CN111582406A - Power equipment state monitoring data clustering method and system - Google Patents
Power equipment state monitoring data clustering method and system Download PDFInfo
- Publication number
- CN111582406A CN111582406A CN202010481567.0A CN202010481567A CN111582406A CN 111582406 A CN111582406 A CN 111582406A CN 202010481567 A CN202010481567 A CN 202010481567A CN 111582406 A CN111582406 A CN 111582406A
- Authority
- CN
- China
- Prior art keywords
- clustering
- power equipment
- monitoring data
- dbscan
- sample
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
- H02J13/00002—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by monitoring
Abstract
The application discloses a method and a system for clustering power equipment state monitoring data, wherein the method comprises the following steps: obtaining a data set X to be clusteredm×nThe method comprises the steps of containing m samples, wherein each sample contains n types of variables, and normalizing the n types of variables of each sample; setting a density threshold parameter k of the DBSCAN model; according to standardized treatment Xm×nPlotting a k-distance map of the distance of each sample from the rest of the samples; determining a lower limit threshold value E of a neighborhood radius parameter E according to the category number of the clustering result of the DBSCAN (0, k) model0(ii) a Determining an upper threshold E of E from the k-distance mapmax(ii) a According to E0And EmaxDrawing a curve of class number-E; the optimal value of E is determined according to the "class number-E" curve. The optimization of neighborhood radius parameter E and density threshold parameter k of the DBSCAN model is realized by designing and drawing a 'class number-E' curve, and the DBSCAN model is used for clustering the online monitoring data of the power equipment for online monitoringAnd monitoring pattern recognition and abnormal detection of real-time data, and judging the type of normal data and the type of abnormal data.
Description
Technical Field
The invention belongs to the technical field of pattern recognition and anomaly detection, and particularly relates to a method and a system for clustering power equipment state monitoring data.
Background
The power equipment is the most important core part of the smart grid, and the normal operation of the power equipment is the fundamental guarantee of the safety of the power grid. The power equipment in the smart grid includes: large-scale power transformer: transmission and distribution networks (overhead lines, cable tunnels); the relay protection and control equipment can also comprise equipment such as a generator and the like. With the rapid development of network technology, sensing technology and computer technology, from the research trend analysis of the latest smart grid, the online monitoring and state maintenance of power equipment by adopting an artificial intelligence method becomes the development trend in the field.
The DBSCAN (Density-Based Spatial Clustering of Applications with Noise, abbreviated as Density Clustering) model has wide application in the field of mode classification and abnormality detection of state monitoring data of power systems and power equipment.
In the DBSCAN model, a neighborhood radius parameter E and a density threshold parameter k have a significant influence on the performance of the DBSCAN model, but under the unsupervised condition of lacking of data labels, a cross validation method cannot be used for parameter optimization, and the traditional optimization model based on a k distance graph can enable the parameter optimization result to present strong subjectivity and inaccuracy. In order to ensure the performance of the density clustering model, the invention provides a power equipment state monitoring method and system based on the density clustering model, and aims to improve the clustering of power equipment state monitoring data by improving the accuracy and effectiveness of the density clustering model on the clustering result of the power equipment state monitoring data.
Disclosure of Invention
In order to overcome the defects in the prior art, the method and the system for clustering the state monitoring data of the electric power equipment are provided, and the optimization of the neighborhood radius parameter E and the density threshold parameter k of the DBSCAN model is realized by designing and drawing a class number-E curve, so that the performance of the model is remarkably improved, and the state monitoring data clustering of the electric power equipment is further improved.
