop sources are in the waveform after temporarily drop initial time T since temporarily drop initial time
Apparent irregular variation will occur, and one of modal components waveform burr is obvious, then regards voltage sag source as transformation
Device switching；
If the exclusive modal components in all kinds of temporary drop sources the waveform after temporarily drop initial time T be after temporary drop starts and when
Between the straight line, unchanged that coincides of axis, and one of modal components have slight variation, but tend to be steady in general, then
Voltage sag source is regarded as electric motor starting.
Embodiment 2
A kind of discrimination method has been obtained since the present invention summarizes in embodiment 1, it can directly leading foundation
Above-mentioned parameter is packaged, to immediately arrive at the identification device of a kind of adaptation and the above method.
Voltage sag source device for identifying based on variation mode decomposition,
Include:
Obtain the input unit that signal temporarily drops in primary voltage；
The construction device of variation mode decomposition model is constructed, it is pre-set when constructing variation mode decomposition model
Decomposition scale K value is encapsulated as 6；
The processing output device of 6 modal components obtained is handled by variation mode decomposition；
Obtain I MF6 modal components and I MF3 modal components, I MF4 modal components in 6 modal components
Acquisition device；
Extract the extraction dress of the temporary drop initial time T in the I MF6 modal components that can represent different temporarily drop sources simultaneously
It sets；
Analyze I MF3 modal components, I MF4 modal components temporarily drop initial time T after waveform state and
The device of Source Type temporarily drops in output.
Exporting the temporary device for dropping Source Type includes:
The extraction dress of waveform after extracting I MF3 modal components, temporarily dropping initial time T in I MF4 modal components
It sets；
If the waveform of I MF3 modal components, I MF4 modal components after temporarily drop initial time T is all obviously dashed forward
When change, label output short-circuit class failure output device；
If the waveform of I MF3 modal components, I MF4 modal components after temporarily drop initial time T all occurs bright
Show irregular variation but I MF4 modal components the waveform after temporarily drop initial time T occur burr it is obvious when, then mark
Export the output device of transformer switching；
If I MF3 modal components are and time shaft in waveform of the temporarily drop initial time T after temporarily drop initial time T
The straight line and I MF4 modal components to coincide has slightly in waveform of the temporarily drop initial time T after temporarily drop initial time T
Variation when, then mark output motor start output device.
It should be understood that since the variation mode decomposition of use of the invention is carried out according to existing design substantially, this hair
Bright creativeness is variation mode decomposition algorithm being applied to voltage sag source type identification, and has found a kind of feasibility
Identification decision process, if therefore the variation mode decomposition that is previously mentioned of the present invention there is principle and detailed content, do not repeating.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects
It is described in detail, it should be understood that being not intended to limit the present invention the foregoing is merely a specific embodiment of the invention
Protection scope, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all include
Within protection scope of the present invention.
Claims (10)
1. the voltage sag source discrimination method based on variation mode decomposition, comprising the following steps:
Primary voltage is obtained from electric energy quality monitoring wave recording device temporarily drops signal；
Signal is temporarily dropped to primary voltage, and one original time-domain signal of acquisition is normalized；
Variation mode decomposition is carried out to original time-domain signal, obtains K modal components；
The modal components IMF N that can represent different temporarily drop sources simultaneously is filtered out from K modal components；
It is corresponding according to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The feature of the exclusive modal components in all kinds of temporary drop sources is recognized.
2. the voltage sag source discrimination method according to claim 1 based on variation mode decomposition, which is characterized in that original
The time-domain signal of beginning carries out variation mode decomposition, obtains the detailed process of K modal components are as follows: decomposition constructs K center frequency
Rate is ωkMode function ukIt and is ω to K centre frequencykMode function ukResolving is carried out to handle to obtain K mode point
Amount, wherein K is pre-set decomposition scale, also referred to as decomposition mode number.
3. the voltage sag source discrimination method according to claim 2 based on variation mode decomposition, which is characterized in that use
Variation Mode Decomposition obtains the detailed process of K modal components are as follows: including construction variational problem and resolves variational problem.
Constructing variational problem includes:
First constructing K same centre frequency to original time-domain signal decomposition is ωkMode function uk；
Again by each mode function ukIt converts to obtain its analytic signal by Hilbert；
Analytic signal and centre frequency ω will be estimated againkIt is mixed, in each mode function ukThe sum of be equal to original time-domain signal
Under the constraint condition of f (t), restrictive variational problem equation is obtained；
Resolving variational problem includes:
Binding character is become using Lagrange multiplier operator λ (t) and secondary penalty factor α for restrictive variational problem equation
Divide problem equations turned for without restrictive variational problem equation；
The saddle point without restrictive variational problem equation is sought using alternating direction orange algorithm, wherein
When modal components meet iteration stopping condition, stop updating, the modal components after output decomposition.
