CN103592587A - Partial discharge diagnosis method based on data mining - Google Patents
Partial discharge diagnosis method based on data mining Download PDFInfo
- Publication number
- CN103592587A CN103592587A CN201310631966.0A CN201310631966A CN103592587A CN 103592587 A CN103592587 A CN 103592587A CN 201310631966 A CN201310631966 A CN 201310631966A CN 103592587 A CN103592587 A CN 103592587A
- Authority
- CN
- China
- Prior art keywords
- shelf depreciation
- method based
- data mining
- discharge
- diagnostic method
- 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
Images
Landscapes
- Testing Relating To Insulation (AREA)
Abstract
The invention discloses a partial discharge diagnosis method based on data mining. The method comprises the steps of 1) acquiring a partial discharge signal of power equipment; 2) extracting characteristic parameters from the partial discharge signal; 3) implementing interval division in a competitive gathering method according to the extracted characteristic parameters, mining a training database of partial discharge types after interval division via an association rule mining algorithm, obtaining a classification rule according to a set minimal confidence and a set minimal support degree, and according to the classification rule, calculating the subordination degrees of the to-be-classified partial discharge signal to the partial discharge types in a fuzzy inference method; and 4) according to the calculated subordination degrees, determining the possibility that the to-be-classified partial discharge signal represents a certain partial discharge type. The partial discharge diagnosis method disclosed by the invention is high in recognition rate and recognition speed.
Description
Technical field
The present invention relates to a kind of electrical signal detection method, relate in particular to a kind of diagnostic method of shelf depreciation.
Background technology
For the large scale electrical power unit that has moved certain time limit, its Insulation Problems can be day by day remarkable along with the growth of tenure of use.The local defect of rugged environment factor and power equipment itself also can shorten the serviceable life of power equipment greatly simultaneously, causes insulation ag(e)ing serious, Frequent Troubles.
As the Important Parameters of reflection large scale electrical power unit state of insulation, the insulation status of shelf depreciation and power equipment has close ties.The local discharge signal feature that different defects produce is different, and dissimilar shelf depreciation is also different to the destructiveness of Electric Power Equipment Insulation, and therefore, the diagnosis of power equipment shelf depreciation is very important.
Existing shelf depreciation diagnostic method has neural net method, traditional decision-tree, fuzzy reasoning method etc.Wherein, neural net method, because having the advantages such as self-organized learning ability and Nonlinear Processing, is usually used in carrying out PD Pattern Recognition, yet the method is explanatory poor, can not produce identifying comparatively intuitively, and when neural network nodes is more, can consume a large amount of training times.It is foundation that Fuzzy Logic Reasoning Algorithm be take artificial experience and expert decision-making, close statement information is concluded and summed up to build shelf depreciation fuzzy inference rule storehouse, the method can overcome the deficiency of neural metwork training time length and explanatory difference, but the structure of rule base needs a large amount of human expert experiences, thereby system constructing expends hugely, actual operation is poor.Traditional decision-tree is directly inferred the method for classifying rules from sample data, the advantage that the method has rule-based reasoning system has reduced required expense and the time of constructing system simultaneously, but it only can differentiate for a feature at every turn, therefore the rule producing is too simple, and feature recognition capability is lower.
As can be seen here, while adopting above several existing methods to carry out shelf depreciation type identification, it is explanatory poor to exist respectively, the low and not high many-sided deficiency of discrimination of efficiency.And in PD Pattern Recognition, the correctness that needs on the one hand abundant expertise check training storehouse, the characteristic of on-line monitoring is had higher requirement to recognition speed and discrimination simultaneously, so need to find a kind of advantage of new method synthesis each side, complete PD Pattern Recognition and fault diagnosis.
Summary of the invention
The object of the present invention is to provide a kind of shelf depreciation diagnostic method based on data mining, it is for the deficiency of existing shelf depreciation diagnostic method, aims to provide a kind of explanatory good, recognition speed is fast, discrimination is high shelf depreciation diagnostic method.
Technical solutions according to the invention have adopted the fuzzy reasoning recognition methods of correlation rule, its basic thought is to each local discharge characteristic, to adopt Competition Clustering method dynamically to divide fuzzy interval respectively, and continuous eigenwert is integrated in respective bins with discretize; The mutual relationship of excavating between discrete features by association rules is extracted classifying rules, and then these vagueness of regulations are obtained to final sorter.The method can, according to each classifying rules and degree of membership thereof, be carried out comprehensively, finally providing believable diagnosis according to certain rule.
