CN110210107A - Relay contact arcing state of strength detection method, device, equipment and medium - Google Patents
Relay contact arcing state of strength detection method, device, equipment and medium Download PDFInfo
- Publication number
- CN110210107A CN110210107A CN201910453959.3A CN201910453959A CN110210107A CN 110210107 A CN110210107 A CN 110210107A CN 201910453959 A CN201910453959 A CN 201910453959A CN 110210107 A CN110210107 A CN 110210107A
- Authority
- CN
- China
- Prior art keywords
- state
- data
- relational model
- strength
- arcing
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/327—Testing of circuit interrupters, switches or circuit-breakers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention discloses a kind of relay contact arcing state of strength detection methods, are related to relay technical field, for solving existing cumbersome modeling process, method includes the following steps: obtaining sample data;The corresponding actual strength information of contact striking state of strength is obtained, relational model is constructed;SE state is predicted according to detection data and the relational model, obtains the predicted intensity information of the corresponding SE state of the detection data.The invention also discloses a kind of relay contact arcing state of strength detection device, electronic equipment and computer storage mediums.The present invention obtains SE status information by modeling to relay coil.
Description
Technical field
The present invention relates to a kind of relay technical field more particularly to a kind of relay contact arcing state of strength detection sides
Method, device, equipment and medium.
Background technique
Relay is mainly made of armature, coil and contact, controls armature according to the power on/off of coil and energization size
With the cooperation of contact.Therefore pass through the status data of acquisition relay contact arcing intensity (SE), it will be appreciated that relay contact
The case where being subjected to impact arcing.
Existing is by modeling to relay to the research of contact striking state of strength, and this mode needs longer build
Mold process, and its result is only applicable to laboratory environment, is difficult to stablize utilization in actual condition.
Summary of the invention
For overcome the deficiencies in the prior art, one of the objects of the present invention is to provide a kind of relay contact arcing intensity
Condition detection method by modeling to relay coil, and then obtains SE status information.
An object of the present invention is implemented with the following technical solutions:
A kind of relay contact arcing state of strength detection method, comprising the following steps:
Sample data is obtained, the sample data is to carry out multiple repairing weld, the coil-end number of acquisition to relay coil end
According to;
The corresponding actual strength information of contact striking state of strength is obtained, the contact striking state of strength is denoted as SE shape
State, according to the sample data and actual strength information architecture relational model, the relational model is coil end data and SE shape
The mathematical relationship of state;
SE state is predicted according to detection data and the relational model, obtains the corresponding SE of the detection data
The predicted intensity information of state, the detection data are the coil end data obtained in SE state-detection.
Further, the sample data and detection data are basic parameter or/and transformation parameter, the underlying parameter are
Low-frequency current Ilow, high-frequency current Ihigh, one of voltage U or a variety of, the transformation parameter is one or more base
Plinth parameter passes through the derivative parameter being calculated.
Further, the transformation parameter is coil equivalent impedance, coil transient state induction reactance, voltage change ratio and curent change
One of rate is a variety of.
Further, the relational model is machine learning model.
Further, the building relational model, including feature annotation step and model training step, in which:
The feature annotation step, comprising:
SE state is described;
Wherein, SE ' indicates the state description set in SE state, sehThe h next state represented in state description set is retouched
It states, the state description characterization impact arcing intensity, h is the sum of state description;
Experiment method is passed through to all state descriptions of state description set respectively and finds corresponding actual strength information;Its
I-th state description se in middle SE stateiCorresponding actual strength informationThe actual strength informationFor i-th state
Corresponding real impact arcing intensity, 1≤i≤h are described;
The model training step, comprising:
Relational model is created, using sample data as input,Relational model is trained as true tag, is obtained
It is exported to predictionIt is expressed asWherein, S is sample data, and alg (S) is SE state and coil-end number
Relational model between;
Calculate prediction outputAnd true tagBetween error, if the error be not more than preset threshold value,
Then the corresponding relational model of SE state is alg (S), conversely, then continuing to train, until prediction outputAnd true tag
Between error be not more than preset threshold value;
Finally obtain the relational model of SE state:
Alg (S)=t 'se
t′seFor using coil end data as the corresponding predicted intensity information of SE state when input.