In order to achieve the above object, the first invention of the present application adopts the following technical solutions:
a method for clustering power equipment state monitoring data, the method comprising the steps of:
step 1: acquiring online monitoring data set X of electric power equipment to be clusteredm×n，Xm×nContaining m samples, each sample containing n classes of variables, for Xm×nNormalizing the n-type variables of each sample;
step 2: setting a density threshold parameter k of the DBSCAN model;
and step 3: according to standardized treatment Xm×nPlotting a k-distance map of the distance of each sample from the rest of the samples;
and 4, step 4: normalizing the post-X according to DBSCAN (0, k) modelm×nThe clustering result category number of the clustering is carried out, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
And 5: determining an upper threshold E of a neighborhood radius parameter E according to the k-distance graph drawn in the step 3max；
Step 6: according to E0And EmaxDrawing a curve of class number-E;
and 7: determining the optimal value of the neighborhood radius parameter E according to the 'category number-E' curve;
and 8: clustering on-line monitoring data of the power equipment by using the DBSCAN model formed in the step 7, wherein the on-line monitoring data is used for mode identification and abnormal detection of on-line monitoring real-time data, and judging the type of normal data and abnormal data;
DBSCAN refers to density-based clustering.
Preferably, in step 1, X is treatedm×nRespectively carrying out z-score standardization on n types of variables of each sample, wherein the standardization formula is as follows:
in the formula (I), the compound is shown in the specification,
is the normalized measurement value of the t sample of the j variable;
and σjThe mean and standard deviation of the j-th category variables, respectively.
Preferably, in step 2, the density threshold parameter k of the DBSCAN model is set to 2 n.
Preferably, step 3 comprises the steps of:
step 3.1: x after calculation standardizationm×nThe distance of each sample from the remaining samples;
step 3.2: calculating the average distance between each sample and the k samples closest to the sample;
step 3.3: and performing ascending sorting on the m average distance values, taking the obtained ascending sorting sequence number as an abscissa, and drawing a k distance graph by taking the corresponding m average distance values as an ordinate.
Preferably, the distance in step 3 is a euclidean distance.
Preferably, in step 4, the DBSCAN (E, k) model pairs the normalized Xm×nThe clustering result category number for clustering begins to increase along with the increase of the E value and is more than N0When the value of E is E0Wherein N is0The category number of the clustering result of the DBSCAN (0, k) model.
Preferably, in step 5, a point on the k-distance map at which the ordinate is half the maximum ordinate is found, a point on the k-distance map near the point at which the change in the slope of the polyline is maximum is further found, and the ordinate of the point is determined as Emax。
Preferably, step 6 comprises the steps of:
step 6.1: at E0And EmaxGenerates a logarithmic grid vector [ E (1), E (2), E (3), … E (9), E (10) containing 10 values](ii) a Wherein E (1) ═ E0，E(10)＝Emax；
Step 6.2: using DBSCAN (E (i), k) model to standardize the processed X under different values of E (i)m×nThe sample data in (1) is clustered, and the number of cluster categories n (i) is counted, and a "category-E" curve is drawn with E (i) as the abscissa and the corresponding n (i) as the ordinate, where i is 1,2, …,10.
Preferably, in step 7, the first local minimum point to the right of the maximum in the "category number-E" curve or the left end point of the first horizontal line is determined as the optimal value of the neighborhood radius parameter E.
The application also discloses another invention, namely an electric power equipment state monitoring data clustering system, which comprises an acquisition module, a setting module, a first drawing module, a neighborhood radius parameter lower limit threshold determining module, a neighborhood radius parameter upper limit threshold determining module, a second drawing module, a neighborhood radius parameter optimal value determining module and a data clustering module, and is characterized in that:
the acquisition module is used for acquiring an online monitoring data set X of the electric power equipment to be clusteredm×n，Xm×nContaining m samples, each sample containing n classes of variables, for Xm×nNormalizing the n-type variables of each sample;
the setting module is used for setting a density threshold parameter k of the DBSCAN model;
the first drawing module is used for drawing according to the standardized Xm×nPlotting a k-distance map of the distance of each sample from the rest of the samples;
the neighborhood radius parameter lower limit threshold determining module is used for determining the standardized X according to a DBSCAN (0, k) modelm×nThe category number of the clustering result is clustered, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
The neighborhood radius parameter upper limit threshold determining module is used for determining the upper limit threshold E of the neighborhood radius parameter E according to the k-distance map drawn in the step 3max；
The second drawing module is used for drawing according to E0And EmaxDrawing a curve of class number-E;
the neighborhood radius parameter optimal value determining module is used for determining the optimal value of a neighborhood radius parameter E according to a class number-E curve;
and the data clustering module is used for clustering the online monitoring data of the equipment by using the DBSCAN model determined by the neighborhood radius parameter optimal value determining module, and is used for performing mode identification and abnormal detection on online monitoring real-time data and judging the type of normal data and abnormal data.
