CN105469156A - MOA condition management and fault prediction method and MOA condition management and fault prediction system - Google Patents
MOA condition management and fault prediction method and MOA condition management and fault prediction system Download PDFInfo
- Publication number
- CN105469156A CN105469156A CN201410460875.XA CN201410460875A CN105469156A CN 105469156 A CN105469156 A CN 105469156A CN 201410460875 A CN201410460875 A CN 201410460875A CN 105469156 A CN105469156 A CN 105469156A
- Authority
- CN
- China
- Prior art keywords
- moa
- failure prediction
- health control
- health
- knowledge base
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an MOA condition management and fault prediction method and an MOA condition management and fault prediction system. The method comprises that an MOA condition management and fault prediction model is established, and takes MOA characteristic parameter data as input and the MOA condition as output; operation data related to different MOA conditions is calculated, and a training knowledge base is established based on the calculated operation data; the MOA condition management and fault prediction model is trained by utilizing data stored in the training knowledge base; and the MOA characteristic parameter data measured in real time is input to the trained MOA condition management and fault prediction model to obtain the MOA condition. According to the technical scheme, the operation data that reflects the different MOA conditions is calculated to train the MOA condition management and fault prediction model, and then the real-time MOA characteristic parameter data is input to the trained MOA condition management and fault prediction model to obtain the accurate MOA condition, and basis is provided for MOA maintenance.
Description
Technical field
The present invention relates to field of system management, particularly, relate to a kind of MOA health control and failure prediction method and system.
Background technology
Zinc-Oxide Arrester (being called for short MOA) is the important proterctive equipment ensureing that high-tension electricity device endangers from thunder and lightning and switching overvoltage, and its safe and reliable operation is very important to electric system, has been widely used at present in China 10 ~ 500kV electrical network.MOA is owing to bearing power-frequency voltage for a long time, and the impact of the factor such as surge voltage and internal wetted, causes the deterioration of inner valve block, current in resistance property increases, and power consumption increases, and causes the inner valve block temperature of MOA to raise, even produce thermal runaway, cause the blast of MOA, thus jeopardize the safe operation of electric system.
Along with the development of the technology such as electronic technology, sensor technology, optical fiber technology, computer technology, information processing, on-line monitoring and fault diagonosing technology is made to enter practical stage.
Therefore, it is desirable to carry out health control and failure prediction so that monitor MOA running status to MOA.
Summary of the invention
The object of this invention is to provide a kind of method and system, so that effectively health control and failure prediction can be carried out to MOA, provide effective foundation for MOA safeguards.
To achieve these goals, the invention provides a kind of Zinc-Oxide Arrester MOA health control and failure prediction method, the method comprises: set up MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting; Add up the service data relevant to the health status of various MOA, and the service data of Corpus--based Method builds training knowledge base; Utilize the data stored in described training knowledge base to described MOA health control and the training of failure prediction model; And the MOA health control that the MOA characteristic parameter data measured in real time input trained and failure prediction model, obtain the health status of MOA.
Further, described MOA health control and failure prediction model comprise: MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module, the filthy macrostate reasoning module of MOA.
Further, the method comprises each State reasoning module comprised corresponding to described MOA health control and failure prediction model, and structure MOA fault training knowledge base, the aging training knowledge base of MOA, MOA make moist and train knowledge base, MOA filth training knowledge base respectively; And the training knowledge base constructed by utilizing is trained each State reasoning module, accordingly to complete the training to described MOA health control and failure prediction model.
Further, the step of the described MOA health control the MOA characteristic parameter data measured in real time input trained and failure prediction model comprises: based on the MOA characteristic parameter data construct real-time testing knowledge base measured in real time, and this real-time testing knowledge base comprises MOA fault test knowledge base, MOA burn-in test knowledge base, MOA make moist knowledge on testing storehouse, the filthy knowledge on testing storehouse of MOA; And the MOA characteristic parameter data in each knowledge on testing storehouse are input to each State reasoning module accordingly as input.
Further, described MOA health control and failure prediction model also comprise final reasoning module, and the method comprises the health status of the judged result output MOA being gathered each State reasoning module by this final reasoning module.