To achieve these goals, the present invention proposes a kind of shelf depreciation diagnostic method based on data mining, it comprises step:
(1) local discharge signal of induction power equipment, carries out analog to digital conversion to convert digital signal to after being amplified, this digital signal is carried out to continuous acquisition;
(2) by the digital signal stack of continuous acquisition, obtain
three-dimensional collection of illustrative plates, wherein
represent operating frequency phase, q represents discharge capacity, and n represents discharge time; From three-dimensional collection of illustrative plates, extract maximum pd quantity phase resolved plot
mean discharge magnitude phase resolved plot
discharge time phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
four kinds of two-dimensional maps extract measure of skewness Sk and steepness Ku as characteristic parameter, at maximum pd quantity phase resolved plot in above-mentioned four kinds of two-dimensional maps
mean discharge magnitude phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
in three kinds of two-dimensional maps, extract mutual relationship and count C and discharge capacity factor Q as characteristic parameter;
Wherein, measure of skewness Sk is the tolerance to statistics distribution skew direction and degree.Measure of skewness is the third moment of data, for characterization data, distributes about the degree of asymmetry of normal distribution, is defined as follows:
Wherein, x
idiscrete value, P
ix
iprobability, u is the mean value of data, σ is the standard deviation of data.
Steepness Ku is for reflecting the index of the high and steep or flat degree of curve of frequency distribution top point.Steepness is the Fourth-order moment of data, and the deviation for characterization data distribution with normal distribution, is defined as follows:
Wherein, x
idiscrete value, P
ix
iprobability, u is the mean value of data, σ is the standard deviation of data.
What so-called cross-correlation coefficient C characterized is the degree of correlation that positive half cycle electric discharge distributes and negative half period discharges between distributing.
So-called discharge factor Q refers to the degree of asymmetry that positive half cycle electric discharge distributes and negative half period discharges between distributing.
(3) characteristic parameter of extraction is adopted competition method for congregating carry out interval division, association rule digging algorithm excavates the training storehouse of the shelf depreciation type through interval division, according to the min confidence of setting and the minimum support of setting, obtain classifying rules, according to classifying rules, by fuzzy reasoning method, calculate local discharge signal to be sorted for the degree of membership of shelf depreciation type, this degree of membership has characterized the similarity degree of local discharge characteristic parameter and shelf depreciation type;
(4) according to calculate degree of membership judge that local discharge signal to be sorted represents the possibility of certain shelf depreciation type.
Further, in the shelf depreciation diagnostic method based on data mining of the present invention, between described step (1) and step (2), also comprise Signal Pretreatment step: the digital signal of continuous acquisition is amplified and denoising, remove narrow band noise and white noise.
Further, in the step (1) of the shelf depreciation diagnostic method based on data mining of the present invention, adopt at least local discharge signal of one of them induction power equipment of uhf sensor, sonac, High Frequency Current Sensor, coupling capacitance sensor.
Further, in the step (1) of the shelf depreciation diagnostic method based on data mining of the present invention, the time of continuous acquisition is no less than the integral multiple of at least one power frequency period.
Further, in the step (3) of the shelf depreciation diagnostic method based on data mining of the present invention, shelf depreciation type comprises corona discharge, suspension electrode electric discharge, internal air gap electric discharge and creeping discharge.
Further, in the step (3) of the shelf depreciation diagnostic method based on data mining of the present invention, the min confidence of setting is 0.4.
Further, in the step (3) of the shelf depreciation diagnostic method based on data mining of the present invention, the minimum support of setting is 0.4.
Further, in the shelf depreciation diagnostic method based on data mining of the present invention, described step (4) according to calculate degree of membership judge that local discharge signal to be sorted represents that the possibility of certain shelf depreciation type comprises: capping threshold value and lower threshold; If calculate degree of membership be more than or equal to upper limit threshold, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is for high; If calculate degree of membership be less than upper limit threshold and be greater than lower threshold, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is; If calculate degree of membership be less than or equal to lower threshold, judge that the possibility that local discharge signal characterizes a certain shelf depreciation type is low.
Further, described upper limit threshold is set as 0.8.
Further, described lower threshold is set as 0.5.