The beneficial effect that this application reached:
the optimization of neighborhood radius parameter E and density threshold parameter k of the DBSCAN model and the improvement of model performance are realized by designing and drawing a 'category number-E' curve, the accuracy and effectiveness of the DBSCAN density clustering model on the state monitoring data clustering result and the abnormal detection result of the power equipment can be remarkably improved, reasonable and effective data analysis results are provided for power equipment management personnel during the cleaning of raw data and the detection and diagnosis of equipment faults, and the method and the system for clustering the state monitoring data of the power equipment are improved.
Drawings
FIG. 1 is a flow chart of a method for optimizing parameters of a density clustering model under an unsupervised condition according to the present application;
FIG. 2 is a k-distance map in an embodiment of the present application;
FIG. 3 is a "class number-E" curve in an embodiment of the present application;
fig. 4 is a result of F-metric calculation for performance evaluation in the embodiment of the present application.
Detailed Description
The present application is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present application is not limited thereby.
As shown in fig. 1, the method for clustering power equipment state monitoring data includes the following steps:
step 1: obtaining an online monitoring data set X of dissolved gas in transformer oil containing 101 abnormal values780×6The data set comprises 780 samples, each sample comprises 6 types of variables, the 6 types of variables of each sample are respectively subjected to z-score standardization, the jth type of variable is taken as an example, and the standardization method is shown as formula (1).
In the formula (I), the compound is shown in the specification,
is the normalized measurement value of the ith sample of the jth variable;
and σjThe mean and standard deviation of the j-th category variables, respectively. z-score normalization is only one non-limiting method of normalization and one skilled in the art can also use Min-max normalization (Min-max normalization), log function transformation, atan function transformation, and the like.
It can be clearly understood by those skilled in the art that the online monitoring data of the dissolved gas in the transformer oil is used for example only, and the method provided by the present invention can process the online monitoring data of any power equipment, so as to implement online monitoring of the power equipment, such as but not limited to, the partial discharge amount, the discharge position, the dissolved gas in the oil, the dielectric loss value, the equipment capacitance, the leakage current, and the like of the transformer; capacitance, leakage current, dielectric loss, and the like of capacitive devices such as a coupling capacitor, a CVT, and a current transformer.
Step 2: the lower limit of the density threshold parameter k must be greater than n and greater than 3 (when n is less than 3), and the upper limit of k is generally not greater than 5 n;
since the computational complexity of the dbss model increases significantly with the increase of k, it is recommended in the embodiment of the present application to set the density threshold parameter k of the dbss model to 2n to 12;
and step 3: according to standardized treatment X780×6Plotting a k-distance map of the distance of each sample from the remaining samples, comprising the steps of:
step 3.1: x after calculation standardization780×6The distance of each sample from the remaining samples;
step 3.2: calculating the average distance between each sample and the k samples with the nearest distance, namely 12 samples;
step 3.3: the m average distance values are sorted in ascending order, the obtained ascending sort sequence number is used as an abscissa, and the corresponding m average distance values are used as an ordinate to draw a line graph, i.e., a k distance graph, as shown in fig. 2.
The distance function in the example selects the euclidean distance.