Further, the health status of described MOA comprises following health indicator: MOA fault, MOA make moist, MOA is aging, MOA is filthy, and wherein each health indicator comprises multiple grade.
Further, the method also comprises use LIBSVM to set up MOA health control and failure prediction model.
Further, the method also comprise use multiple LIBSVM wrap each module set up accordingly in MOA health control and failure prediction model.
Another aspect of the present invention additionally provides a kind of Zinc-Oxide Arrester MOA health control and failure prediction system, and this system comprises: MOA characteristic parameter extraction unit, is configured to measure MOA characteristic parameter data in real time; Real-time testing knowledge base, is configured to the MOA characteristic parameter data storing the measurement of described MOA characteristic parameter extraction unit; Training knowledge base, is configured to the service data relevant to the health status of various MOA storing statistics; MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting, the data that this MOA health control and failure prediction model are configured to according to storing in described training knowledge base are trained; And after having trained, this MOA health control and failure prediction model are configured to the health status according to the data acquisition MOA stored in described real-time testing knowledge base.
Further, described MOA health control and failure prediction model comprise: MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module, the filthy macrostate reasoning module of MOA and final reasoning module, and the one wherein in the following health indicator that comprises the health status of MOA respectively of each State reasoning module judges: MOA fault, MOA make moist, MOA is aging, MOA is filthy; And described final reasoning module is configured to the judged result gathering each State reasoning module, to export the health status of MOA.
Pass through technique scheme, utilize the service data of the health status of the various MOA of reflection of statistics to MOA health control and failure prediction model training, the MOA the health control then input of real-time MOA characteristic parameter data trained and failure prediction model, the health status of accurate MOA can be obtained, provide foundation for MOA safeguards.
Other features and advantages of the present invention are described in detail in embodiment part subsequently.
Accompanying drawing explanation
Accompanying drawing is used to provide a further understanding of the present invention, and forms a part for instructions, is used from explanation the present invention, but is not construed as limiting the invention with embodiment one below.In the accompanying drawings:
Fig. 1 is MOA health control according to embodiment of the present invention and failure prediction method process flow diagram;
Fig. 2 is MOA health control according to the preferred embodiment of the present invention and failure prediction method process flow diagram;
Fig. 3 trains knowledge base structure schematic diagram according to the MOA of embodiment of the present invention;
Fig. 4 is the MOA knowledge on testing library structure schematic diagram according to embodiment of the present invention;
Fig. 5 A-5D is training, the test of each macrostate analysis ratiocination module according to embodiment of the present invention and exports schematic diagram;
Fig. 6 is the input and output schematic diagram of MOA health control according to embodiment of the present invention and failure prediction model; And
Fig. 7 is MOA health control according to embodiment of the present invention and failure prediction system structural representation.
Embodiment
Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described in detail.Adopt based on distributed support vector machine in embodiments of the present invention, with the example that is applied as of LIBSVM software package, principle of the present invention is described.Should be understood that, embodiment described herein, only for instruction and explanation of the present invention, is not limited to the present invention, and such as principle provided by the present invention also can be realized by artificial neural network.
Fig. 1 is MOA health control according to embodiment of the present invention and failure prediction method process flow diagram.As shown in Figure 1, a kind of MOA health control provided by the invention and failure prediction method, the method can comprise: S100, sets up MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting; S102, adds up the service data relevant to the health status of various MOA, and the service data of Corpus--based Method builds training knowledge base; S104, utilizes the data stored in described training knowledge base to described MOA health control and the training of failure prediction model; And S106, the MOA health control trained the MOA characteristic parameter data measured in real time input and failure prediction model, obtain the health status of MOA.
Pass through technique scheme, utilize the service data of the health status of the various MOA of reflection of statistics to MOA health control and failure prediction model training, the MOA the health control then input of real-time MOA characteristic parameter data trained and failure prediction model, the health status of accurate MOA can be obtained, provide foundation for MOA safeguards.