Further, the relational model is wavelet transformation model;
The building relational model, comprising:
By experiment method, sample data when the corresponding actual strength information of SE state, the actual strength letter are obtained
Breath is the corresponding real impact arcing intensity of SE state, and by Fourier transformation, by the corresponding actual strength of the SE state
Sample data expansion when information, obtains sample expanding data;
Wavelet basis is established with the sample expanding data;
The mathematical relationship of SE state and coil end data is established by detecting wavelet basis.
The second object of the present invention is to provide a kind of relay contact arcing state of strength detection system, by after
Electric apparatus coil is modeled, and then obtains SE status information.
The second object of the present invention is implemented with the following technical solutions:
A kind of relay contact arcing state of strength detection device comprising:
Data acquisition module, for obtaining sample data, the sample data is repeatedly to be adopted to relay coil end
Sample, the coil end data of acquisition;
Model construction module fires the contact for obtaining the corresponding actual strength information of contact striking state of strength
Arc state of strength is denoted as SE state, according to the sample data and actual strength information architecture relational model, the relational model
For the mathematical relationship of coil end data and SE state;
As a result output module obtains institute for predicting according to detection data and the relational model SE state
The predicted intensity information of the corresponding SE state of detection data is stated, the detection data is the coil-end obtained in SE state-detection
Data.
The third object of the present invention is to provide the electronic equipment for executing one of goal of the invention comprising processor, storage
Medium and computer program, the computer program are stored in storage medium, and the computer program is executed by processor
Shi Shixian above-mentioned relay contact arcing state of strength detection method.
The fourth object of the present invention is to provide the computer readable storage medium of one of storage goal of the invention, store thereon
There is computer program, the computer program realizes above-mentioned relay contact arcing state of strength detection when being executed by processor
Method.
Compared with prior art, the beneficial effects of the present invention are:
The present invention is studied by the sampled data to relay coil end, is obtained between coil end data and SE state
Connection, and then obtain the case where relay contact is subjected to impact arcing whether occur and contact is subjected to impact the intensity of arcing
Size has abandoned the cumbersome of existing mathematical modeling.
Detailed description of the invention
Fig. 1 is the schematic diagram of invention relay contact striking state of strength detection method;
Fig. 2 is the flow chart of the relay contact arcing state of strength detection method of embodiment one;
Fig. 3 is the process of the relay contact arcing state of strength detection method of embodiment three;
Fig. 4 is the structural block diagram of the relay contact arcing state of strength detection device of embodiment five;
Fig. 5 is the structural block diagram of the electronic equipment of embodiment six.
Specific embodiment
Below with reference to attached drawing, the present invention is described in more detail, it should be noted that lower reference attached drawing is to this
The description that invention carries out is only illustrative, and not restrictive.It can be combined with each other between each difference embodiment, with
Constitute the other embodiments not shown in the following description.
Embodiment one
Embodiment one provides a kind of relay contact arcing state of strength detection method, it is intended to by carrying out to coil-end
Data acquisition, and then calculate the relationship between coil end data and relay contact arcing state of strength, in this way, energy
Enough sizes for effectively understanding relay and impacting arcing intensity during being attracted release etc., instead of cumbersome mathematical modeling mistake
Journey, to obtain the case where relay contact is subjected to impact arcing and contact when is subjected to impact arcing and impact arcing is strong
The size of degree.
It please refers to shown in Fig. 1, the distortion, that is, disturbance of magnetic field and contact striking (contact impingement arcing) in magnetic field are to correspond
, contact striking can bring disturbance of magnetic field, and disturbance of magnetic field will lead to the variation of coil end data, therefore, pass through phenomenon pair
The result of phenomenon can analyze the relationship between the variation and contact striking of coil end data, therefore, according to coil end data
The available relay contact of variation the case where being subjected to impact arcing and contact when be subjected to impact arcing and impact combustion
The size of arc intensity.
According to above-mentioned principle, relational model is built, to realize that acquisition coil end data is input to relational model and can obtain
Relay contact is subjected to impact the case where arcing out and when contact is subjected to impact arcing and impacts the size of arcing intensity.