And 4, step 4: normalizing the normalized X according to DBSCAN (0, k) model780×6The category number of the clustering result is clustered, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
Specifically, the number N of categories of the clustering result of the DBSCAN (0,12) model when E is 00When E increases to 0.01, the number of classes N of the clustering result starts to rise and is greater than 1, let E0＝0.01。
And 5: finding out the point with half maximum ordinate value on the k-distance map, further finding out the point with maximum change of the broken line slope on the k-distance map near the point, and determining the ordinate of the point as Emax。
In FIG. 2, the k-distance plot is shown at the far rightThe maximum value was taken at the end, the ordinate of the point was approximately 21, and the degree of change in the slope of the curve was the greatest at coordinates (784,10.97) near the point (10) half the maximum value, and therefore 10.97 was determined as Emax。
Step 6: according to E0And EmaxDrawing a 'category number-E' curve, comprising the following steps:
step 6.1: at E0And EmaxGenerates a logarithmic grid vector [0.0100,0.02176,0.04738,0.10313,0.2245,0.48866,1.06366,0.31529,5.03971, 10.97) composed of 10 values of E (i)]；
Step 6.2: using DBSCAN (E (i),12) model to standardize the processed X under different values of E (i)780×6The sample data in (1) is clustered, and the number n (i) of the cluster categories is counted, and a line graph, i.e. a curve of "category number-E", is drawn by taking E (i) as an abscissa and the corresponding n (i) as an ordinate, as shown in fig. 3, where i is 1,2, …, and 10.
And 7: the first local minimum point to the right of the maximum in the "category number-E" curve or the left end point of the first horizontal line is determined as the optimal value of the neighborhood radius parameter E, as shown in fig. 3, i.e., the fourth E (i) value point is selected: e-0.10313.
And 8: and (4) clustering the online monitoring data of the power equipment by using the DBSCAN model formed in the step (7) for mode identification and abnormal detection of online monitoring real-time data and judging the type of normal data and abnormal data.
A power equipment state monitoring data clustering system comprises an acquisition module, a setting module, a first drawing module, a neighborhood radius parameter lower limit threshold determination module, a neighborhood radius parameter upper limit threshold determination module, a second drawing module and a neighborhood radius parameter optimal value determination module;
the acquisition module is used for acquiring a data set X to be clusteredm×n，Xm×nContaining m samples, each sample containing n classes of variables, for Xm×nNormalizing the n-type variables of each sample;
the setting module is used for setting a density threshold parameter k of the DBSCAN model;
the first mentionedA drawing module for processing the processed X according to the standardm×nPlotting a k-distance map of the distance of each sample from the rest of the samples;
the neighborhood radius parameter lower limit threshold determining module is used for determining the standardized X according to a DBSCAN (0, k) modelm×nThe category number of the clustering result is clustered, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
The neighborhood radius parameter upper limit threshold determining module is used for determining the upper limit threshold E of the neighborhood radius parameter E according to the k-distance map drawn in the step 3max；
The second drawing module is used for drawing according to E0And EmaxDrawing a curve of class number-E;
the neighborhood radius parameter optimal value determining module is used for determining the optimal value of a neighborhood radius parameter E according to a class number-E curve;
and the data clustering module is used for clustering the online monitoring data of the equipment by using the DBSCAN model determined by the neighborhood radius parameter optimal value determining module, and is used for performing mode identification and abnormal detection on online monitoring real-time data and judging the type of normal data and abnormal data.
In order to verify the effectiveness of the application in improving the clustering performance of the DBSCAN model, the clustering performance of the DBSCAN model is improved according to X786×6The F metric is used as a performance evaluation index, and the calculation method is shown in formula (2).
Wherein, P is precision rate, R is recall rate; f is the geometric mean of P and R. A higher value for the F metric indicates better performance of the model.
When DBSCAN (E (i),12) is used to normalize X780×6After clustering, F values of clustering results under different e (i) are counted, as shown in fig. 4. As can be seen from fig. 4, F takes a maximum value when E takes the optimal value 0.10313, which indicates that the present application can improve the performance of the density clustering method.