Fig. 2 is MOA health control according to the preferred embodiment of the present invention and failure prediction method process flow diagram.Clearly show in fig. 2 and how to utilize data in training knowledge base to MOA health control and the training of failure prediction model, thus to determine the model parameter in MOA health control and failure prediction model.After model parameter is determined, the MOA characteristic parameter that may be used for real-time is carried out model calculation by this MOA health control and failure prediction model, and expection will obtain MOA failure prediction result accurately.
In embodiments, MOA health control and failure prediction model can comprise multiple reasoning module, and such as MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module and the filthy macrostate reasoning module of MOA.The following health indicator that the division of each reasoning module above-mentioned or setting can comprise according to the health status of MOA: MOA fault, MOA make moist, MOA is aging, MOA is filthy.Although utilize a LIBSVM software package just can realize the judgement of each health indicator to above-mentioned MOA, but more preferably utilize multiple LIBSVM software package to judge each health indicator in a distributed fashion, can improve operation efficiency like this, the method that embodiment of the present invention is provided is more applicable for real-time working condition.Wherein, each health indicator can comprise multiple grade, such as, can comprise for fault indices: non-fault, minor failure, fault, inefficacy.Predicting the outcome specifically can be provided like this for plant maintenance personnel.
Wherein, utilizing LIBSVM software package to set up the process of four MOA State reasoning submodules respectively can be for: 1) prepare training and testing data set (or knowledge base) according to the form required by LIBSVM software package;
2) characteristic processing (comprising data normalization process) is carried out to data;
3) Svmtrain function is utilized to train modeling:
svmtrain[options]training_set_file[model_file]
Select SVM type to be C-SVC, select kernel function to be RBF;
4) cross validation is adopted to select optimal parameter C and g;
5) optimal parameter C and g is adopted to carry out whole training set training four the support vector machine reasoning submodules obtained;
6) function Svmpredict is utilized to predict the model trained:
svmpredicttest_filemodel_fileoutput_file
Model_file is the model file produced by svmtrain;
Test_file is the data file will carrying out predicting;
Output_file is the output file of svmpredict.
The structure in knowledge base and MOA knowledge on testing storehouse is trained to be described below in conjunction with Fig. 3 and Fig. 4 to according to the MOA of embodiment of the present invention.
In order to improve the efficiency of health control and failure prediction further, each reasoning module can be corresponded to training knowledge base and knowledge on testing storehouse are set.
In embodiments, method provided by the invention can comprise each State reasoning module comprised corresponding to described MOA health control and failure prediction model, and structure MOA fault training knowledge base, the aging training knowledge base of MOA, MOA make moist and train knowledge base, MOA filth training knowledge base respectively.Training knowledge base constructed by utilization can be trained each State reasoning module accordingly, to complete the training to described MOA health control and failure prediction model.In a preferred embodiment, the knowledge base of each grade be divided into corresponding to multiple grade can be continued corresponding to the training knowledge base of each health indicator.The mode of such refinement can make to the training of MOA health control and failure prediction model effectively, can for certain grade of each health indicator to MOA health control and the training of failure prediction model.
In embodiments, this real-time testing knowledge base comprises MOA fault test knowledge base, MOA burn-in test knowledge base, MOA make moist knowledge on testing storehouse, the filthy knowledge on testing storehouse of MOA.Corresponding with input characteristic parameter data step, the MOA health control MOA characteristic parameter data measured in real time input trained in method provided by the invention and the step of failure prediction model can comprise: based on the MOA characteristic parameter data construct real-time testing knowledge base measured in real time; And the MOA characteristic parameter data in each knowledge on testing storehouse are input to each State reasoning module accordingly as input.Similar with effect above, real-time testing knowledge base is classified according to health indicator, can forecasting efficiency be improved.
The training of each macrostate analysis ratiocination module, test and export as shown in figs. 5 a-5d.Such as, as shown in Figure 5A, MOA fault is utilized to train knowledge base to train MOA fault macrostate reasoning module, the data of the MOA fault macrostate reasoning module after then can utilizing training and MOA fault real-time testing knowledge base judge fault health indicator, obtain one of following judged result: non-fault, minor failure, catastrophic failure, inefficacy.