Shown in referring to figure 2., a kind of relay contact arcing state of strength detection method comprising following steps:
S110, sample data is obtained, the sample data is to carry out multiple repairing weld, the coil of acquisition to relay coil end
End sample data.
These sample datas can be that (period of motion includes relay actuation → disconnection in the period of motion of armature
→ be attracted) in, multiple data acquisition is carried out with fixed or unfixed frequency, can be and repeatedly counted in multiple periods of motion
According to acquisition, it is also possible to based on experiment number, experiment acquires one or more sample datas every time, does not do specific limit here
It is fixed.Sample data acquisition is more, and the relational model finally obtained is also more accurate.
S120, the corresponding actual strength information of SE state is obtained, according to the sample data and actual strength information architecture
Relational model, the relational model are the mathematical relationship of coil end data and SE state.
Obtaining the corresponding actual strength information of SE state is obtained by experiment method, is rushed for example, being grabbed with oscillograph
The current status of moment is hit, and then the intensity description that impact occurs can be read out on its screen, timing at this time is t1, then t1
As the actual strength description of SE state, the description of this actual strength are referred to as actual strength information, refer to true combustion
Arc intensity.
Here the manifestation mode of relational model is not limited, as long as coil end data can be characterized and the mathematical algorithm of SE state is equal
It can, it should be noted that the mathematical algorithm for characterizing coil end data and SE state here is carried out with the strength information of SE state
, i.e., by the relational model, it can predict the strength information of corresponding SE state when any coil end data, i.e. predicted intensity
Information.
S130, SE state is predicted according to detection data and the relational model, obtains the detection data pair
The predicted intensity information for the SE state answered, the detection data are the coil end data obtained in SE state-detection.
Due to the mathematical relationship of above-mentioned relational model available coil end data and SE state, pass through the mathematics
Relationship carries out inverting, can obtain the real data of the corresponding coil-end of SE state, can also pass through the coil-end of actual measurement
Data (detection data) calculate the contact striking intensity of SE state, i.e. predicted intensity information.
For the acquisition of predicted intensity information, mode without limitation, in the base for the actual strength information for knowing SE state
On plinth, marriage relation model can predict the contact striking intensity of SE state.
It should be understood that the present embodiment first is that using changes of magnetic field principle, by the coil end data with intensive properties
It is associated with SE state, it is hereby achieved that the relationship between the variation of coil end data and SE state, then using this
Relationship is it can be concluded that when the case where relay contact is subjected to impact arcing under some coil end data and contact are rushed
It hits arcing and impacts the size of arcing intensity.
Embodiment two
Embodiment is second is that the improvement carried out on the basis of embodiment one, mainly to the type of sample data and detection data
It is explained and illustrated.
The data type of sample data and detection data and without limitation, because influence of the changes of magnetic field to coil-end can be with
Lead to the variation of any coil-end electrical parameter, therefore one or more electrical parameters of any coil-end that can be obtained can be used as
Sample data and detection data, and acquire these coil-ends electrical parameter mode also without limitation, pass through mutual inductor, voltage
Table, ammeter, oscillograph, the various sample circuits built etc., as long as the corresponding electrical parameter of coil-end can be obtained.
It should be noted that sample data and detection data be theoretically it is corresponding, that is, acquire certain or certain types
Sample data, to construct relational model, then in acquisition testing data, it is also desirable to directly or indirectly obtain and sample data
Identical data type.Here be indirectly because the certain parameter types of coil-end are by certain operation, it is available in addition
Parameter type, i.e. transformation parameter.