Compared with the prior art, the method and the system have the advantages that optimization of neighborhood radius parameters E and density threshold parameters k of the DBSCAN model and improvement of model performance are achieved by designing and drawing a class number-E curve, accuracy and effectiveness of the DBSCAN density clustering model on power equipment state monitoring data clustering results and abnormal detection results can be remarkably improved, reasonable and effective data analysis results are provided for power equipment managers when raw data cleaning, equipment fault detection and diagnosis are carried out, and the power equipment state monitoring data clustering method and system are improved.
The present applicant has described and illustrated embodiments of the present invention in detail with reference to the accompanying drawings, but it should be understood by those skilled in the art that the above embodiments are merely preferred embodiments of the present invention, and the detailed description is only for the purpose of helping the reader to better understand the spirit of the present invention, and not for limiting the scope of the present invention, and on the contrary, any improvement or modification made based on the spirit of the present invention should fall within the scope of the present invention.
Claims (10)
1. A clustering method for power equipment state monitoring data is characterized in that:
the method comprises the following steps:
step 1: acquiring online monitoring data set X of electric power equipment to be clusteredm×n，Xm×nContaining m samples, each sample containing n classes of variables, for Xm×nNormalizing the n-type variables of each sample;
step 2: setting a density threshold parameter k of the DBSCAN model;
and step 3: according to standardized treatment Xm×nPlotting a k-distance map of the distance of each sample from the rest of the samples;
and 4, step 4: normalizing the post-X according to DBSCAN (0, k) modelm×nThe clustering result category number of the clustering is carried out, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
And 5: determining an upper threshold E of a neighborhood radius parameter E according to the k-distance graph drawn in the step 3max；
Step 6: root of herbaceous plantAccording to E0And EmaxDrawing a curve of class number-E;
and 7: determining the optimal value of the neighborhood radius parameter E according to the 'category number-E' curve;
and 8: clustering on-line monitoring data of the power equipment by using the DBSCAN model formed in the step 7, wherein the on-line monitoring data is used for mode identification and abnormal detection of on-line monitoring real-time data, and judging the type of normal data and abnormal data;
DBSCAN refers to density-based clustering.
2. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
in step 1, for Xm×nRespectively carrying out z-score standardization on n types of variables of each sample, wherein the standardization formula is as follows:
in the formula (I), the compound is shown in the specification,
is the normalized measurement value of the t sample of the j variable;
and σjThe mean and standard deviation of the j-th category variables, respectively.
3. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
in step 2, the density threshold parameter k of the DBSCAN model is set to 2 n.
4. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
the step 3 comprises the following steps:
step 3.1: x after calculation standardizationm×nEach sample inDistance of the specimen from the rest of the specimen;
step 3.2: calculating the average distance between each sample and the k samples closest to the sample;
step 3.3: and performing ascending sorting on the m average distance values, taking the obtained ascending sorting sequence number as an abscissa, and drawing a k distance graph by taking the corresponding m average distance values as an ordinate.
5. The power equipment state monitoring data clustering method according to claim 1 or 4, characterized in that:
and 3, the distance is the Euclidean distance.
6. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
in step 4, the DBSCAN (E, k) model pairs the standardized Xm×nThe clustering result category number for clustering begins to increase along with the increase of the E value and is more than N0When the value of E is E0Wherein N is0The category number of the clustering result of the DBSCAN (0, k) model.
7. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
step 5, finding out a point on the k-distance map when the ordinate value is half of the maximum ordinate value, further finding out a point on the k-distance map near the point where the change of the slope of the broken line is maximum, and determining the ordinate of the point as Emax。
8. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
step 6 comprises the following steps:
step 6.1: at E0And EmaxGenerates a logarithmic grid vector [ E (1), E (2), E (3), … E (9), E (10) containing 10 values](ii) a Wherein E (1) ═ E0，E(10)＝Emax；
Step 6.2: using DBSCAN (E (i), k) model pair standard under different values of E (i)After chemical treatment Xm×nThe sample data in (1) is clustered, and the number of cluster categories n (i) is counted, and a "category-E" curve is drawn with E (i) as the abscissa and the corresponding n (i) as the ordinate, where i is 1,2, …,10.