In embodiments, MOA health control and failure prediction model can also comprise final reasoning module, export the health status of MOA with the judged result being gathered each State reasoning module by this final reasoning module.
It should be noted that, in embodiments of the present invention, as an example, use LIBSVM to set up MOA health control and failure prediction model.Preferably, multiple LIBSVM can be used to wrap each module set up accordingly in MOA health control and failure prediction model.
Be described according to the input and output of the MOA health control of embodiment of the present invention and the training process of failure prediction model below in conjunction with Fig. 6, utilize the failure prediction process of real-time measuring data similar.
In embodiments, establish four classes training knowledge bases, be respectively the training of MOA fault knowledge base, MOA and make moist training knowledge base, the aging training knowledge base of MOA, MOA filth training knowledge base.Be described for each knowledge base respectively below with reference to Fig. 6.
(1) MOA fault training knowledge base, the reflection MOA fault characteristic value deposited has:
The Wavelet Energy Spectrum FACTOR P in leakage current circuital 2 cycles
1, P
2
The first-harmonic effective value I in 2 cycles of leakage current
1, I
2
The first-harmonic peak I in 2 cycles of leakage current
max1, I
max2,
The first-harmonic effective value I in 2 cycles of leakage current resistive component
r1, I
r2
Leakage current resistive component peak value: positive peak Ir+, negative peak Ir-
The effective value I in 2 cycles of leakage current resistive component 3 subharmonic
r31, Ir
32
Namely the input vector X1 of MOA fault macrostate LIBSVM reasoning submodule is:
X1＝[P
1P
2I
1I
2I
max1I
max2I
r1I
r2Ir+Ir-I
r31Ir
32]
To should input vector X1, MOA fault macrostate LIBSVM reasoning submodule output state being: non-fault, minor failure, fault, inefficacy; The corresponding MOA fault training knowledge base of this submodule and fault test knowledge base.Corresponding four kinds of malfunctions, good MOA fault training knowledge base prepared in advance can train this submodule in embodiments of the present invention, carry out fault test analysis with the on-the-spot real time data of preparation for acquiring, as shown in table 1.
Table 1
(2) MOA makes moist training knowledge base, and the reflection MOA the deposited characteristic quantity that makes moist has:
The first-harmonic effective value I in 2 cycles of leakage current resistive component
r1, I
r2
The effective value I in 2 cycles of leakage current resistive component 3 subharmonic
r31, Ir
32
The real-time working humidity S of MOA valve block
The m-Acetyl chlorophosphonazo content U1% in 2 cycles of voltage, U2%
MOA Temperature Distribution coefficient F
Namely the make moist input vector X2 of macrostate LIBSVM reasoning submodule of MOA is:
X2＝[I
r1I
r2I
r31Ir
32SU1％U2％]
Corresponding input vector X2, the MOA macrostate LIBSVM reasoning submodule output state that makes moist can be: do not make moist, slightly make moist, moderate is made moist, seriously make moist; The corresponding MOA of this submodule makes moist and trains knowledge base and the knowledge on testing storehouse that makes moist.In embodiments of the present invention can corresponding four kinds of states of making moist, good MOA prepared in advance training knowledge base of making moist trains this submodule, so that collection site real time data carries out making moist test analysis, as shown in table 2.
Table 2
(3) the aging training knowledge base of MOA, the reflection MOA aging characteristics deposited has:
5, the 7 subharmonic effective value I in 2 cycles of leakage current resistive component
r51, Ir
52, I
r71, Ir
72number of lightning strokes L
Sound signal singular entropy J during electric discharge
Fundamental harmonic resistant leakage current U-I correlation coefficient G
MOA valve block work temperature
Insulation loss angle θ
Lightning arrester power consumption P
Capacity current 2 cycle first-harmonic effective value I
c1, Ic
2
Full voltage phase angle difference φ
Total current phase angle difference ψ
Voltage 2 cycle effective value U
1, U
2
Namely the input vector of MOA aging macrostate LIBSVM reasoning submodule is X3:
X3＝[I
r51，Ir
52I
r71Ir
72LGTθPI
C1I
C2φψU1U2]
To should input vector X3, MOA aging macrostate LIBSVM reasoning submodule output state being: unaged, slightly aging, mittlere alterung, serious aging; The corresponding aging training knowledge base of MOA of this submodule and burn-in test knowledge base.In embodiments of the present invention can corresponding four kinds of ageing states, the aging training knowledge base of good MOA prepared in advance trains this submodule, carries out burn-in test analysis with the on-the-spot real time data of preparation for acquiring, as shown in table 3.