Specifically, sample data and detection data are one of underlying parameter and transformation parameter or a variety of, here
Underlying parameter is low-frequency current Ilow, high-frequency current Ihigh, one of voltage U or a variety of, transformation parameter be one or one with
On underlying parameter by the derivative parameter that is calculated, transformation parameter includes that peace is not limited to coil equivalent impedance, coil transient state
One of induction reactance, voltage change ratio and current changing rate are a variety of one or more.Wherein:
Coil equivalent impedance Rz can pass through the voltage U and low-frequency current I in underlying parameterlow(or high-frequency current Ihigh)
It is divided by obtain: Rz=U (t)/I (t);Coil transient state induction reactance RL can be obtained by coil equivalent impedance Rz and coil DC internal resistance
Arrive, that is, pass through Rz=| Rcoil+RL*j | obtain RL=sqrt ((Rz)2-(Rcoil)2), wherein Rcoil is in coil direct current
Resistance can be obtained by relay handbook or directly measurement;Voltage change ratio U ' can be by seeking the voltage U in underlying parameter
It leads to obtain: U '=dU/dt;Current changing rate I ' can be high-frequency current change rate, be also possible to low-frequency current change rate, and I '=
dI/dt。
In fact, transformation parameter mainly has two major classes, it is electric parameter class and timing class respectively, electric parameter class is mainly led to
It crosses underlying parameter to be calculated, timing class is obtained by being converted such as derivation to underlying parameter, coil equivalent impedance, coil
Transient state induction reactance belongs to electric parameter class, and voltage change ratio and current changing rate belong to timing class.Above are only to transformation parameter into
Capable citing, the transformation parameter that other are calculated accordingly by underlying parameter and timing converts belong to the present invention
Protection scope.
Embodiment three
Embodiment is third is that carry out on the basis of embodiment one or/and embodiment two.In the third embodiment, relational model
Using machine learning model, i.e., the mathematical relationship of coil end data Yu SE state is obtained by way of machine learning.
Machine learning model includes but is not limited to decision tree, random forest, logistic regression, support vector machines, Bayes, K
Neighbour, K mean value and deep learning (various artificial neural networks) etc..Machine learning is that computer is based on sample data progress machine
Tool study, learning from instruction, analogical learning or event selection etc. obtain corresponding relational model.Sample data is more, engineering
The relational model that acquistion is arrived is also more accurate.The relational model that machine learning obtains can be tested by some sample datas
Card, to judge whether relational model needs to optimize or update, optimization or update are by way of increasing sample data to pass
It is that model improves process.According to its learning strategy, slightly different for each machine learning, but arbitrarily obtains machine learning mould
The mode of type can be applied in the present invention.
Below by taking deep learning as an example, the building of machine learning model is explained and illustrated.Shown in referring to figure 3.,
Itself the following steps are included:
S210, sample data is obtained, the sample data is to carry out multiple repairing weld, the coil of acquisition to relay coil end
End data.
It is grabbed using data of any acquisition mode to coil-end, obtains sample data:
Wherein: S is sample data sets, and k represents the component of coil end data, including but not limited to low-frequency current Ilow, high
Frequency electric current Ihigh, the underlying parameter of one or more of signals such as voltage U or the combination of transformation parameter.N represents sampled data
Maximum length (number), it should be noted that the value of n only needs to be adjusted according to the actual situation, be not fix
Length.
S220, feature mark.
1, the state description set of SE state is indicated with SE ', it may be assumed thatWherein sehIt represents the h times
The arcing time is impacted in sports immunology.
2, the state description se of the first time experiment in SE ' is taken out1, se can be found by being looked for by experiment1Entirely it is being moved through
Corresponding practical arcing intensity in journeySuch as assume se1Experiment relay impact arcing intensity description for the first time is represented, then
Its se can be obtained by following experiment1The actual strength of generation: the current status of impact moment is grabbed with oscillograph, in turn
The intensity description that impact occurs can be read out on its screen, and timing at this time isThenAs se1Actual time
Description.
3, other states se in SE ' is successively taken out2~seh, and corresponding state generation is marked out according to corresponding experiment
Temporal informationIt can be expressed as
S230, model training.
The result obtained for step S210And step S220
Obtained resultIt performs the following operations:
1, se is taken out1Corresponding Label resultUsing S as input,Algorithm alg is carried out as model answer
Training obtains prediction outputIt is expressed as
2, the error Loss calculated between prediction output and true Label is expressed as
It should be noted that not confining the functional based method for calculating Loss herein, it is only necessary to react the error of actual application scenarios i.e.
Can, such as can be difference of two squares Loss calculation method.
