CN109164382A - A kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults - Google Patents
A kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults Download PDFInfo
- Publication number
- CN109164382A CN109164382A CN201811053083.5A CN201811053083A CN109164382A CN 109164382 A CN109164382 A CN 109164382A CN 201811053083 A CN201811053083 A CN 201811053083A CN 109164382 A CN109164382 A CN 109164382A
- Authority
- CN
- China
- Prior art keywords
- parameter
- bat
- support vector
- vector machines
- contact
- 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
- 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
- G01R31/3271—Testing of circuit interrupters, switches or circuit-breakers of high voltage or medium voltage devices
- G01R31/3275—Fault detection or status indication
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Relating To Insulation (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The present invention relates to circuit breaker diagnosis fields, specifically disclose a kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults, comprising: obtain the contact ablation assessment parameter of breaker: resistance-stroke curve and static resistance value signal；The contact ablation status parameter values that parameter obtains breaker are assessed according to contact ablation；It uses the parameter of bat algorithm optimization support vector machines to obtain optimized parameter, optimal Nonlinear Support Vector Machines is established using optimized parameter；Set up sample data；Optimal Nonlinear Support Vector Machines are trained using sample data, parameter is assessed in input contact ablation, corresponding contact ablation status parameter values is exported, to obtain the non-linear support vector machines for being able to carry out assessment；It is predicted using contact ablation assessment parameter of the Nonlinear Support Vector Machines after training to breaker to be assessed.This method can precisely assess high-voltage circuitbreaker electric shock corrosion trouble.
Description
Technical field
The invention belongs to high-voltage circuitbreaker diagnostic field, in particular to a kind of High Voltage Circuit Breaker Contacts electroerosion fault diagnosis
Method.
Background technique
What the contact resistance or work generated in the motion process of the contact experience of electric switch generated in vibration environment
Contact resistance is presented as dynamic contact resistance, and characteristic is the true reflection and embodiment of electrical contact electrical surface contact state,
Therefore it can be used as the main foundation of electrical contact assessment.
For the contact of high-voltage circuitbreaker to being mainly formed in parallel by main contact and arcing contact, main contact carries nominal operation electricity
Stream, arcing contact bear electrical arc erosion.Arcing contact state is to influence the most important factor of high-voltage circuitbreaker electric life.China's electric power is set
Standby maintenance is in the transitional period from scheduled overhaul to repair based on condition of component, and the state-detection of arcing contact is arc-chutes repair based on condition of component
Pith.The electroerosion of arcing contact will cause the reduction that arc-chutes cut-off short circuit current ability and insulating capacity, extreme case
Under may cause arc-chutes drop-out current and unsuccessfully cause to explode, seriously threatened the reliability of electric system.Therefore, based on to arc
Contact state detection and the research of Endurance Prediction have most important theories meaning and Practical to stability of power system is improved
Value.
Breaker close or interrupting process in contact resistance between contact be also understood that and slide stroke for contact
Function.Contact resistance under the contact closure state of usual high-voltage circuitbreaker is in 10~20u Ω magnitude, to make surveyed contact pressure
Signal is dropped from interference, and there is good robustness, and the constant-current source that specification regulation applies in industry is direct current, and amplitude is many
In 1000A.Pass through the test to contact resistance, it can be estimated that contact ablation degree realizes the repair based on condition of component of breaker.
Such as application No. is 201710953217.8 patent documents to disclose a kind of breaker arc touching neural network based
Head ablation state evaluating method, it discloses carry out assessment judgement, but mind to arcing contact ablation state using neural network algorithm
It is too long through network algorithm learning time, it is inefficient, local minimum may be fallen into, the problems such as to cause accuracy not high.
Summary of the invention
The purpose of the present invention is to provide a kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults, after optimization
Support vector machines High Voltage Circuit Breaker Contacts electroerosion failure is precisely assessed.