Table 3
(4) MOA filth training knowledge base, the filthy characteristic quantity of the reflection MOA deposited has:
The first-harmonic effective value I in 2 cycles of leakage current resistive component
r1, I
r2
The effective value I in 2 cycles of leakage current resistive component 3 subharmonic
r31, Ir
32
Insulation loss angle θ
MOA surface salt density p
1
MOA surface gray density ρ
2
MOA surface conductivity λ
Pollution flashover voltage effective value U
w
Namely the input vector of MOA filthy macrostate LIBSVM reasoning submodule is X4:
X4＝[I
r1，Ir
2I
r31Ir
32θρ
1ρ
2λU
W]
To should input vector X4, MOA filthy macrostate LIBSVM reasoning submodule output state being: by filthy, slight filthy, moderate is filthy, pollution severity; This submodule corresponding MOA filth training knowledge base and the filthy knowledge on testing storehouse of MOA.In embodiments of the present invention can corresponding four kinds of filthy states, good MOA filth training knowledge base prepared in advance trains this submodule, carries out filthy test analysis with the on-the-spot real time data of preparation for acquiring, as shown in table 4.
Table 4
In embodiments of the present invention can using in distributed LIBSVM model 4 analyze and the possible Output rusults of reasoning submodule as the input of MOA health control and the final reasoning module of failure prediction LIBSVM, namely the input vector of final reasoning module is 4, output vector is 1, as shown in table 5.
Table 5
In embodiments, association obtains final MOA health status, and predicts fault and MOA ruuning situation is analyzed, and the possible output of above-mentioned final reasoning module is 4*4*4*4=256 kind health status.Such as: if final reasoning module be input as XX=[Y11Y23Y34Y42], i.e. XX=[1000111011111100], then the output obtained is that Y represents that the state of MOA is non-fault, slightly filthy, but device moderate makes moist, and seriously aging.
Another aspect of the present invention additionally provides a kind of Zinc-Oxide Arrester MOA health control and failure prediction system.Fig. 7 is MOA health control according to embodiment of the present invention and failure prediction system structural representation.As shown in Figure 7, system provided by the invention can comprise: MOA characteristic parameter extraction unit, is configured to measure MOA characteristic parameter data in real time; Real-time testing knowledge base, is configured to the MOA characteristic parameter data storing the measurement of described MOA characteristic parameter extraction unit; Training knowledge base, is configured to the service data relevant to the health status of various MOA storing statistics; MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting, the data that this MOA health control and failure prediction model are configured to according to storing in described training knowledge base are trained; And after having trained, this MOA health control and failure prediction model are configured to the health status according to the data acquisition MOA stored in described real-time testing knowledge base.
In embodiments, said system provided by the invention can be built by LIBSVM.
In embodiments, MOA health control and failure prediction model can comprise: MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module, the filthy macrostate reasoning module of MOA and final reasoning module, and the one wherein in the following health indicator that comprises the health status of MOA respectively of each State reasoning module judges: MOA fault, MOA make moist, MOA is aging, MOA is filthy; And described final reasoning module is configured to the judged result gathering each State reasoning module, to export the health status of MOA.
The beneficial effect of MOA health control provided by the invention and failure prediction method and system is as follows:
1, adopt fault, aging, make moist, filthy four kinds of macroscopical dynamic indicators have carried out health control to MOA and failure prediction, accurately provide again multiple (such as four kinds) health evaluating index accordingly below often kind of macroscopical dynamic indicator, altogether can assess 4*4*4*4=256 kind MOA health status.