3, other arcing state of strength are successively taken out and describe se2~sehAnd S respectively repeats steps 1~2 as input;
According to LossseAs a result adjustment algorithm model alg is finally reached certain Loss threshold value beta.Not fixed threshold β herein
Only acceptable threshold value beta need to be arranged according to practical application scene in specific value.May finally be expressed as [alg (S)=
E’se]
It should be noted that here and being not fixed the model of alg.That is alg (S) can be any one in machine learning
Kind, it can be any capable mathematical algorithm for indicating relationship between this SE state and coil end data.
It is by BP backpropagation, using S as input, and really by taking the building of convolutional neural networks CNN algorithm as an example
SE point is as true tag, constantly training and to obtain SE state by the Loss backpropagation Optimized model with true tag online
Enclose corresponding mathematical relationship in end data S.
For the other types of other deep learnings and machine learning, due to itself being routine techniques, only will here
Conventional machine learning is combined with coil end data, is not improved to algorithm model, therefore, according to specific machine
Mode of learning come obtain sample data and according to sample data training machine learn process no longer specifically describe herein.
S240, result output.
From the relational model of step S230: will arbitrarily sample obtained detection data and be input to the corresponding pass of SE state
It is the predicted intensity information that the SE state can be obtained in model, thus help to grasp the properties of relay, into
And obtain the case where relay contact is subjected to impact arcing and when contact is subjected to impact arcing.
Example IV
Example IV is carried out on the basis of embodiment one or/and embodiment two.In example IV, relational model
Using wavelet transformation model, i.e., the mathematical relationship of coil end data Yu SE state is obtained by way of establishing wavelet basis.
Specifically, by experiment method, sample data when the corresponding actual strength information of SE state, this process are obtained
Be in embodiment three it is similar, coil end data is acquired under the corresponding actual strength information of SE state, so
Coil end data when actual strength information that the SE state is corresponding is unfolded by Fourier transformation afterwards, by the line after expansion
Circle end data is denoted as sample expanding data;
Wavelet basis is established according to the sample expanding data;Wavelet basis has intensive properties and frequency, can by wavelet basis
To find the mathematical relationship of SE state and coil end data, therefore SE can be established by detecting the wavelet basis of SE status data section
The mathematical relationship of state and coil end data.
Embodiment five
The relay contact arcing state of strength detection method that embodiment five discloses a kind of corresponding above-described embodiment is corresponding
Device, be above-described embodiment virtual device structure, referring to figure 4. shown in, comprising:
Data acquisition module 310, for obtaining sample data, the sample data is to carry out repeatedly to relay coil end
Sampling, the coil end data of acquisition;
Model construction module 320, for obtaining the corresponding actual strength information of contact striking state of strength, by the contact
Arcing state of strength is denoted as SE state, according to the sample data and actual strength information architecture relational model, the relationship mould
Type is the mathematical relationship of coil end data and SE state;
As a result output module 330 are obtained for being predicted according to detection data and the relational model SE state
The predicted intensity information of the corresponding SE state of the detection data, the detection data is the coil obtained in SE state-detection
End data.
Preferably, the sample data and detection data are basic parameter or/and transformation parameter, and the underlying parameter is low
Frequency electric current Ilow, high-frequency current Ihigh, one of voltage U or a variety of, the transformation parameter is one or more basis
Parameter passes through the derivative parameter being calculated.The transformation parameter is coil equivalent impedance, coil transient state induction reactance, voltage change ratio
With one of current changing rate or a variety of.
Preferably, relational model uses machine learning model.The building relational model, including feature annotation step and mould
Type training step, in which:
The feature annotation step, comprising:
SE state is described;
Wherein, SE ' indicates the state description set in SE state, sehThe h next state represented in state description set is retouched
It states, the state description characterization impact arcing intensity, h is the sum of state description;
Experiment method is passed through to all state descriptions of state description set respectively and finds corresponding actual strength information;Its
I-th state description se in middle SE stateiCorresponding actual strength informationThe actual strength informationFor i-th shape
State describes corresponding real impact arcing intensity, 1≤i≤h;
The model training step, comprising:
Relational model is created, using sample data as input,Relational model is trained as true tag, is obtained
It is exported to predictionIt is expressed asWherein, S is sample data, and alg (S) is SE state and coil end data
Between relational model;
Calculate prediction outputAnd true tagBetween error, if the error be not more than preset threshold value,
Then the corresponding relational model of SE state is alg (S), conversely, then continuing to train, until prediction outputAnd true tag
Between error be not more than preset threshold value;
Finally obtain the relational model of SE state:
Alg (S)=t 'se
t’seFor using coil end data as the corresponding predicted intensity information of SE state when input.