Shelf depreciation diagnostic method based on data mining of the present invention can effectively be found incidence relation complicated between each insulation defect feature, has made up the deficiency that traditional decision-tree simple tree type is divided; Can determine corresponding degree of membership relation according to regular degree of confidence and support, effectively avoided general rule inference system to need expertise to determine the limitation of weight simultaneously; In addition, this method with respect to neural net method also have advantages of explanatory good, recognition speed is fast.
Shelf depreciation diagnostic method based on data mining of the present invention can be widely used in the shelf depreciation diagnostic application of different power equipments, different method of testings.
Accompanying drawing explanation
Fig. 1 is the shelf depreciation diagnostic method based on data mining of the present invention schematic flow sheet in one embodiment.
Fig. 2 is the schematic flow sheet that the shelf depreciation diagnostic method based on data mining of the present invention adopts association rules mining algorithm to excavate the training storehouse of local electric discharge type under a kind of embodiment.
Fig. 3 be the shelf depreciation diagnostic method based on data mining of the present invention in one embodiment, according to calculate degree of membership judge that local discharge signal to be sorted represents the schematic flow sheet of the possibility of certain shelf depreciation type.
Embodiment
Below in conjunction with Figure of description and specific embodiment, the shelf depreciation diagnostic method based on data mining of the present invention is made to further explanation and explanation.
As shown in Figure 1, adopt in the present embodiment Matlab development language that the diagnostic method of the shelf depreciation based on data mining is provided, it comprises the following steps:
(1) collection signal: the local discharge signal that adopts sensor sensing power equipment, after being amplified, local discharge signal carries out analog to digital conversion to convert digital signal to, this digital signal is carried out to continuous acquisition, and the time of continuous acquisition signal is two power frequency periods (2 * 20=40 milliseconds).
(2) pre-service: the digital signal collecting is amplified to the processing in earlier stage such as denoising, adopt FIR filtering and wavelet method to suppress the narrow band noise in signal and white noise interference, better highlight partial discharge pulse, improve the accuracy of follow-up recognition methods.
(3) feature extraction, the input using the characteristic parameter extracting as sorter: by continuous acquisition to digital signal stack after to obtain
three-dimensional collection of illustrative plates, by operating frequency phase
uniformly-spaced be divided into 60 intervals according to 0 °～360 °, discharge capacity q is uniformly-spaced divided into 200 intervals according to 0～1000pC, statistics
discharge time n in each grid section of plane, can obtain
three-dimensional collection of illustrative plates.From three-dimensional collection of illustrative plates, can extract maximum pd quantity phase resolved plot
mean discharge magnitude phase resolved plot
discharge time phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
four kinds of two-dimensional maps.In above-mentioned four kinds of two-dimensional maps, extract measure of skewness Sk and steepness Ku totally 14 parameters as characteristic parameter (collection of illustrative plates that phase place is relevant need be chosen respectively in positive-negative half-cycle, wherein+,-represent respectively positive half cycle and negative half period):
Maximum pd quantity phase resolved plot
sk+, Sk-, Ku+, Ku-
Mean discharge magnitude phase resolved plot
sk+, Sk-, Ku+, Ku-
Discharge time phase resolved plot
sk+, Sk-, Ku+, Ku-
Discharge capacity number of times distribution collection of illustrative plates
sk, Ku
In like manner, at maximum pd quantity phase resolved plot
mean discharge magnitude phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
in three kinds of two-dimensional maps, extract mutual relationship count C and discharge capacity factor Q totally 6 parameters as characteristic parameter.
(4) sorter identification, Fig. 2 has shown the schematic flow sheet that adopts association rules mining algorithm to excavate the training storehouse of local electric discharge type:
(4a) adopt competitive agglomeration algorithm to divide the scalar type feature of insulation defect: carry out in the process of association rule mining in the optimization interval at application Competition Clustering algorithm gained, if interval division is meticulous, easily there is the phenomenon of over-fitting, can restrict the efficiency of association rule mining simultaneously, if interval number is very few, will reduce the validity of rule set.Choose the interval number corresponding data of a stack features parameter within the specific limits as the object of rule digging.After interval division, can obtain a new data set, establishing this data set has n sample, and each sample represents strip defect data; If this data set has r bar attribute, by fuzzy interval grade and the classification results grade of scalar type feature, formed, adopt maximum subjection principle that scalar type feature set is converted into Boolean type feature set.