4. the voltage sag source discrimination method according to claim 2 based on variation mode decomposition, which is characterized in that original
Time-domain signal are as follows: y=f (t)；
Mode function ukIt is defined as an AM/FM amplitude modulation/frequency modulation signal, mode function ukAre as follows:
uk(t)=Ak(t)cosΦk(t)； (1)
In formula: phase Φ 'kIt (t) >=0 is a non-decreasing function, i.e. Φ 'k(t)≥0；AkIt (t) >=0 is envelope line function；Becoming
Change A in speedk(t) and instantaneous frequency ωk(t)=Φ 'k(t) compare Φk(t) slow；
Restrictive variational problem equation are as follows:
Wherein:To seek local derviation to t；δ (t) is impulse function；For center frequency index.
Without restrictive variational problem equation are as follows:
Wherein: α is secondary penalty factor；λ (t) is Lagrange multiplier operator.
5. the voltage sag source discrimination method according to claim 2 based on variation mode decomposition, which is characterized in that utilize
Alternating direction orange algorithm seeks the specific implementation step of the saddle point without restrictive variational problem equation are as follows:
A, it initializesn；
B, circulation n=n+1 is executed；
C, it when to all ω > 0, updatesThat is:
D, ω is updatedk；That is:
E, λ is updated；That is:
F, b~e is repeated, it is out of service until meeting formula (7), K modal components are just obtained, wherein ε is given discrimination precision；
6. the voltage sag source discrimination method according to claim 5 based on variation mode decomposition, which is characterized in that wherein
For the parameter setting during variation mode decomposition are as follows: sample frequency 48000, sampling number 1800, α=2000 are made an uproar
Acoustic capacitance limit 0.3, decomposes mode number K=6, and no direct current component selects equality initialization.
7. the voltage sag source discrimination method according to claim 1 based on variation mode decomposition, which is characterized in that from K
The modal components that can represent different temporarily drop sources simultaneously are filtered out in a modal components method particularly includes:
1 temporarily drop apparent modal components IMF N of initial time T boundary is filtered out from K modal components, temporarily drops initial time
When the apparent modal components IMF N of boundary refers to that it can identify apparent starting for the voltage sag source of all classes
Carve T.
8. the voltage sag source discrimination method according to claim 7 based on variation mode decomposition, which is characterized in that
It is corresponding according to waveform of the exclusive modal components in all kinds of temporary drop sources after the temporary drop initial time T of modal components IMF N
The detailed process that the feature of the exclusive modal components in all kinds of temporary drop sources is recognized are as follows:
If waveform of the exclusive modal components in all kinds of temporary drop sources after temporarily drop initial time T all can be significantly mutated, dropped temporarily
In lasting process, waveform is unchanged, and with time overlapping of axles, then regard voltage sag source as short-circuit class failure；
If the exclusive modal components in all kinds of temporary drop sources all can since temporarily drop initial time in the waveform after temporarily drop initial time T
Apparent irregular variation occurs, and one of modal components waveform burr is obvious, then regards voltage sag source as transformer throwing
It cuts；
If the exclusive modal components in all kinds of temporary drop sources are and time shaft after temporary drop starts in the waveform after temporarily drop initial time T
It is the straight line that coincides, unchanged, and one of modal components have slight variation, but tend to be steady in general, then regard electricity
Temporarily drop source is electric motor starting to pressure.
9. the voltage sag source device for identifying based on variation mode decomposition, which is characterized in that
Include:
Obtain the input unit that signal temporarily drops in primary voltage；
The construction device of variation mode decomposition model is constructed, when constructing variation mode decomposition model, the K of K modal components
Value is encapsulated as 6；
The processing output device of 6 modal components obtained is handled by variation mode decomposition；
Obtain I MF6 modal components and I MF3 modal components in 6 modal components, I MF4 modal components obtain
Take device；
Extract the extraction element of the temporary drop initial time T in the I MF6 modal components that can represent different temporarily drop sources simultaneously；
Analysis I MF3 modal components, I MF4 modal components temporarily drop initial time T after waveform state and export
The device of fault type.