To achieve the above object, the present invention provides a kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults, comprising:
S101 after multiple breakers are run a period of time under different voltage and currents respectively, acquires the dynamic of breaker
The dynamic contact resistance signal and static resistance value signal of arcing contact, obtain dynamic contact resistance-time graph；It acquires described dynamic
Dynamic stroke when dynamic contact resistance signal occurs for arcing contact, obtains stroke-time graph；
It is bent to obtain resistance-stroke according to the dynamic contact resistance-time graph and the stroke-time graph by S102
Line；
S103 assesses parameter using static resistance value signal and resistance-stroke curve as contact ablation, according to the contact
The contact ablation status parameter values that parameter obtains breaker are assessed in ablation；
S104 is used the parameter of bat algorithm optimization support vector machines to obtain optimized parameter, is established using optimized parameter
The optimal non-linear support vector machines；
S105, using the contact ablation of each breaker assess parameter and corresponding contact ablation status parameter values as
One group of sample data；
S106 is trained the optimal non-linear support vector machines using multiple groups sample data, inputs the touching
Parameter is assessed in head ablation, and the non-linear support vector machines exports corresponding contact ablation status parameter values, can be right with acquisition
The non-linear support vector machines that High Voltage Circuit Breaker Contacts ablation failure is assessed；
S107 is commented using the contact ablation that the non-linear support vector machines after training treats the breaker of diagnosis
Estimate parameter to be predicted, directly carries out ablation status assessment according to the contact ablation status parameter values of output.
Preferably, it in above-mentioned technical proposal, is specifically included in step 102:
The parameter of the support vector machines: the parameter model of punishment parameter C, RBF nuclear parameter δ, loss function ε is arranged in S201
It encloses；Initialize bat group's relevant parameter: setting initial population number of individuals n, pulse loudness A0, impulse ejection rate r0, bat pulse hair
Firing rate rate increases coefficient gamma, pulse loudness attenuation coefficient α, bat and searches for arteries and veins frequency limits fmin,fmax, maximum number of iterations
tmaxWith search precision ε；
S202 initializes the position x of batiWith speed vi；
S203 determines fitness evaluating function f (x), x=(x1,…xd)T, evaluated according to the fitness evaluating function every
The fitness value of a bat is to find current optimal solution x*；
S204 adjusts bat search pulse frequency, and speed and the position of bat are updated by formula (1), (2), (3):
fi=fmin+(fmax-fmin)β (1)
In formula: β is the uniform random number that [0,1] is randomly generated；fiIndicate the frequency of sound wave；X* indicates current global optimum
Solution；It indicates in the position of i-th bat of t moment,Indicate speed this moment；
S205 generates equally distributed random number rand, if rand > ri, then enter S206, otherwise enter S207,
In, riFor the impulse ejection rate of i-th bat；
S206 carries out random perturbation to current optimal solution to generate a new explanation, and carries out processing of crossing the border to the new explanation, i.e.,
A local solution is searched near the optimal solution currently selected, records current optimal solution；
S207 generates a new explanation by random flight, if rand < AiAnd f (xi) < f (x*), then enter S208, it is no
Then enter S209, wherein AiFor the pulse loudness of i-th bat；
S208 records this new explanation, and updates r using formula (4), (5)iWith Ai；
ri t+1=ri 0[1-exp(-γ*t)] (4)
In formula, ri t+1Indicate i-th bat in the impulse ejection rate in t+1 generation, ri 0Indicate the maximum impulse hair of i-th bat
Rate is penetrated, γ is that impulse ejection rate increases coefficient, wherein γ > 0,I-th bat is respectively indicated in the arteries and veins in t+1 and t generation
Loudness is rushed, a ∈ [0,1] is pulse loudness attenuation coefficient；
S209 is ranked up the fitness value of all bats in bat group, finds out current optimal solution and optimal suitable
Answer angle value；
S210 then goes to step S211, otherwise returns if meeting preset search precision or reaching maximum search number
Step S204；
S211 exports current globally optimal solution, based on currently choosing optimal Nonlinear Support Vector Machines model and its ginseng
Number.
Preferably, in above-mentioned technical proposal, step S103 is specifically included:
Sample data is divided into training sample set and test sample collection by S301；
The data of test sample collection and training sample set are normalized in S302；
S303, according to the training parameter of the step S211 optimized parameter setting support vector machines chosen, to training sample set
It is trained study, and test sample is trained with the support vector machines；
S304 obtains the prediction result of test sample collection.
Preferably, in above-mentioned technical proposal, based on currently choosing optimal Nonlinear Support Vector Machines mould in step S211
Type and its parameter include: training parameter, the type of model, kernel function type, loss function and its parameter.
Preferably, it in above-mentioned technical proposal, is connect using the dynamic of breaker dynamic contact resistance tester acquisition breaker
Touch resistance signal.
Preferably, in above-mentioned technical proposal, using the motion profile of stroke sensor measurement moving arcing contact.