2, propose and utilize distributed LIBSVM structure to set up MOA health control and failure prediction model, travelling speed is fast.
3, set up distributed training knowledge base, deposit the historical data knowledge of all kinds of state of reflection MOA, farthest can reflect the health status that MOA is possible;
4, according to the real-time output of distributed macrostate analysis ratiocination module, failure prediction and MOA overall operation status monitoring and analysis accurately can be realized.
Below the preferred embodiment of the present invention is described in detail by reference to the accompanying drawings; but; the present invention is not limited to the detail in above-mentioned embodiment; within the scope of technical conceive of the present invention; can carry out multiple simple variant to technical scheme of the present invention, these simple variant all belong to protection scope of the present invention.
It should be noted that in addition, each the concrete technical characteristic described in above-mentioned embodiment, in reconcilable situation, can be combined by any suitable mode.In order to avoid unnecessary repetition, the present invention illustrates no longer separately to various possible array mode.
In addition, also can carry out combination in any between various different embodiment of the present invention, as long as it is without prejudice to thought of the present invention, it should be considered as content disclosed in this invention equally.
Claims (10)
1. Zinc-Oxide Arrester MOA health control and a failure prediction method, is characterized in that, the method comprises:
Set up MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting;
Add up the service data relevant to the health status of various MOA, and the service data of Corpus--based Method builds training knowledge base;
Utilize the data stored in described training knowledge base to described MOA health control and the training of failure prediction model; And
The MOA health control trained the MOA characteristic parameter data measured in real time input and failure prediction model, obtain the health status of MOA.
2. MOA health control according to claim 1 and failure prediction method, it is characterized in that, described MOA health control and failure prediction model comprise: MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module, the filthy macrostate reasoning module of MOA.
3. MOA health control according to claim 2 and failure prediction method, it is characterized in that, the method comprises each State reasoning module comprised corresponding to described MOA health control and failure prediction model, and structure MOA fault training knowledge base, the aging training knowledge base of MOA, MOA make moist and train knowledge base, MOA filth training knowledge base respectively; And
Training knowledge base constructed by utilization is trained each State reasoning module accordingly, to complete the training to described MOA health control and failure prediction model.
4. MOA health control according to claim 2 and failure prediction method, is characterized in that, the step of the described MOA health control the MOA characteristic parameter data measured in real time input trained and failure prediction model comprises:
Based on the MOA characteristic parameter data construct real-time testing knowledge base measured in real time, this real-time testing knowledge base comprises MOA fault test knowledge base, MOA burn-in test knowledge base, MOA make moist knowledge on testing storehouse, the filthy knowledge on testing storehouse of MOA; And
MOA characteristic parameter data in each knowledge on testing storehouse are input to each State reasoning module accordingly as input.
5. the failure prediction method of MOA health control according to claim 2, it is characterized in that, described MOA health control and failure prediction model also comprise final reasoning module, and the method comprises the health status of the judged result output MOA being gathered each State reasoning module by this final reasoning module.
6. the failure prediction method of MOA health control according to claim 1, is characterized in that, the health status of described MOA comprises following health indicator: MOA fault, MOA make moist, MOA is aging, MOA is filthy, and wherein each health indicator comprises multiple grade.
7. the failure prediction method of MOA health control according to claim 1, is characterized in that, the method also comprises use LIBSVM to set up MOA health control and failure prediction model.
8. the failure prediction method of MOA health control according to claim 5, is characterized in that, the method also comprises the multiple LIBSVM of use and wraps each module set up accordingly in MOA health control and failure prediction model.
9. Zinc-Oxide Arrester MOA health control and a failure prediction system, is characterized in that, this system comprises:
MOA characteristic parameter extraction unit, is configured to measure MOA characteristic parameter data in real time;
Real-time testing knowledge base, is configured to the MOA characteristic parameter data storing the measurement of described MOA characteristic parameter extraction unit;
Training knowledge base, is configured to the service data relevant to the health status of various MOA storing statistics;
MOA health control and failure prediction model, this MOA health control and failure prediction model with MOA characteristic parameter data be input, with the health status of MOA for exporting, the data that this MOA health control and failure prediction model are configured to according to storing in described training knowledge base are trained; And after having trained, this MOA health control and failure prediction model are configured to the health status according to the data acquisition MOA stored in described real-time testing knowledge base.