10. the voltage sag source device for identifying according to claim 9 based on variation mode decomposition, which is characterized in that
Output fault type device include:
The extraction element of waveform after extracting I MF3 modal components, temporarily dropping initial time T in I MF4 modal components；
If I MF3 modal components, I MF4 modal components temporarily drop initial time T after waveform all obviously be mutated when,
Mark the output device of output short-circuit class failure；
If obvious nothing all occurs for the waveform of I MF3 modal components, I MF4 modal components after temporarily drop initial time T
Rule variation but I MF4 modal components temporarily drop initial time T after waveform occur burr it is obvious when, then mark output
The output device of transformer switching；
If I MF3 modal components are mutually be overlapped with time shaft in waveform of the temporarily drop initial time T after temporarily drop initial time T
The straight line and I MF4 modal components of conjunction have slight change in waveform of the temporarily drop initial time T after temporarily drop initial time T
When change, then the output device of output motor starting is marked.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610664.2A CN110363130B (en) | 2019-07-08 | 2019-07-08 | Voltage sag source identification method and identification device based on variational modal decomposition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610664.2A CN110363130B (en) | 2019-07-08 | 2019-07-08 | Voltage sag source identification method and identification device based on variational modal decomposition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110363130A true CN110363130A (en) | 2019-10-22 |
| CN110363130B CN110363130B (en) | 2023-01-13 |
Family
ID=68218515
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910610664.2A Active CN110363130B (en) | 2019-07-08 | 2019-07-08 | Voltage sag source identification method and identification device based on variational modal decomposition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110363130B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111189639A (en) * | 2020-01-08 | 2020-05-22 | 重庆交通大学 | Bearing fault diagnosis method based on instantaneous frequency optimization VMD |
| CN112152224A (en) * | 2020-11-24 | 2020-12-29 | 国网四川省电力公司电力科学研究院 | Voltage sag treatment optimization method and system |
| CN114184889A (en) * | 2021-12-01 | 2022-03-15 | 厦门大学 | VMD-based power grid voltage sag fault signal detection method and system |
| WO2022119426A1 (en) * | 2020-12-01 | 2022-06-09 | Université Internationale de RABAT | Intelligent system for immediate detection and notification of disturbances in electrical signal quality |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004336846A (en) * | 2003-05-01 | 2004-11-25 | Tokyo Electric Power Co Inc:The | Method of determining variation factor of electrical quantity |
| CN103424600A (en) * | 2013-08-20 | 2013-12-04 | 昆明理工大学 | Voltage sag source identification method based on Hilbert-Huang transformation and wavelet packet energy spectra |
| US20140268458A1 (en) * | 2013-03-14 | 2014-09-18 | Innovolt, Inc. | Systems and methods for detecting and determining sources of power disturbances in connection with effective remediation |
| CN106199338A (en) * | 2016-07-20 | 2016-12-07 | 东南大学 | A kind of discrimination method of short trouble type voltage sag source |
| CN106226587A (en) * | 2016-07-01 | 2016-12-14 | 浙江工业大学 | A kind of exchange micro-capacitance sensor voltage dip method for quick based on LES HHT |
| CN108075437A (en) * | 2017-11-27 | 2018-05-25 | 国网江苏省电力有限公司电力科学研究院 | A kind of voltage sag source identification method based on wavelet transformation |
| CN109470985A (en) * | 2018-06-19 | 2019-03-15 | 国网浙江省电力有限公司湖州供电公司 | A kind of voltage sag source identification methods based on more resolution singular value decompositions |
| CN109633368A (en) * | 2018-12-03 | 2019-04-16 | 三峡大学 | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA |
-
2019
- 2019-07-08 CN CN201910610664.2A patent/CN110363130B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004336846A (en) * | 2003-05-01 | 2004-11-25 | Tokyo Electric Power Co Inc:The | Method of determining variation factor of electrical quantity |
| US20140268458A1 (en) * | 2013-03-14 | 2014-09-18 | Innovolt, Inc. | Systems and methods for detecting and determining sources of power disturbances in connection with effective remediation |
| CN103424600A (en) * | 2013-08-20 | 2013-12-04 | 昆明理工大学 | Voltage sag source identification method based on Hilbert-Huang transformation and wavelet packet energy spectra |
| CN106226587A (en) * | 2016-07-01 | 2016-12-14 | 浙江工业大学 | A kind of exchange micro-capacitance sensor voltage dip method for quick based on LES HHT |
| CN106199338A (en) * | 2016-07-20 | 2016-12-07 | 东南大学 | A kind of discrimination method of short trouble type voltage sag source |
| CN108075437A (en) * | 2017-11-27 | 2018-05-25 | 国网江苏省电力有限公司电力科学研究院 | A kind of voltage sag source identification method based on wavelet transformation |