Compared with prior art, High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults of the invention obtains height and breaks
Road device contact ablation data establish nonlinear compensation model using bat algorithm BA-SVM support vector regression, to different
The contact ablation status parameter values that high-voltage circuitbreaker obtains are measured under voltage and current precisely to be predicted.
Detailed description of the invention
Fig. 1 is the flow chart of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to the present invention.
Fig. 2 is BA-SVM algorithm flow chart according to the present invention.
Fig. 3 is the dynamic electric resistor and time waveform figure of the A phase of test breaker according to the present invention.
Fig. 4 is the dynamic electric resistor and moving contact stroke waveform diagram of the A phase of test breaker according to the present invention.
Fig. 5 is the dynamic electric resistor and time waveform figure of the C phase of test breaker according to the present invention.
Fig. 6 is the dynamic electric resistor and moving contact stroke waveform diagram of the C phase of test breaker according to the present invention.
Specific embodiment
With reference to the accompanying drawing, specific embodiments of the present invention will be described in detail, it is to be understood that guarantor of the invention
Shield range is not limited by the specific implementation.
As shown in Figure 1, the High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults in the embodiment, main thought are as follows: set
Contact ablation assessment parameter (resistance-stroke curve and Static Electro resistance value) of the breaker of acquisition is set as x, by contact ablation state
Parameter value is as target parameter y, it is clear that y=f (x) is non-linear relation, is supported using contact ablation assessment parameter x as BA-SVM
The input sample of vector machine model, output contact ablation status parameter values schedule to last after the processing of BA-SVM supporting vector machine model
Hope eliminate voltage and current influence after target parameter y, specifically, this method specifically includes the following steps:
Step S101 after multiple breakers are run a period of time under different voltage and currents respectively, acquires breaker
Moving arcing contact dynamic contact resistance signal and static resistance value signal, obtain dynamic contact resistance-time graph；Acquisition institute
Dynamic stroke when dynamic contact resistance signal occurs for moving arcing contact is stated, stroke-time graph is obtained；
Specifically, using adjustable voltage current source, utilizing super electricity when high-voltage circuitbreaker does acquisition dynamic electric resistor experiment
Holding and generates dash current, the dash current of super capacitor output can achieve 2500A, it measures by the way of on/off switch test,
The combined floodgate experiment for setting duration 250mS, the separating brake experiment of a length of 250mS, can be obtained electricity by voltage and current when then doing
Resistance, specific DB-8016 breaker dynamic contact resistance tester pass through voltage data, the electricity of collected contact resistance two sides
Flow data calculates the dynamic contact resistance signal and static resistance signal of contact resistance, then according to the pass of resistance and time
System, draws out the curve of dynamic contact resistance and time；Instrument can draw out the curve of voltage and time, electric current and time
Curve etc..
Step S102 obtains resistance-stroke curve according to dynamic contact resistance-time graph and stroke-time graph.
In the step, the motion profile of moving arcing contact can be measured simultaneously by stroke sensor tester, and be transmitted to DB-
8016 breaker dynamic contact resistance testers can draw the curve of stroke and time on liquid crystal display.According to dynamic Contact electricity
The curve of resistance and time, the curve of stroke and time derive the relationship of dynamic contact resistance and stroke, and can be on liquid crystal display
Draw the curve of contact resistance and moving arcing contact stroke.
Step S103 assesses parameter using static resistance value signal and resistance-stroke curve as contact ablation, according to contact
The contact ablation status parameter values that parameter obtains breaker are assessed in ablation.
In the step, contact ablation can be obtained according to expert survey and assess parameter (resistance-stroke curve and static resistance
Value) and contact ablation status parameter values between relationship be divided into three such as numerical value of the contact ablation status parameter values between 0-1
A segment, i.e. 0-0.4,0.4-0.6,0.6-1.0；First second interval be it is normal, third be it is undetermined, the 4th the 5th section is then
It is serious for ablation, it need to be replaced, such as such as:
1. resistance-stroke curve is normal and Static Electro resistance value is normal；Resistance-stroke curve is micro- abnormal and Static Electro resistance value
Normally；Resistance-stroke curve is micro- abnormal and the micro- exception of Static Electro resistance value；Resistance-stroke curve is normal and Static Electro resistance value is micro-
It is abnormal；Above situation is section 0-0.2.