10. Zinc-Oxide Arrester MOA according to claim 9 health control and failure prediction system, it is characterized in that, described MOA health control and failure prediction model comprise: MOA fault macrostate reasoning module, the aging macrostate reasoning module of MOA, MOA make moist macrostate reasoning module, the filthy macrostate reasoning module of MOA and final reasoning module, wherein
One in the following health indicator that each State reasoning module comprises the health status of MOA respectively judges: MOA fault, MOA make moist, MOA is aging, MOA is filthy; And
Described final reasoning module is configured to the judged result gathering each State reasoning module, to export the health status of MOA.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410460875.XA CN105469156A (en) | 2014-09-11 | 2014-09-11 | MOA condition management and fault prediction method and MOA condition management and fault prediction system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410460875.XA CN105469156A (en) | 2014-09-11 | 2014-09-11 | MOA condition management and fault prediction method and MOA condition management and fault prediction system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105469156A true CN105469156A (en) | 2016-04-06 |
Family
ID=55606825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410460875.XA Pending CN105469156A (en) | 2014-09-11 | 2014-09-11 | MOA condition management and fault prediction method and MOA condition management and fault prediction system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105469156A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106203642A (en) * | 2016-07-18 | 2016-12-07 | 王力 | The prediction of a kind of fault of electric locomotive and the method for health control |
| CN106771513A (en) * | 2016-11-22 | 2017-05-31 | 华北电力大学 | The determination method and device of arrester early warning value |
| CN106886481A (en) * | 2017-02-28 | 2017-06-23 | 深圳市华傲数据技术有限公司 | A kind of system health degree static analysis Forecasting Methodology and device |
| CN107679445A (en) * | 2017-08-14 | 2018-02-09 | 南京理工大学 | A kind of arrester ageing failure diagnosis method based on wavelet-packet energy entropy |
| CN110442898A (en) * | 2019-06-14 | 2019-11-12 | 广东电网有限责任公司江门供电局 | A kind of electric power pylon health status model method for on-line optimization |
| CN112163297A (en) * | 2020-09-30 | 2021-01-01 | 厦门科灿信息技术有限公司 | Equipment health prediction system |
| CN113158452A (en) * | 2021-04-09 | 2021-07-23 | 国网福建省电力有限公司龙岩供电公司 | Lightning arrester defect reason analysis method and system based on support vector machine |
| CN114660387A (en) * | 2022-03-18 | 2022-06-24 | 国网安徽省电力有限公司马鞍山供电公司 | Lightning arrester monitoring method based on leakage current sensor and BP neural network algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101907868A (en) * | 2010-08-05 | 2010-12-08 | 暨南大学珠海学院 | Intelligent trouble diagnosis method for tractive power supply system and system thereof |
| CN102722759A (en) * | 2012-05-17 | 2012-10-10 | 河海大学 | Method for predicting power supply reliability of power grid based on BP neural network |
| CN103455658A (en) * | 2013-07-10 | 2013-12-18 | 西北工业大学自动化学院 | Weighted grey target theory based fault-tolerant motor health status assessment method |
-
2014
- 2014-09-11 CN CN201410460875.XA patent/CN105469156A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101907868A (en) * | 2010-08-05 | 2010-12-08 | 暨南大学珠海学院 | Intelligent trouble diagnosis method for tractive power supply system and system thereof |
| CN102722759A (en) * | 2012-05-17 | 2012-10-10 | 河海大学 | Method for predicting power supply reliability of power grid based on BP neural network |
| CN103455658A (en) * | 2013-07-10 | 2013-12-18 | 西北工业大学自动化学院 | Weighted grey target theory based fault-tolerant motor health status assessment method |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106203642A (en) * | 2016-07-18 | 2016-12-07 | 王力 | The prediction of a kind of fault of electric locomotive and the method for health control |
| CN106771513A (en) * | 2016-11-22 | 2017-05-31 | 华北电力大学 | The determination method and device of arrester early warning value |