Preferably, the relational model is wavelet transformation model;
The building relational model, comprising:
By experiment method, sample data when the corresponding actual strength information of SE state, the actual strength letter are obtained
Breath is the corresponding real impact arcing intensity of SE state, and by Fourier transformation, by the corresponding actual strength of the SE state
Sample data expansion when information, obtains sample expanding data;
Wavelet basis is established with the sample expanding data;
The mathematical relationship of SE state and coil end data is established by detecting wavelet basis.
Embodiment six
Fig. 5 is the structural schematic diagram for a kind of electronic equipment that the embodiment of the present invention six provides, as shown in figure 5, the electronics is set
Standby includes processor 410, memory 420, input unit 430 and output device 440;The number of processor 410 in computer equipment
It measures and can be one or more, in Fig. 5 by taking a processor 410 as an example;Processor 410, memory 420 in electronic equipment,
Input unit 430 can be connected with output device 440 by bus or other modes, in Fig. 5 for being connected by bus.
Memory 420 is used as a kind of computer readable storage medium, can be used for storing software program, journey can be performed in computer
Sequence and module, as in the embodiment of the present invention the corresponding program instruction/module of relay SE condition detection method (for example, after
Data acquisition module 310, model construction module 320 and result output module 330 in electric appliance SE condition detection method device).
Software program, instruction and the module that processor 410 is stored in memory 420 by operation, thereby executing electronic equipment
Various function application and data processing, i.e. the relay contact arcing state of strength of realization above-described embodiment one to example IV
Detection method.
Memory 420 can mainly include storing program area and storage data area, wherein storing program area can store operation system
Application program needed for system, at least one function;Storage data area, which can be stored, uses created data etc. according to terminal.This
Outside, memory 420 may include high-speed random access memory, can also include nonvolatile memory, for example, at least one
Disk memory, flush memory device or other non-volatile solid state memory parts.In some instances, memory 420 can be into one
Step includes the memory remotely located relative to processor 410, these remote memories can be set by network connection to electronics
It is standby.The example of above-mentioned network includes but is not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
Input unit 430 can be used for receiving subscriber identity information, sample data and detection data of input etc..Output dress
Setting 440 may include that display screen etc. shows equipment.
Embodiment seven
The embodiment of the present invention seven also provides a kind of storage medium comprising computer executable instructions, and the computer can be held
Row instruction by computer processor when being executed for executing relay contact arcing state of strength detection method, this method packet
It includes:
Sample data is obtained, the sample data is to carry out multiple repairing weld, the coil-end number of acquisition to relay coil end
According to;
The corresponding actual strength information of contact striking state of strength is obtained, the contact striking state of strength is denoted as SE shape
State, according to the sample data and actual strength information architecture relational model, the relational model is coil end data and SE shape
The mathematical relationship of state;
SE state is predicted according to detection data and the relational model, obtains the corresponding SE of the detection data
The predicted intensity information of state, the detection data are the coil end data obtained in SE state-detection.
Certainly, a kind of storage medium comprising computer executable instructions, computer provided by the embodiment of the present invention
The method operation that executable instruction is not limited to the described above can also be performed provided by any embodiment of the invention based on relay
Relevant operation in device contact striking state of strength detection method.
By the description above with respect to embodiment, it is apparent to those skilled in the art that, the present invention
It can be realized by software and required common hardware, naturally it is also possible to which by hardware realization, but in many cases, the former is more
Good embodiment.Based on this understanding, technical solution of the present invention substantially in other words contributes to the prior art
Part can be embodied in the form of software products, which can store in computer readable storage medium
In, floppy disk, read-only memory (Read-Only Memory, ROM), random access memory (Random such as computer
Access Memory, RAM), flash memory (FLASH), hard disk or CD etc., including some instructions use so that an electronic equipment
(can be mobile phone, personal computer, server or the network equipment etc.) executes method described in each embodiment of the present invention.