2. resistance-stroke curve is normal and Static Electro resistance value moderate abnormality；Resistance-stroke curve moderate abnormality and it is static
Resistance value is normal；Resistance-stroke curve moderate abnormality and Static Electro resistance value moderate abnormality；Above situation is section 0.4-
0.6。
3. Static Electro resistance value severe abnormality within resistance-stroke curve moderate abnormality；Resistance-stroke curve severe is different
Often and within Static Electro resistance value moderate abnormality；Resistance-stroke curve severe abnormality and Static Electro resistance value severe abnormality；Above-mentioned feelings
Condition is section 0.6-1.0.
Step S104 uses the parameter of bat algorithm optimization support vector machines to obtain optimized parameter, using optimized parameter
Establish the optimal non-linear support vector machines；
Step S105 assesses parameter and corresponding contact ablation status parameter values with the contact ablation of each breaker
As one group of sample data；
Step S106 is trained the optimal non-linear support vector machines using multiple groups sample data, inputs institute
Contact ablation assessment parameter is stated, the non-linear support vector machines exports corresponding contact ablation status parameter values, can with acquisition
With the non-linear support vector machines assessed High Voltage Circuit Breaker Contacts ablation failure.
In step S106, specifically include:
Sample data is divided into training sample set and test sample collection, before randomly drawing sample data by step S301
90% is used as training set, and rear 10% is test set；
The data of test sample collection and training sample set are normalized in step S302；
Step S303 is trained training sample set according to the BA-SVM support vector machines that the optimized parameter of selection is arranged
Study, and test sample is trained with the support vector machines；
Step S304 obtains the prediction result of test sample collection, by test data to the pre- of SVM gamma correction model
Effect is surveyed to be assessed and analyzed.
Step S107 is burnt using the contact that the non-linear support vector machines after training treats the breaker of diagnosis
Erosion assessment parameter is predicted, directly carries out ablation status assessment according to the contact ablation status parameter values of output.
With continued reference to Fig. 2, in the embodiment, specifically included in step 102:
Step S201, the parameter of SVM support vector machines is arranged: the parameter area of punishment parameter C is [1,100], RBF core is joined
The range of number δ is [0.1,100], the parameter area of loss function ε is [0.001,1]；Initialize bat group's relevant parameter: setting
Initial population number of individuals n, pulse loudness A0, impulse ejection rate r0, bat impulse ejection rate increases coefficient gamma, pulse loudness and declines
Subtract factor alpha, bat search arteries and veins frequency limits fmin,fmax, maximum number of iterations tmaxWith search precision ε；
Step S202 initializes the position x of batiWith speed vi；
Step S203 determines fitness evaluating function f (x), x=(x1,…xd)T, commented according to the fitness evaluating function
The fitness value of each bat of valence is to find current optimal solution x*；
Step S204 adjusts bat search pulse frequency, and speed and the position of bat are updated by formula (1), (2), (3):
fi=fmin+(fmax-fmin)β (1)
In formula: β is the uniform random number that [0,1] is randomly generated；fiIndicate the frequency of sound wave；X* indicates current global optimum
Solution；It indicates in the position of i-th bat of t moment,Indicate speed this moment；
Step S205 generates equally distributed random number rand, if rand > ri, then enter S206, otherwise enter
S207, wherein riFor the impulse ejection rate of i-th bat；
Step S206 carries out random perturbation to current optimal solution to generate a new explanation, and carries out place of crossing the border to the new explanation
Reason searches for a local solution that is, near the optimal solution currently selected, records current optimal solution；
Step S207 generates a new explanation by random flight, if rand < AiAnd f (xi) < f (x*), then enter
Otherwise S208 enters S209, wherein AiFor the pulse loudness of i-th bat；
Step S208 records this new explanation, and updates r using formula (4), (5)iWith Ai, i.e., (increase ri, reduce Ai)；
ri t+1=ri 0[1-exp(-γ*t)] (4)
In formula, ri t+1Indicate i-th bat in the impulse ejection rate in t+1 generation, ri 0Indicate the maximum impulse hair of i-th bat
Rate is penetrated, γ is that impulse ejection rate increases coefficient, wherein γ > 0,I-th bat is respectively indicated in the arteries and veins in t+1 and t generation
Loudness is rushed, a ∈ [0,1] is pulse loudness attenuation coefficient；
Step S209 is ranked up the fitness value of all bats in bat group, finds out current optimal solution and most
Excellent fitness value；
Step S210 then goes to step S211, otherwise if meeting preset search precision or reaching maximum search number
Return step S204；
Step S211 exports current globally optimal solution, based on currently choose optimal Nonlinear Support Vector Machines model and
Its parameter, comprising: training parameter (including penalty factor, Radial basis kernel function parameter etc.), the type of model, kernel function type,
Loss function and its parameter.