| CN106886481A (en) * | 2017-02-28 | 2017-06-23 | 深圳市华傲数据技术有限公司 | A kind of system health degree static analysis Forecasting Methodology and device |
| CN107679445A (en) * | 2017-08-14 | 2018-02-09 | 南京理工大学 | A kind of arrester ageing failure diagnosis method based on wavelet-packet energy entropy |
| CN110442898A (en) * | 2019-06-14 | 2019-11-12 | 广东电网有限责任公司江门供电局 | A kind of electric power pylon health status model method for on-line optimization |
| CN110442898B (en) * | 2019-06-14 | 2023-06-09 | 广东电网有限责任公司江门供电局 | Power transmission line health condition model online optimization method |
| CN112163297A (en) * | 2020-09-30 | 2021-01-01 | 厦门科灿信息技术有限公司 | Equipment health prediction system |
| CN112163297B (en) * | 2020-09-30 | 2023-07-18 | 厦门科灿信息技术有限公司 | Equipment health prediction system |
| CN113158452A (en) * | 2021-04-09 | 2021-07-23 | 国网福建省电力有限公司龙岩供电公司 | Lightning arrester defect reason analysis method and system based on support vector machine |
| CN114660387A (en) * | 2022-03-18 | 2022-06-24 | 国网安徽省电力有限公司马鞍山供电公司 | Lightning arrester monitoring method based on leakage current sensor and BP neural network algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105469156A (en) | MOA condition management and fault prediction method and MOA condition management and fault prediction system | |
| CN105353256B (en) | A kind of power transmission and transformation equipment state method for detecting abnormality | |
| CN106199305B (en) | Underground coal mine power supply system dry-type transformer insulation health state evaluation method | |
| CN105589040B (en) | Battery regulation and control method based on aging adjustment battery operation interval | |
| CN103576050B (en) | A kind of running status appraisal procedure of capacitance type potential transformer | |
| CN102221651A (en) | Fault on-line diagnosis and early warning method of flameproof dry-type transformer for mine | |
| CN102508055B (en) | Device and method for detecting wind power generation grid-connected system | |
| CN106054036B (en) | A kind of main time constant computational methods of paper oil insulation based on extension debye equivalent circuit | |
| CN103399241A (en) | Distribution transformer fault diagnosis system and method based on relation between temperature rise and load | |
| CN109034461A (en) | A kind of voltage dip Stochastic prediction method based on actual electric network monitoring information | |
| CN107886171A (en) | A kind of circuit-breaker status inline diagnosis method and system based on PMU data | |
| CN106546859A (en) | A kind of on-line monitoring system for cable arrester | |
| CN102590652B (en) | Electric-information-based equipment performance evaluation system and method | |
| CN112684274B (en) | Lightning arrester reliability state assessment method under different air water contents | |
| CN104655967B (en) | Distribution transformer basket vibration signal characteristic quantity extracting method | |
| CN108459269A (en) | A kind of 10kV pvs (pole-mounted vacuum switch)s state on-line evaluation method and apparatus | |
| CN106370913A (en) | Node voltage dip severity evaluation system and evaluation method | |
| CN103413033A (en) | Method for predicting storage battery faults | |
| CN114779151B (en) | Comprehensive evaluation method, device, equipment and medium for capacitor voltage transformer | |
| CN111025096B (en) | XLPE cable aging state evaluation method based on leakage current characteristic factor | |
| CN106443253A (en) | Power transmission line parameter identification method based on PMU (phasor measurement unit) data | |
| CN105259449A (en) | System for evaluating state of EHV (extra-high voltage) lightning arrester | |
| CN105486917A (en) | Method for detecting energy-efficient power transmission line loss and faults | |
| CN205067637U (en) | State evaluation system of extra -high voltage arrester | |
| CN114778956B (en) | Marine substation sleeve insulation performance evaluation method considering saline water influence factors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160406 |
|
| RJ01 | Rejection of invention patent application after publication |