9. The power equipment state monitoring data clustering method according to claim 1, characterized in that:
in step 7, the first local minimum point on the right side of the maximum value in the "category number-E" curve or the left end point of the first horizontal line is determined as the optimal value of the neighborhood radius parameter E.
10. The utility model provides an electric power equipment state monitoring data clustering system, includes acquisition module, setting module, first drawing module, neighborhood radius parameter lower limit threshold determine module, neighborhood radius parameter upper limit threshold determine module, second drawing module and neighborhood radius parameter optimal value determine module and data clustering module, its characterized in that:
the acquisition module is used for acquiring an online monitoring data set X of the electric power equipment to be clusteredm×n，Xm×nContaining m samples, each sample containing n classes of variables, for Xm×nNormalizing the n-type variables of each sample;
the setting module is used for setting a density threshold parameter k of the DBSCAN model;
the first drawing module is used for drawing according to the standardized Xm×nPlotting a k-distance map of the distance of each sample from the rest of the samples;
the neighborhood radius parameter lower limit threshold determining module is used for determining the standardized X according to a DBSCAN (0, k) modelm×nThe category number of the clustering result is clustered, and the lower limit threshold value E of the neighborhood radius parameter E is determined0；
The neighborhood radius parameter upper limit threshold determining module is used for determining the upper limit threshold E of the neighborhood radius parameter E according to the k-distance map drawn in the step 3max；
The second drawing module is used for drawing according to E0And EmaxPlotting "number of categories-E "curve;
the neighborhood radius parameter optimal value determining module is used for determining the optimal value of a neighborhood radius parameter E according to a class number-E curve;
and the data clustering module is used for clustering the online monitoring data of the equipment by using the DBSCAN model determined by the neighborhood radius parameter optimal value determining module, and is used for performing mode identification and abnormal detection on online monitoring real-time data and judging the type of normal data and abnormal data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010481567.0A CN111582406A (en) | 2020-05-31 | 2020-05-31 | Power equipment state monitoring data clustering method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010481567.0A CN111582406A (en) | 2020-05-31 | 2020-05-31 | Power equipment state monitoring data clustering method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111582406A true CN111582406A (en) | 2020-08-25 |
Family
ID=72114362
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010481567.0A Pending CN111582406A (en) | 2020-05-31 | 2020-05-31 | Power equipment state monitoring data clustering method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111582406A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112381130A (en) * | 2020-11-10 | 2021-02-19 | 国网上海市电力公司 | Cluster analysis-based power distribution room multivariate data anomaly detection method |
| CN112565422A (en) * | 2020-12-04 | 2021-03-26 | 杭州佳速度产业互联网有限公司 | Method, system and storage medium for identifying fault data of power internet of things |
| CN113011472A (en) * | 2021-02-26 | 2021-06-22 | 广东电网有限责任公司电力调度控制中心 | Method and device for judging similarity of multi-section power quotation curves |
| CN113177597A (en) * | 2021-04-30 | 2021-07-27 | 平安国际融资租赁有限公司 | Model training data determination method, detection model training method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558821A (en) * | 2018-11-21 | 2019-04-02 | 哈尔滨工业大学（深圳） | The clothes article number calculating method of particular persons in a kind of video |
| CN109583763A (en) * | 2018-12-03 | 2019-04-05 | 国网浙江省电力有限公司台州供电公司 | Branch trade custom power load growth feature mining algorithm |
| CN110110802A (en) * | 2019-05-14 | 2019-08-09 | 南京林业大学 | Airborne laser point cloud classification method based on high-order condition random field |
| CN110263230A (en) * | 2019-04-25 | 2019-09-20 | 北京科技大学 | A kind of data cleaning method and device based on Density Clustering |
-
2020