Further, in step S102, resistance-is obtained according to dynamic contact resistance-time graph and stroke-time graph
Stroke curve is specific as follows:
Dynamic electric resistor experiment carried out to SF6 high-voltage circuitbreaker to certain substation, breaker A phase dynamic electric resistor and time,
The waveform of stroke is as shown in Figure 3, Figure 4:
By repeatedly measure the waveform comparison stablize, after dynamic electric resistor test, the dynamic of B phase are carried out to the B phase of breaker
Resistance is identical as the waveform of A phase.
When carrying out the test of breaker C phase dynamic electric resistor, the dynamic electric resistor measured and time, stroke waveform diagram such as Fig. 5, figure
Shown in 6:
Breaker closing time point 100mS, moving contact of breaker stablize the time about in 150mS.Afterwards repeatedly to breaker C
The test of phase dynamic electric resistor, waveform are as shown in Figure 5, Figure 6.
For visible break device in the combined floodgate excess of stroke stage, dynamic contact resistance fluctuation is larger.A in opposite Fig. 3, Fig. 4 is in contact
The smooth waveform of resistance, thus it is speculated that breaker C phase contact should have more serious ablation situation.
Data analysis analyzes software by Dynamic Resistance Tester PC, extract the moving contact stroke of A phase and C phase with contact
The data of resistance carry out data analysis.
Resistance-run-length data of above-mentioned A phase and C phase is made into data and curves, by comparison A, C phase data curve it is found that A
Phase dynamic electric resistor increases with the stroke of moving contact, dynamic electric resistor monotone decreasing, can be inferred that the dynamic/static contact contact of A phase is good
It is good, apparent ablation situation is not present.C phase dynamic electric resistor increases with the stroke of moving contact, and dynamic electric resistor bounce is larger, i.e. table
Bright A phase resistance-stroke curve is normal, C phase resistance-stroke curve severe abnormality, touches finally, forming in conjunction with static resistance value signal
PSO-SVM support vector regression after head ablation assessment parameter input optimization is trained.
High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults of the invention obtains High Voltage Circuit Breaker Contacts ablation data,
Nonlinear compensation model is established using bat algorithm BA-SVM support vector regression, to measuring high pressure under different voltage and currents
The contact ablation status parameter values that breaker obtains precisely are predicted.
The aforementioned description to specific exemplary embodiment of the invention is in order to illustrate and illustration purpose.These descriptions
It is not wishing to limit the invention to disclosed precise forms, and it will be apparent that according to the above instruction, can much be changed
And variation.The purpose of selecting and describing the exemplary embodiment is that explaining specific principle of the invention and its actually answering
With so that those skilled in the art can be realized and utilize a variety of different exemplary implementation schemes of the invention and
Various chooses and changes.The scope of the present invention is intended to be limited by claims and its equivalents.
Claims (6)
1. a kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults characterized by comprising
S101 after multiple breakers are run a period of time under different voltage and currents respectively, acquires the dynamic arc touching of breaker
The dynamic contact resistance signal and static resistance value signal of head, obtain dynamic contact resistance-time graph；Acquire the dynamic arc touching
Dynamic stroke when hair life dynamic contact resistance signal, obtains stroke-time graph；
S102 obtains resistance-stroke curve according to the dynamic contact resistance-time graph and the stroke-time graph；
S103 assesses parameter using static resistance value signal and resistance-stroke curve as contact ablation, according to the contact ablation
Assess the contact ablation status parameter values that parameter obtains breaker；
S104 is used the parameter of bat algorithm optimization support vector machines to obtain optimized parameter, is established using optimized parameter optimal
The non-linear support vector machines；
S105 assesses parameter and corresponding contact ablation status parameter values as one group using the contact ablation of each breaker
Sample data；
S106 is trained the optimal non-linear support vector machines using multiple groups sample data, inputs the contact and burns
Erosion assessment parameter, the non-linear support vector machines export corresponding contact ablation status parameter values, can be to high pressure with acquisition
The non-linear support vector machines that contact of breaker ablation failure is assessed；
S107 treats the contact ablation assessment ginseng of the breaker of diagnosis using the non-linear support vector machines after training
Number is predicted, directly carries out ablation status assessment according to the contact ablation status parameter values of output.
