CN111582357A - Electric power fingerprint identification method based on multi-dimensional integration - Google Patents
Electric power fingerprint identification method based on multi-dimensional integration Download PDFInfo
- Publication number
- CN111582357A CN111582357A CN202010367254.2A CN202010367254A CN111582357A CN 111582357 A CN111582357 A CN 111582357A CN 202010367254 A CN202010367254 A CN 202010367254A CN 111582357 A CN111582357 A CN 111582357A
- Authority
- CN
- China
- Prior art keywords
- classifier
- probability
- appliance
- result
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 230000010354 integration Effects 0.000 title claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 31
- 230000001052 transient effect Effects 0.000 claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims description 32
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000012937 correction Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000003066 decision tree Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000012706 support-vector machine Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 description 6
- 230000006872 improvement Effects 0.000 description 5
- 238000007689 inspection Methods 0.000 description 5
- 230000006399 behavior Effects 0.000 description 3
- 238000013145 classification model Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000049 pigment Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
- G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Probability & Statistics with Applications (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a multi-dimensional integration-based power fingerprint identification method, which comprises the following steps of: step S1, acquiring power fingerprint data of the electrical appliance; step S2, training a classifier C1 according to the time dimension data set; s3, training a classifier C2 according to the transient dimension data set; s4, training a classifier C3 according to the transient dimension data set; step S5, inputting the data of the electric appliance to be identified into classifiers C1, C2 and C3 to obtain results; step S6, correcting the initial probability of each electric appliance type according to the output result of the classifier to obtain the posterior probability; s7, comparing the posterior probabilities of all the electric appliances, selecting the largest one as the recognition result, and finishing the recognition; in step S8, the result obtained in step S7 is checked, and if the result passes the check, the recognition is completed.
Description
Technical Field
The invention relates to the field of electrical load identification, in particular to a multi-dimensional integration-based power fingerprint identification method.
Background
The current proposed methods for identifying power loads mainly focus on identifying transient characteristics, steady-state characteristics or user behavior characteristics of the electrical appliances by using different machine learning algorithms and artificial rules.
A convolutional neural network-based non-invasive load recognition method (tanguo, pigment, mild, tang army. a convolutional neural network-based non-invasive load recognition method [ J ]. yunnan power technology, 2019,47(02):2-4+10.) proposed by tanguo et al, by collecting discrete power points, and then training a recognition model with a convolutional neural network; an LSTM model-based power load identification method (Liu Heng Yong Li, Liu Yong Li, Deng Shi Smart, Schaibin, Min Lin, Zhou Dong Guo.) proposed by Liu Heng Yong, et al (an LSTM model-based power load identification method [ J ]. an electrical measuring and measuring instrument, 2019,56(23):62-69.), monitors a load event by a Gauss window moving variable point optimizing algorithm, extracts a harmonic component as a load characteristic label and identifies the load as the input of the LSTM model; hua liang et al proposed a DTW algorithm-based non-invasive home load behavior recognition method (Hualiang, Huangwei, Yang Zi power, Wang Yu, Zhang Jia.) the DTW algorithm-based non-invasive home load behavior recognition method [ J ] an electrical measurement and instrument, 2019,56(14):17-22.), which recognized the load by collecting the sudden change of the steady-state power waveform of the device and using the dynamic time warping algorithm to match the database. The methods have the following two main characteristics in general: the method has the advantages that the identification rate is high, the identification rate of individual appliances is low, the identification rate of self samples is high, and the generalization capability is poor, and the methods are essentially to use single-dimension data to identify loads without considering the different identification degrees of the loads in different dimensions. The electric power fingerprint identification technology is developed from an electric power load identification technology, the defect problems of identification models under different algorithms and different characteristics are made up by integrating the identification results of a plurality of dimensional characteristics, the high overall identification rate can be kept, and the problem of low identification rate of part of electric appliances can be solved.
Disclosure of Invention
Based on the method, the invention provides a multi-dimensional integration-based power fingerprint identification method. The method is based on the Bayesian principle, can solve the comprehensive problem of a plurality of dimensional feature recognition results, and has certain flexibility and expansibility; meanwhile, the magnification factor of the probability is used as the judgment basis of the recognition result, so that the improvement effect of the method on the probability of a correct result can be embodied, and the problem that the electric appliance cannot be recognized due to a small initial probability can be solved; and finally, a novel inspection mode is provided, wherein the result is substituted into a classifier with higher recall ratio for observation.