It is worth noting that, in the above-mentioned embodiment based on relay contact arcing state of strength detection method device, institute
Including each unit and module be only divided according to the functional logic, but be not limited to the above division, as long as energy
Enough realize corresponding function;In addition, the specific name of each functional unit is also only for convenience of distinguishing each other, it is not used to
It limits the scope of the invention.
It will be apparent to those skilled in the art that can make various other according to the above description of the technical scheme and ideas
Corresponding change and deformation, and all these changes and deformation all should belong to the protection scope of the claims in the present invention
Within.
Claims (9)
1. a kind of relay contact arcing state of strength detection method, which comprises the following steps:
Sample data is obtained, the sample data is to carry out multiple repairing weld, the coil end data of acquisition to relay coil end;
The corresponding actual strength information of contact striking state of strength is obtained, the contact striking state of strength is denoted as SE state,
According to the sample data and actual strength information architecture relational model, the relational model is coil end data and SE state
Mathematical relationship;
SE state is predicted according to detection data and the relational model, obtains the corresponding SE state of the detection data
Predicted intensity information, the detection data is the coil end data obtained in SE state-detection.
2. relay contact arcing state of strength detection method as described in claim 1, which is characterized in that the sample data
It is basic parameter or/and transformation parameter with detection data, the underlying parameter is low-frequency current Ilow, high-frequency current Ihigh, voltage
One of U or a variety of, the transformation parameter are that one or more underlying parameter passes through the derivative parameter being calculated.
3. relay contact arcing state of strength detection method as claimed in claim 2, which is characterized in that the transformation parameter
For one of coil equivalent impedance, coil transient state induction reactance, voltage change ratio and current changing rate or a variety of.
4. relay contact arcing state of strength detection method as described in claim 1, which is characterized in that the relational model
For machine learning model.
5. relay contact arcing state of strength detection method as claimed in claim 4, which is characterized in that the building relationship
Model, including feature annotation step and model training step, in which:
The feature annotation step, comprising:
SE state is described;
Wherein, SE ' indicates the state description set in SE state, sehRepresent the h next state description in state description set, institute
State description characterization impact arcing intensity is stated, h is the sum of state description;
Experiment method is passed through to all state descriptions of state description set respectively and finds corresponding actual strength information;Wherein SE
I-th state description se in stateiCorresponding actual strength informationThe actual strength informationIt is retouched for i-th state
State corresponding real impact arcing intensity, 1≤i≤h;
The model training step, comprising:
Relational model is created, using sample data as input,Relational model is trained as true tag, is obtained pre-
Survey outputIt is expressed asWherein, S is sample data, and alg (S) is between SE state and coil end data
Relational model;
Calculate prediction outputAnd true tagBetween error, if the error be not more than preset threshold value, SE shape
The corresponding relational model of state is alg (S), conversely, then continuing to train, until prediction outputAnd true tagBetween
Error is not more than preset threshold value;
Finally obtain the relational model of SE state:
Alg (S)=t 'se
t’seFor using coil end data as the corresponding predicted intensity information of SE state when input.
6. relay contact arcing state of strength detection method as described in claim 1, which is characterized in that the relational model
For wavelet transformation model;
The building relational model, comprising:
By experiment method, sample data when the corresponding actual strength information of SE state is obtained, the actual strength information is
The corresponding real impact arcing intensity of SE state, and by Fourier transformation, by the corresponding actual strength information of the SE state
When sample data expansion, obtain sample expanding data;
Wavelet basis is established with the sample expanding data;
The mathematical relationship of SE state and coil end data is established by detecting wavelet basis.
7. a kind of relay contact arcing state of strength detection device, characterized in that it comprises:
Data acquisition module, for obtaining sample data, the sample data is to carry out multiple repairing weld to relay coil end, is obtained
The coil end data obtained;
Model construction module obtains the corresponding actual strength information of contact striking state of strength, by the contact striking intensity shape
State is denoted as SE state, and according to the sample data and actual strength information architecture relational model, the relational model is coil-end
The mathematical relationship of data and SE state;
As a result output module obtains the inspection for predicting according to detection data and the relational model SE state
The predicted intensity information of the corresponding SE state of measured data, the detection data are the coil-end number obtained in SE state-detection
According to.