It should be noted that, the optimum number of interval division is obtained by competitive agglomeration algorithm, for each scalar type feature, all can obtain an optimum interval number.Generally speaking, the scalar type feature of finally choosing, its optimum interval number can be positioned at all optimum intervals and count a near scope of intermediate value.For example, scalar type feature has 5, and optimum interval number is respectively 3,4,5,6,7, and the scalar type feature of finally choosing has 3, and 3 optimum interval numbers corresponding to feature should be 4,5,6 so.
(4b) adopt association rules method to excavate above-mentioned data: setting min confidence is 0.4, and minimum support is 0.05, scanning new data set, calculates the number of times of every appearance, produces frequent 1-item collection; On the basis of frequent 1-item collection, search for frequent 2-item collection, so iterative loop, obtains frequent k-item collection according to frequent (k-1)-collection set; Deletion is less than the item collection of minimum support, until obtain all k-frequent item sets; From frequent item set, obtain being not less than the correlation rule of min confidence.
(4c) obtain after effective correlation rule, the fuzzy fact of known each regular former piece is compared according to its connected mode, infer the matching degree that each is regular.In rule former piece, adopt " AND " conjunction, calculating accordingly operator is T normal form, and the mode of selecting algebraically to take advantage of in the present embodiment can obtain the output fuzzy set of every rule, like this by producing quantitative output after de-fuzzy.Adopt gravity model appoach to complete de-fuzzy process, domain output is discrete.Comparative sample X respectively organizes data for the discriminant function of each classification, obtains discharging accordingly degree of membership.
(5) possibility is differentiated: the degree of membership being provided by sorter is released corresponding possibility, is specially: if calculate degree of membership be more than or equal to upper limit threshold 0.8, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is for high; If calculate degree of membership be less than 0.8 and be greater than 0.5, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is; If calculate degree of membership be less than or equal to 0.5, judge that the possibility that local discharge signal characterizes a certain shelf depreciation type is low.
(6) make final shelf depreciation type conclusion.
Be noted that above enumerate only for specific embodiments of the invention, obviously the invention is not restricted to above embodiment, have many similar variations thereupon.If all distortion that those skilled in the art directly derives or associates from content disclosed by the invention, all should belong to protection scope of the present invention.
Claims (10)
1. the shelf depreciation diagnostic method based on data mining, is characterized in that, comprises step:
(1) local discharge signal of induction power equipment, carries out analog to digital conversion to convert digital signal to after being amplified, this digital signal is carried out to continuous acquisition;
(2) by the digital signal stack of continuous acquisition, obtain
three-dimensional collection of illustrative plates, wherein
represent operating frequency phase, q represents discharge capacity, and n represents discharge time; From three-dimensional collection of illustrative plates, extract maximum pd quantity phase resolved plot
mean discharge magnitude phase resolved plot
discharge time phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
four kinds of two-dimensional maps extract measure of skewness Sk and steepness Ku as characteristic parameter, at maximum pd quantity phase resolved plot in above-mentioned four kinds of two-dimensional maps
mean discharge magnitude phase resolved plot
and discharge capacity number of times distribution collection of illustrative plates
in three kinds of two-dimensional maps, extract mutual relationship and count C and discharge capacity factor Q as characteristic parameter;
(3) characteristic parameter of extraction is adopted competition method for congregating carry out interval division, association rule digging algorithm excavates the training storehouse of the shelf depreciation type through interval division, according to the min confidence of setting and the minimum support of setting, obtain classifying rules, according to classifying rules, by fuzzy reasoning method, calculate local discharge signal to be sorted for the degree of membership of shelf depreciation type;
(4) according to calculate degree of membership judge that local discharge signal to be sorted represents the possibility of certain shelf depreciation type.
2. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, it is characterized in that, between described step (1) and step (2), also comprise Signal Pretreatment step: the digital signal of continuous acquisition is amplified and denoising, remove narrow band noise and white noise.
3. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, is characterized in that, in described step (1), the time of continuous acquisition is no less than the integral multiple of at least one power frequency period.
4. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, is characterized in that, in described step (3), shelf depreciation type comprises corona discharge, suspension electrode electric discharge, internal air gap electric discharge and creeping discharge.
5. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, it is characterized in that, described step (4) according to calculate degree of membership judge that local discharge signal to be sorted represents the possibility of certain shelf depreciation type, it comprises: capping threshold value and lower threshold; If calculate degree of membership be more than or equal to upper limit threshold, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is for high; If calculate degree of membership be less than upper limit threshold and be greater than lower threshold, judge that possibility that local discharge signal characterizes a certain shelf depreciation type is; If calculate degree of membership be less than or equal to lower threshold, judge that the possibility that local discharge signal characterizes a certain shelf depreciation type is low.
6. the shelf depreciation diagnostic method based on data mining as claimed in claim 5, is characterized in that, described upper limit threshold is set as 0.8.
7. the shelf depreciation diagnostic method based on data mining as claimed in claim 5, is characterized in that, described lower threshold is set as 0.5.
8. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, is characterized in that, in described step (3), the min confidence of setting is 0.4.
9. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, is characterized in that, in described step (3), the minimum support of setting is 0.4.
10. the shelf depreciation diagnostic method based on data mining as claimed in claim 1, it is characterized in that, in described step (1), adopt at least local discharge signal of one of them induction power equipment of uhf sensor, sonac, High Frequency Current Sensor, coupling capacitance sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310631966.0A CN103592587A (en) | 2013-12-02 | 2013-12-02 | Partial discharge diagnosis method based on data mining |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310631966.0A CN103592587A (en) | 2013-12-02 | 2013-12-02 | Partial discharge diagnosis method based on data mining |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103592587A true CN103592587A (en) | 2014-02-19 |
Family
ID=50082803
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310631966.0A Pending CN103592587A (en) | 2013-12-02 | 2013-12-02 | Partial discharge diagnosis method based on data mining |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103592587A (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104375064A (en) * | 2014-11-07 | 2015-02-25 | 深圳供电局有限公司 | Partial discharge online monitoring alarm confidence coefficient analysis method and device |
| CN104375066A (en) * | 2014-11-07 | 2015-02-25 | 国家电网公司 | GIS partial discharge mode identification method under oscillation mode impulse voltage |
| CN104407278A (en) * | 2014-11-07 | 2015-03-11 | 国家电网公司 | Extraction method and device of ultrasonic wave characteristic parameters |
| CN104535905A (en) * | 2014-12-11 | 2015-04-22 | 国家电网公司 | Partial discharge diagnosis method based on naive bayesian classification |
| CN104573335A (en) * | 2014-12-25 | 2015-04-29 | 武汉大学苏州研究院 | Extraction method for feature information of partial-discharge mass and real-time electric physical quantity |
| CN104865508A (en) * | 2015-06-08 | 2015-08-26 | 中国科学院半导体研究所 | Partial discharge recognition method based on data grouping quantification |
| CN104991171A (en) * | 2015-06-25 | 2015-10-21 | 国家电网公司 | Method for drawing GIS partial discharge frequency division fault spectrogram based on ultrahigh frequency signal |
| CN105182200A (en) * | 2015-09-28 | 2015-12-23 | 苏州光格设备有限公司 | Noise reduction processing method for cable local discharging signal |
| CN106156316A (en) * | 2016-07-04 | 2016-11-23 | 长江大学 | Special name under a kind of big data environment and native place correlating method and system |
| CN106526434A (en) * | 2016-10-11 | 2017-03-22 | 国网上海市电力公司 | Partial discharge mode identifying method and device |
| CN107247995A (en) * | 2016-09-29 | 2017-10-13 | 上海交通大学 | Transmission line of electricity running status association rule mining and Forecasting Methodology based on Bayesian model |
| CN108414905A (en) * | 2018-04-08 | 2018-08-17 | 国网上海市电力公司 | Partial discharge of transformer recognition methods based on manifold distance neighbour's propagation clustering |