| CN109470985A (en) * | 2018-06-19 | 2019-03-15 | 国网浙江省电力有限公司湖州供电公司 | A kind of voltage sag source identification methods based on more resolution singular value decompositions |
| CN109633368A (en) * | 2018-12-03 | 2019-04-16 | 三峡大学 | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA |
Non-Patent Citations (8)
| Title |
|---|
| LIU YAJUAN等: "Classification and Identification of Voltage Sag Sources in Distribution Network with Harmonic", 《3RD INTERNATIONAL CONFERENCE ON MATERIALS ENGINEERING, MANUFACTURING TECHNOLOGY AND CONTROL (ICMEMTC 2016)》 * |
| M.MANJULA等: "Detection and Classification of Voltage Sag Causes based on Empirical Mode Decomposition", 《2011 ANNUAL IEEE INDIA CONFERENCE》 * |
| PANKAJ D. ACHLERKAR等: "Detection of Voltage Variation Events Using Variational Mode Decomposition", 《2017 IEEE PES ASIA-PACIFIC POWER AND ENERGY ENGINEERING CONFERENCE (APPEEC)》 * |
| 丁宁等: "电压暂降源识别方法研究", 《电网技术》 * |
| 于婷: "电压暂降的检测与定位方法研究", 《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》 * |
| 张琛等: "基于变分模态分解奇异值熵的滚动轴承微弱故障辨识方法", 《振动与冲击》 * |
| 祁博等: "基于EMD-SVM的化工厂配电网电压暂降源识别", 《计算机与应用化学》 * |
| 祁博等: "基于Hilbert-Huang 变换和小波包能量谱的电压暂降源识别", 《中国电力》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111189639A (en) * | 2020-01-08 | 2020-05-22 | 重庆交通大学 | Bearing fault diagnosis method based on instantaneous frequency optimization VMD |
| CN111189639B (en) * | 2020-01-08 | 2021-09-14 | 重庆交通大学 | Bearing fault diagnosis method based on instantaneous frequency optimization VMD |
| CN112152224A (en) * | 2020-11-24 | 2020-12-29 | 国网四川省电力公司电力科学研究院 | Voltage sag treatment optimization method and system |
| WO2022119426A1 (en) * | 2020-12-01 | 2022-06-09 | Université Internationale de RABAT | Intelligent system for immediate detection and notification of disturbances in electrical signal quality |
| CN114184889A (en) * | 2021-12-01 | 2022-03-15 | 厦门大学 | VMD-based power grid voltage sag fault signal detection method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110363130B (en) | 2023-01-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110363130A (en) | Voltage sag source discrimination method and device for identifying based on variation mode decomposition | |
| CN103713237B (en) | A kind of power system transmission line short trouble diagnostic method | |
| Gu et al. | High impedance fault detection in overhead distribution feeders using a DSP-based feeder terminal unit | |
| CN105044556B (en) | Low current grounding self-adaption route selection method based on the sparse evidence fusion of atom | |
| CN103760464B (en) | Solve and the low current neutral grounding system fault route selecting method of SVM based on analysis diagram | |
| CN103941163A (en) | Resonant earthed system fault line selection method utilizing fuzzy K-means clustering | |
| Luo et al. | Stacked auto-encoder based fault location in VSC-HVDC | |
| CN103745119A (en) | Oil-immersed transformer fault diagnosis method based on fault probability distribution model | |
| CN108414872A (en) | A kind of earth-fault detecting method for adjusting power source system for power plant's system | |
| CN104330676A (en) | Transformer substation overvoltage intelligence monitoring system and method | |
| CN111308260B (en) | Electric energy quality monitoring and electric appliance fault analysis system based on wavelet neural network and working method thereof | |
| CN105092997A (en) | Identification method of lightning shielding failure and lightning back flashover of high-voltage transmission line | |
| CN109375051A (en) | Lightning Transient signal recognition method and system based on spectral density decaying | |
| CN103323741A (en) | Method for judging D-shaped cable hybrid circuit fault section directing at strong fault and based on fault voltage initial row amplitude value comparison | |
| CN112462193B (en) | Automatic reclosing judgment method for power distribution network based on real-time fault filtering data | |
| CN104198890A (en) | Intelligent-line-selecting small-current grounding system of medium-and-low-voltage power distribution network | |
| Liang et al. | Multilabel classification model for type recognition of single-phase-to-ground fault based on KNN-Bayesian method | |
| Damala et al. | A simple decision tree-based disturbance monitoring system for VSC-based HVDC transmission link integrating a DFIG wind farm | |
| CN103823158B (en) | Use the malfunction route selection method for resonant grounded system of not bending moment | |
| Nguyen et al. | Wavelet-based Neural Network for recognition of faults at NHABE power substation of the Vietnam power system | |
| CN111896890B (en) | Micro-grid line fault diagnosis method and system based on Hilbert-Huang transform | |
| CN116298509A (en) | Online identification method for harmonic resonance of power system | |
| Chen et al. | Coherent clustering method based on weighted clustering of multi-indicator panel data | |
| CN116298675A (en) | Intelligent algorithm-based power distribution network line loss anomaly detection method and system | |
| Hu et al. | A data-driven method of users-transformer relationship identification in the secondary power distribution 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 |