8. a kind of electronic equipment comprising processor, storage medium and computer program, the computer program are stored in
In storage media, which is characterized in that the computer program is realized as claimed in any one of claims 1 to 6 when being executed by processor
Relay contact arcing state of strength detection method.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
Processor realizes relay contact arcing state of strength detection method as claimed in any one of claims 1 to 6 when executing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910453959.3A CN110210107A (en) | 2019-05-28 | 2019-05-28 | Relay contact arcing state of strength detection method, device, equipment and medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910453959.3A CN110210107A (en) | 2019-05-28 | 2019-05-28 | Relay contact arcing state of strength detection method, device, equipment and medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110210107A true CN110210107A (en) | 2019-09-06 |
Family
ID=67789144
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910453959.3A Withdrawn CN110210107A (en) | 2019-05-28 | 2019-05-28 | Relay contact arcing state of strength detection method, device, equipment and medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110210107A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111169487A (en) * | 2020-02-19 | 2020-05-19 | 齐鲁工业大学 | Collector shoe arc-discharge early-warning intelligent measurement and control device, metro vehicle and control method thereof |
CN113687222A (en) * | 2021-08-24 | 2021-11-23 | 青岛理工大学 | SF (sulfur hexafluoride)6Method and system for evaluating state of arc contact of circuit breaker |
-
2019
- 2019-05-28 CN CN201910453959.3A patent/CN110210107A/en not_active Withdrawn
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111169487A (en) * | 2020-02-19 | 2020-05-19 | 齐鲁工业大学 | Collector shoe arc-discharge early-warning intelligent measurement and control device, metro vehicle and control method thereof |
CN113687222A (en) * | 2021-08-24 | 2021-11-23 | 青岛理工大学 | SF (sulfur hexafluoride)6Method and system for evaluating state of arc contact of circuit breaker |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Cheng et al. | A hybrid remaining useful life prognostic method for proton exchange membrane fuel cell | |
AU2016276256B2 (en) | Efficient battery tester | |
CN110943453B (en) | Power system fault sample generation and model construction method facing transfer learning | |
Li et al. | MagNet: An open-source database for data-driven magnetic core loss modeling | |
CN105467971B (en) | A kind of second power equipment monitoring system and method | |
CN105425768B (en) | A kind of second power equipment monitoring device and method | |
CN110210107A (en) | Relay contact arcing state of strength detection method, device, equipment and medium | |
CN111537884B (en) | Method and device for acquiring service life data of power battery, computer equipment and medium | |
CN108152676B (en) | Power distribution network fault positioning method and system based on intelligent agent | |
CN114636900A (en) | Power distribution network multiple fault diagnosis method and system | |
Zheng et al. | State of health estimation for lithium battery random charging process based on CNN-GRU method | |
CN108064391A (en) | A kind of method for quick predicting of battery cycle life and its fast prediction device | |
CN110059451A (en) | Relay contact arcing time condition detection method, device, equipment and medium | |
CN110532314A (en) | The method and terminal device of High-Voltage Electrical Appliances quality testing | |
CN110175411A (en) | Relay contact bounce status detection method, device, equipment and medium | |
CN106792985B (en) | A kind of mobile terminal access point Forecasting Methodology and device | |
CN106707118B (en) | Partial Discharge Pattern Recognition Method and device | |
Li et al. | Deep reinforcement learning for penetration testing of cyber-physical attacks in the smart grid | |
CN113239623B (en) | Fault positioning method suitable for electric power equipment | |
CN106021838A (en) | Residual life prediction method for complicated electronic system | |
CN110083991A (en) | The armature of relay and contact start condition detection method, equipment and the medium of touching | |
CN110209991A (en) | Armature of relay motion state detection method, apparatus, equipment and medium | |
CN112580211A (en) | Lead-acid storage battery SOH estimation method based on SA and ANN algorithm | |
CN113777500A (en) | Method for identifying battery parameters and optimizing precision by adopting simulated annealing method | |
CN104423532B (en) | power supply management method, device and system of mobile terminal |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190906 |
|
WW01 | Invention patent application withdrawn after publication |