| CN108717152A (en) * | 2018-04-09 | 2018-10-30 | 保定华创电气有限公司 | A kind of GIS partial discharge localization method based on interval statistics algorithm |
| CN109031053A (en) * | 2017-06-08 | 2018-12-18 | 许继集团有限公司 | The shelf depreciation alarm method and system of for transformer status monitoring intelligent apparatus |
| CN110837028A (en) * | 2019-09-27 | 2020-02-25 | 中国船舶重工集团公司第七一九研究所 | Method for rapidly identifying partial discharge mode |
| CN111044845A (en) * | 2019-12-25 | 2020-04-21 | 国网天津市电力公司 | Power distribution network accident identification method and system based on Apriori algorithm |
| CN112834876A (en) * | 2020-12-31 | 2021-05-25 | 广东电网有限责任公司广州供电局 | Cable state detection method, device, equipment and computer readable storage medium |
| CN118094264A (en) * | 2024-04-29 | 2024-05-28 | 山东泰开电力电子有限公司 | Intelligent power capacitor partial discharge detection method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11258296A (en) * | 1998-03-09 | 1999-09-24 | Fujikura Ltd | Automatic discrimination method for partial discharge |
| CN101408580A (en) * | 2008-11-21 | 2009-04-15 | 重庆大学 | Method for evaluating oil paper insulation ageing state based on local discharge characteristic parameter |
| CN101614775A (en) * | 2009-07-15 | 2009-12-30 | 河北科技大学 | Transformer State Assessment system and appraisal procedure thereof based on Multi-source Information Fusion |
| CN103323749A (en) * | 2013-05-16 | 2013-09-25 | 上海交通大学 | Multi-classifier information fusion partial discharge diagnostic method |
-
2013
- 2013-12-02 CN CN201310631966.0A patent/CN103592587A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11258296A (en) * | 1998-03-09 | 1999-09-24 | Fujikura Ltd | Automatic discrimination method for partial discharge |
| CN101408580A (en) * | 2008-11-21 | 2009-04-15 | 重庆大学 | Method for evaluating oil paper insulation ageing state based on local discharge characteristic parameter |
| CN101614775A (en) * | 2009-07-15 | 2009-12-30 | 河北科技大学 | Transformer State Assessment system and appraisal procedure thereof based on Multi-source Information Fusion |
| CN103323749A (en) * | 2013-05-16 | 2013-09-25 | 上海交通大学 | Multi-classifier information fusion partial discharge diagnostic method |
Non-Patent Citations (1)
| Title |
|---|
| 姚林朋等: "基于关联规则的XLPE电缆局部放电模糊识别研究", 《电工技术学报》, vol. 27, no. 5, 31 May 2012 (2012-05-31) * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104375066A (en) * | 2014-11-07 | 2015-02-25 | 国家电网公司 | GIS partial discharge mode identification method under oscillation mode impulse voltage |
| CN104407278A (en) * | 2014-11-07 | 2015-03-11 | 国家电网公司 | Extraction method and device of ultrasonic wave characteristic parameters |
| CN104375064A (en) * | 2014-11-07 | 2015-02-25 | 深圳供电局有限公司 | Partial discharge online monitoring alarm confidence coefficient analysis method and device |
| CN104375066B (en) * | 2014-11-07 | 2018-04-03 | 国家电网公司 | GIS partial discharge mode identification method under a kind of oscillation mode surge voltage |
| CN104535905B (en) * | 2014-12-11 | 2017-05-24 | 国家电网公司 | Partial discharge diagnosis method based on naive bayesian classification |
| CN104535905A (en) * | 2014-12-11 | 2015-04-22 | 国家电网公司 | Partial discharge diagnosis method based on naive bayesian classification |
| CN104573335A (en) * | 2014-12-25 | 2015-04-29 | 武汉大学苏州研究院 | Extraction method for feature information of partial-discharge mass and real-time electric physical quantity |
| CN104573335B (en) * | 2014-12-25 | 2017-08-22 | 武汉大学苏州研究院 | A kind of shelf depreciation real-time mass electric physical quantity characteristics information extraction method |
| CN104865508A (en) * | 2015-06-08 | 2015-08-26 | 中国科学院半导体研究所 | Partial discharge recognition method based on data grouping quantification |
| CN104865508B (en) * | 2015-06-08 | 2017-09-29 | 中国科学院半导体研究所 | The Recognition of Partial Discharge quantified based on packet |
| CN104991171A (en) * | 2015-06-25 | 2015-10-21 | 国家电网公司 | Method for drawing GIS partial discharge frequency division fault spectrogram based on ultrahigh frequency signal |
| CN105182200A (en) * | 2015-09-28 | 2015-12-23 | 苏州光格设备有限公司 | Noise reduction processing method for cable local discharging signal |
| CN106156316A (en) * | 2016-07-04 | 2016-11-23 | 长江大学 | Special name under a kind of big data environment and native place correlating method and system |
| CN107247995A (en) * | 2016-09-29 | 2017-10-13 | 上海交通大学 | Transmission line of electricity running status association rule mining and Forecasting Methodology based on Bayesian model |