- 2020-05-31 CN CN202010481567.0A patent/CN111582406A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558821A (en) * | 2018-11-21 | 2019-04-02 | 哈尔滨工业大学（深圳） | The clothes article number calculating method of particular persons in a kind of video |
| CN109583763A (en) * | 2018-12-03 | 2019-04-05 | 国网浙江省电力有限公司台州供电公司 | Branch trade custom power load growth feature mining algorithm |
| CN110263230A (en) * | 2019-04-25 | 2019-09-20 | 北京科技大学 | A kind of data cleaning method and device based on Density Clustering |
| CN110110802A (en) * | 2019-05-14 | 2019-08-09 | 南京林业大学 | Airborne laser point cloud classification method based on high-order condition random field |
Non-Patent Citations (1)
| Title |
|---|
| HANG LIU等: "Anomaly detection for condition monitoring data using auxiliary feature vector and density-based clustering" * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112381130A (en) * | 2020-11-10 | 2021-02-19 | 国网上海市电力公司 | Cluster analysis-based power distribution room multivariate data anomaly detection method |
| CN112565422A (en) * | 2020-12-04 | 2021-03-26 | 杭州佳速度产业互联网有限公司 | Method, system and storage medium for identifying fault data of power internet of things |
| CN112565422B (en) * | 2020-12-04 | 2022-07-22 | 杭州佳速度产业互联网有限公司 | Method, system and storage medium for identifying fault data of power internet of things |
| CN113011472A (en) * | 2021-02-26 | 2021-06-22 | 广东电网有限责任公司电力调度控制中心 | Method and device for judging similarity of multi-section power quotation curves |
| CN113011472B (en) * | 2021-02-26 | 2023-09-01 | 广东电网有限责任公司电力调度控制中心 | Multi-section electric power quotation curve similarity judging method and device |
| CN113177597A (en) * | 2021-04-30 | 2021-07-27 | 平安国际融资租赁有限公司 | Model training data determination method, detection model training method, device and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111582406A (en) | Power equipment state monitoring data clustering method and system | |
| CN108304567B (en) | Method and system for identifying working condition mode and classifying data of high-voltage transformer | |
| CN109298225B (en) | Automatic identification model system and method for abnormal state of voltage measurement data | |
| CN101738998B (en) | System and method for monitoring industrial process based on local discriminatory analysis | |
| CN112904810B (en) | Process industry nonlinear process monitoring method based on effective feature selection | |
| CN110795690A (en) | Wind power plant operation abnormal data detection method | |
| CN110889441B (en) | Power transformation equipment data anomaly identification method based on distance and point density | |
| CN106326915B (en) | A kind of Fault Diagnosis for Chemical Process method based on improvement core Fisher | |
| CN110930057A (en) | Quantitative evaluation method for reliability of distribution transformer test result based on LOF algorithm | |
| CN110794360A (en) | Method and system for predicting fault of intelligent electric energy meter based on machine learning | |
| CN111273125A (en) | RST-CNN-based power cable channel fault diagnosis method | |
| CN115409131A (en) | Production line abnormity detection method based on SPC process control system | |
| CN116780781B (en) | Power management method for smart grid access | |
| CN113128567A (en) | Abnormal electricity consumption behavior identification method based on electricity consumption data | |
| CN112417763A (en) | Defect diagnosis method, device and equipment for power transmission line and storage medium | |
| CN114997256A (en) | Method and device for detecting abnormal power of wind power plant and storage medium | |
| CN115526258A (en) | Power system transient stability evaluation method based on Spearman correlation coefficient feature extraction | |
| CN110689324A (en) | Auxiliary judging method for detection result of distribution transformer | |
| CN114580934A (en) | Early warning method for food detection data risk based on unsupervised anomaly detection | |
| CN113469252A (en) | Extra-high voltage converter valve operation state evaluation method considering unbalanced samples | |
| CN116990633A (en) | Fault studying and judging method based on multiple characteristic quantities | |
| CN110427019B (en) | Industrial process fault classification method and control device based on multivariate discriminant analysis | |
| CN110132566B (en) | OLTC fault diagnosis method based on fuzzy clustering | |
| CN111507374A (en) | Power grid mass data anomaly detection method based on random matrix theory | |
| CN111623905A (en) | Wind turbine bearing temperature early warning method and device |
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 |