2. High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to claim 1, which is characterized in that step 102
In specifically include:
The parameter of the support vector machines: the parameter area of punishment parameter C, RBF nuclear parameter δ, loss function ε is arranged in S201；Just
Beginningization bat group's relevant parameter: setting initial population number of individuals n, pulse loudness A0, impulse ejection rate r0, bat impulse ejection speed
Rate increases coefficient gamma, pulse loudness attenuation coefficient α, bat and searches for arteries and veins frequency limits fmin,fmax, maximum number of iterations tmaxWith
Search precision ε；
S202 initializes the position x of batiWith speed vi；
S203 determines fitness evaluating function f (x), x=(x1,…xd)T, each bat is evaluated according to the fitness evaluating function
The fitness value of bat is to find current optimal solution x*；
S204 adjusts bat search pulse frequency, and speed and the position of bat are updated by formula (1), (2), (3):
fi=fmin+(fmax-fmin)β (1)
In formula: β is the uniform random number that [0,1] is randomly generated；fiIndicate the frequency of sound wave；x*Indicate current globally optimal solution；
It indicates in the position of i-th bat of t moment,Indicate speed this moment；
S205 generates equally distributed random number rand, if rand > ri, then enter S206, otherwise enter S207, wherein ri
For the impulse ejection rate of i-th bat；
S206 carries out random perturbation to current optimal solution to generate a new explanation, and carries out processing of crossing the border to the new explanation, that is, is working as
The optimal solution of preceding selection nearby searches for a local solution, records current optimal solution；
S207 generates a new explanation by random flight, if rand < AiAnd f (xi) < f (x*), then enter S208, otherwise into
Enter S209, wherein AiFor the pulse loudness of i-th bat；
S208 records this new explanation, and updates r using formula (4), (5)iWith Ai；
ri t+1=ri 0[1-exp(-γ*t)] (4)
In formula, ri t+1Indicate i-th bat in the impulse ejection rate in t+1 generation, ri 0Indicate the maximum impulse transmitting of i-th bat
Rate, γ are that impulse ejection rate increases coefficient, wherein γ > 0,I-th bat is respectively indicated in the pulse in t+1 and t generation
Loudness, a ∈ [0,1] are pulse loudness attenuation coefficient；
S209 is ranked up the fitness value of all bats in bat group, finds out current optimal solution and adaptive optimal control degree
Value；
S210 then goes to step S211, otherwise return step if meeting preset search precision or reaching maximum search number
S204；
S211 exports current globally optimal solution, based on currently choosing optimal Nonlinear Support Vector Machines model and its parameter.
3. High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to claim 2, which is characterized in that step S103
It specifically includes:
Sample data is divided into training sample set and test sample collection by S301；
The data of test sample collection and training sample set are normalized in S302；
S303 carries out training sample set according to the training parameter of the step S211 optimized parameter setting support vector machines chosen
Training study, and test sample is trained with the support vector machines；
S304 obtains the prediction result of test sample collection.
4. High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to claim 2, which is characterized in that step S211
In based on optimal Nonlinear Support Vector Machines model and its parameter is currently chosen include: training parameter, the type of model, core letter
Several classes of types, loss function and its parameter.
5. High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to claim 1, which is characterized in that using open circuit
The dynamic contact resistance signal of device dynamic contact resistance tester acquisition breaker.