| CN106526434A (en) * | 2016-10-11 | 2017-03-22 | 国网上海市电力公司 | Partial discharge mode identifying method and device |
| CN109031053A (en) * | 2017-06-08 | 2018-12-18 | 许继集团有限公司 | The shelf depreciation alarm method and system of for transformer status monitoring intelligent apparatus |
| CN108414905A (en) * | 2018-04-08 | 2018-08-17 | 国网上海市电力公司 | Partial discharge of transformer recognition methods based on manifold distance neighbour's propagation clustering |
| CN108717152A (en) * | 2018-04-09 | 2018-10-30 | 保定华创电气有限公司 | A kind of GIS partial discharge localization method based on interval statistics algorithm |
| CN110837028A (en) * | 2019-09-27 | 2020-02-25 | 中国船舶重工集团公司第七一九研究所 | Method for rapidly identifying partial discharge mode |
| CN110837028B (en) * | 2019-09-27 | 2021-08-31 | 中国船舶重工集团公司第七一九研究所 | Method for rapidly identifying partial discharge mode |
| CN111044845A (en) * | 2019-12-25 | 2020-04-21 | 国网天津市电力公司 | Power distribution network accident identification method and system based on Apriori algorithm |
| CN111044845B (en) * | 2019-12-25 | 2021-07-23 | 国网天津市电力公司 | Power distribution network accident identification method and system based on Apriori algorithm |
| CN112834876A (en) * | 2020-12-31 | 2021-05-25 | 广东电网有限责任公司广州供电局 | Cable state detection method, device, equipment and computer readable storage medium |
| CN118094264A (en) * | 2024-04-29 | 2024-05-28 | 山东泰开电力电子有限公司 | Intelligent power capacitor partial discharge detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103592587A (en) | Partial discharge diagnosis method based on data mining | |
| CN103323749B (en) | The partial discharge diagnostic method of multi-classifier information fusion | |
| CN103234767B (en) | Based on the nonlinear fault detection method of semi-supervised manifold learning | |
| Tang et al. | A hybrid approach to integrate fuzzy C-means based imputation method with genetic algorithm for missing traffic volume data estimation | |
| WO2022067562A1 (en) | Method and device for diagnosing fault arc, and computer-readable storage medium | |
| WO2016029570A1 (en) | Intelligent alert analysis method for power grid scheduling | |
| CN103092975A (en) | Detection and filter method of network community garbage information based on topic consensus coverage rate | |
| CN101226133B (en) | Method for specification and recognition of hemocyte pulse signal | |
| CN102542262A (en) | Waveform identification method based on operating-characteristic working condition waveform library of high-speed rail | |
| CN103323702A (en) | Composite power quality disturbing signal identifying method | |
| CN102879677A (en) | Intelligent fault diagnosis method based on rough Bayesian network classifier | |
| CN104729667A (en) | Method for recognizing disturbance type in a distributed optical fiber vibration sensing system | |
| CN103778569A (en) | Distributed generation island detection method based on meta learning | |
| CN104535905A (en) | Partial discharge diagnosis method based on naive bayesian classification | |
| CN104156403A (en) | Clustering-based big data normal-mode extracting method and system | |
| CN105426441B (en) | A kind of automatic preprocess method of time series | |
| CN105447464A (en) | Electric energy quality disturbance recognition and classification method based on PSO | |
| CN110927527A (en) | Transient fault sensing method based on transient energy double channels and information fusion | |
| CN112732748A (en) | Non-invasive household appliance load identification method based on adaptive feature selection | |
| CN102063626A (en) | Power quality disturbance mode discrimination method | |
| CN105373620A (en) | Mass battery data exception detection method and system for large-scale battery energy storage power stations | |
| CN103914558A (en) | Method for mining space-time aggregation patterns of meteorological elements on basis of space-time statistics | |
| CN113887912A (en) | Non-invasive load identification method for deeply learning downward embedded equipment | |
| CN105654392A (en) | Familial defect analysis method of equipment based on clustering algorithm | |
| CN103218614A (en) | Method for identifying severe convection weather based on combination of manifold learning and rough set |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140219 |