6. High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults according to claim 1, which is characterized in that use stroke
The motion profile of sensor measurement moving arcing contact.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811053083.5A CN109164382B (en) | 2018-09-10 | 2018-09-10 | Method for diagnosing electrical erosion fault of high-voltage circuit breaker contact |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811053083.5A CN109164382B (en) | 2018-09-10 | 2018-09-10 | Method for diagnosing electrical erosion fault of high-voltage circuit breaker contact |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109164382A true CN109164382A (en) | 2019-01-08 |
| CN109164382B CN109164382B (en) | 2021-01-05 |
Family
ID=64894661
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811053083.5A Active CN109164382B (en) | 2018-09-10 | 2018-09-10 | Method for diagnosing electrical erosion fault of high-voltage circuit breaker contact |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109164382B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110412373A (en) * | 2019-07-23 | 2019-11-05 | 安徽升隆电气有限公司 | A kind of switchgear fault early warning system and its replacing options |
| CN111505490A (en) * | 2020-03-23 | 2020-08-07 | 温州大学乐清工业研究院 | AC contactor ablation condition evaluation method based on convolutional neural network regression |
| CN112084662A (en) * | 2020-09-11 | 2020-12-15 | 西安高压电器研究院有限责任公司 | Method and device for detecting electrical service life of circuit breaker |
| RU2807606C1 (en) * | 2023-04-25 | 2023-11-17 | Алексей Михайлович Москалёв | Method for assessing electrical erosion wear resistance of material of electrical contact parts |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130035885A1 (en) * | 2011-08-04 | 2013-02-07 | Massachusetts Institute Of Technology | Topology identification in distribution network with limited measurements |
| CN103336243A (en) * | 2013-07-01 | 2013-10-02 | 东南大学 | Breaker fault diagnosis method based on separating/closing coil current signals |
| CN103439650A (en) * | 2013-08-07 | 2013-12-11 | 王岩 | Method and device used for state monitoring and fault diagnosis of relay |
| CN103575525A (en) * | 2013-11-18 | 2014-02-12 | 东南大学 | Intelligent diagnosis method for mechanical fault of circuit breaker |
| CN103616635A (en) * | 2013-12-05 | 2014-03-05 | 国家电网公司 | Method and device for diagnosing mechanical characteristic failures of high-voltage circuit-breaker |
| CN104764993A (en) * | 2014-01-07 | 2015-07-08 | 国家电网公司 | Detection method and device for high-voltage circuit breaker |
| CN104793134A (en) * | 2015-04-29 | 2015-07-22 | 中国电力科学研究院 | Breaker operating mechanism fault diagnosis method based on least square support vector machine |
| CN105467309A (en) * | 2015-12-01 | 2016-04-06 | 天津市电力科技发展公司 | State evaluation method and maintenance strategy for contact of high-voltage circuit breaker |
| CN105548867A (en) * | 2015-12-01 | 2016-05-04 | 天津市电力科技发展公司 | Diagnostic system and diagnostic method of contact state of high-voltage circuit breaker |
| CN106482937A (en) * | 2016-09-30 | 2017-03-08 | 南方电网科学研究院有限责任公司 | A kind of monitoring method of mechanical state of high-voltage circuit breaker |
| CN106556796A (en) * | 2016-10-21 | 2017-04-05 | 中国电力科学研究院 | A kind of SF6 choppers arcing contact abatement detecting method |
| CN107680835A (en) * | 2017-10-13 | 2018-02-09 | 中国电力科学研究院 | A kind of breaker arcing contact ablation state evaluating method based on neutral net |
| CN108121999A (en) * | 2017-12-10 | 2018-06-05 | 北京工业大学 | Support vector machines parameter selection method based on mixing bat algorithm |
-
2018
- 2018-09-10 CN CN201811053083.5A patent/CN109164382B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130035885A1 (en) * | 2011-08-04 | 2013-02-07 | Massachusetts Institute Of Technology | Topology identification in distribution network with limited measurements |
| CN103336243A (en) * | 2013-07-01 | 2013-10-02 | 东南大学 | Breaker fault diagnosis method based on separating/closing coil current signals |
| CN103439650A (en) * | 2013-08-07 | 2013-12-11 | 王岩 | Method and device used for state monitoring and fault diagnosis of relay |
| CN103575525A (en) * | 2013-11-18 | 2014-02-12 | 东南大学 | Intelligent diagnosis method for mechanical fault of circuit breaker |
| CN103616635A (en) * | 2013-12-05 | 2014-03-05 | 国家电网公司 | Method and device for diagnosing mechanical characteristic failures of high-voltage circuit-breaker |
| CN104764993A (en) * | 2014-01-07 | 2015-07-08 | 国家电网公司 | Detection method and device for high-voltage circuit breaker |
| CN104793134A (en) * | 2015-04-29 | 2015-07-22 | 中国电力科学研究院 | Breaker operating mechanism fault diagnosis method based on least square support vector machine |
| CN105467309A (en) * | 2015-12-01 | 2016-04-06 | 天津市电力科技发展公司 | State evaluation method and maintenance strategy for contact of high-voltage circuit breaker |
| CN105548867A (en) * | 2015-12-01 | 2016-05-04 | 天津市电力科技发展公司 | Diagnostic system and diagnostic method of contact state of high-voltage circuit breaker |
| CN106482937A (en) * | 2016-09-30 | 2017-03-08 | 南方电网科学研究院有限责任公司 | A kind of monitoring method of mechanical state of high-voltage circuit breaker |
| CN106556796A (en) * | 2016-10-21 | 2017-04-05 | 中国电力科学研究院 | A kind of SF6 choppers arcing contact abatement detecting method |
| CN107680835A (en) * | 2017-10-13 | 2018-02-09 | 中国电力科学研究院 | A kind of breaker arcing contact ablation state evaluating method based on neutral net |
| CN108121999A (en) * | 2017-12-10 | 2018-06-05 | 北京工业大学 | Support vector machines parameter selection method based on mixing bat algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 魏峻: "基于蝙蝠算法的支持向量机参数优化", 《宝鸡文理学院学报（自然科学版）》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110412373A (en) * | 2019-07-23 | 2019-11-05 | 安徽升隆电气有限公司 | A kind of switchgear fault early warning system and its replacing options |
| CN110412373B (en) * | 2019-07-23 | 2021-08-27 | 安徽升隆电气有限公司 | Switch cabinet fault early warning system and replacement method thereof |
| CN111505490A (en) * | 2020-03-23 | 2020-08-07 | 温州大学乐清工业研究院 | AC contactor ablation condition evaluation method based on convolutional neural network regression |
| CN112084662A (en) * | 2020-09-11 | 2020-12-15 | 西安高压电器研究院有限责任公司 | Method and device for detecting electrical service life of circuit breaker |
| RU2807606C1 (en) * | 2023-04-25 | 2023-11-17 | Алексей Михайлович Москалёв | Method for assessing electrical erosion wear resistance of material of electrical contact parts |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109164382B (en) | 2021-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Qu et al. | Series arc fault detection of indoor power distribution system based on LVQ-NN and PSO-SVM | |
| CN109061462A (en) | A kind of High Voltage Circuit Breaker Contacts ablation assessment of failure method | |
| CN109164382A (en) | A kind of High Voltage Circuit Breaker Contacts electroerosion method for diagnosing faults | |
| Khoshkhoo et al. | On-line dynamic voltage instability prediction based on decision tree supported by a wide-area measurement system | |
| CN103915818A (en) | Series connection fault electric arc detecting method and special device thereof | |
| CN106443435A (en) | Circuit breaker breaking and closing operation simulation test device and test method | |
| CN106771993A (en) | A kind of circuit-breaker switching on-off action simulation experimental rig and test method | |
| Humbert et al. | Serial-arc detection by use of Spectral Dispersion Index (SDI) analysis in a low-voltage network (270V HVDC) | |
| CN110231581A (en) | Accumulator internal resistance measuring circuit and measurement method | |
| Yao et al. | DC arc fault: Characteristic study and fault recognition | |
| CN206431254U (en) | A kind of circuit-breaker switching on-off action simulation experimental rig | |
| Golkar et al. | MATLAB modeling of arc furnace for flicker study | |
| CN206147060U (en) | Circuit breaker divide -shut brake action analogue test device | |
| CN105403799B (en) | Electrostatic transducer analysis method for reliability and system | |
| CN113094983A (en) | Online simulation method for multi-dimensional time-varying characteristics of direct-current fault electric arc of photovoltaic system | |
| CN116742649A (en) | Reactive compensation linear zero-crossing detection circuit and dynamic adjustment zero-crossing switching method | |
| Strobl et al. | Resonant electric arcs in DC microgrids with low system impedance in the VLF-band | |
| CN202057754U (en) | Man-machine interaction type life test and assessment system for fuse | |
| JPH10332758A (en) | Impulse tester | |
| Strobl et al. | Refined fault detection in LVDC-grids with signal processing, system identification and machine learning methods | |
| US11338302B1 (en) | Apparatus and test method for simulating spark discharge of high-voltage electrostatic precipitator | |
| CN105629105B (en) | electrostatic transducer reliability test method and system | |
| CN107269475A (en) | Simulation load test system and method of wind generating set | |
| Chang et al. | Modeling voltage-current characteristics of an electric arc furnace based on actual recorded data: A comparison of classic and advanced models | |
| CN105842618B (en) | Magnetic switch comprehensive tester with process tracking protection and testing method thereof |
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 |