A multi-dimensional integration-based power fingerprint identification method comprises the following steps:
step S1, acquiring power fingerprint data of the electrical appliance;
step S2, time dimension data set D obtained in step S11Training classifier C1And calculating the probability matrix M of each electrical appliance1；
Step S3, the transient dimension data set D obtained in the step S12Training classifier C2And calculating the probability matrix M of each electrical appliance2；
Step S4, the steady dimensional data set D obtained in the step S13Training classifier C3And calculating the probability matrix M of each electrical appliance3；
Step S5, inputting the data of the electric appliance to be identified into a classifier C1Classifier C2Classifier C3In turn, the classification result R is obtained1、R2、R3；
Step S6, inquiring a probability matrix M according to the classification result obtained in the step S51、M2、M3Corresponding data, correcting the initial probability P of each electric appliance category i in turni 0Obtaining the posterior probability P of each electrical appliancei 1；
Step S7, P obtained according to step S5i 1Compared with the initial probability Pi 0Selecting the maximum magnification α as the recognition result;
and step S8, checking the result obtained in step S7, if the result passes the check, finishing the identification, otherwise, the result does not pass the check.
Further, the power fingerprint data acquired at step S1 includes: electric appliance category set A containing N electric appliance categories and initial probability P corresponding to the electric appliance categoriesi 0And a time dimension data set D for all appliance categories in the set A1Transient dimensional data setD2Steady state dimensional data set D3。
Further, the time dimension data set D1Including but not limited to: type A of electric applianceiThe starting time T of the electrical appliance and the duration time T of the electrical appliance1Daily opening frequency f of the appliance, where T and T1In an arbitrary time format, AiExpressed as the ith category;
the transient dimensional data set D2Including but not limited to: type A of electric applianceiAnd transient waveform S at the time of opening electric appliance1And transient waveform S at the closing time of the electric appliance2Transient duration T of the electrical appliance2A transient overcurrent multiple β, wherein the transient waveform S is at the moment of turning on the electrical appliance1And the transient waveform S at the closing time of the electric appliance2Collecting at any sampling frequency;
the steady dimensional data set D3Including but not limited to: type A of electric applianceiActive power P, reactive power Q, apparent power S and power factors during stable operation of electric appliance
Voltage harmonic HVCurrent harmonic HCWherein the voltage harmonic and the current harmonic are in phasor form, and the harmonic frequency is 2-11 or more.
Further, the classifier C of step S21Through a time dimension data set D1Training to obtain, classifier C1The specific form of the method is a Bayesian classifier, a BP (back propagation) neural network or a decision tree; probability matrix M1Is an N × N matrix, in which the j row and k column elements represent the classifier C in the training process1The recognition result for the training data is j, but actually the ratio of k.
Further, the classifier C of step S32By transient dimensional data set D2Training to obtain, classifier C2The specific form is a convolution neural network or a BP neural network; probability matrix M2Is an N × N matrix, wherein the j row and k column elements represent the number of training elements,classifier C2The recognition result for the training data is j, but actually the ratio of k.
Further, the classifier C of step S43Through a time dimension data set D3Training to obtain, classifier C3The specific form is a deep neural network, a random deep forest, a long and short memory network or a support vector machine; probability matrix M3Is an N × N matrix, in which the j row and k column elements represent the classifier C in the training process3The recognition result for the training data is j, but actually the ratio of k.
Further, step S5 inputs data of the appliance to be recognized into the classifier C1～C3Obtaining classification results R in sequence1～R3Wherein the classification result R1、R2、R3Are the categories in the appliance category set a.
Further, step S6 is to obtain the classification result and the probability matrix M according to step S51、M2、M3Correcting the initial probability P of each appliance class ii 0Obtaining the posterior probability P of each electrical appliancei 1：
Wherein: pi'、PiIs a passing probability matrix M1Probability matrix M2Posterior probability of appliance class i after correction, an intermediate quantity in the correction process, M1(R1,i)、M2(R2,i)、M3(R3I) are respectively probability matrices M1、M2、M3Rn row and i column elements, n ═ 1,2, 3.
Further, step S7 is based on P obtained in step S5i 1Comparing the initial probability Pi 0The maximum magnification α is selected as the recognition result R:
further, step S8 substitutes the result obtained in step S7 into the recall REiHighest classifier ClIf the result of the recognition result R is identical to the result of the recognition result R', the result is checked to obtain the recall ratio REiThe calculation is as follows:
wherein M is1(j,i)、M2(i,i)、M3(i, i) are respectively probability matrix M1、M2、M3Row jth and column ith elements.
The electric power fingerprint identification method based on multi-dimensional integration is described. The method is based on the Bayesian principle, can solve the comprehensive problem of a plurality of dimensional feature recognition results, and has certain flexibility and expansibility; meanwhile, the magnification factor of the probability is used as the judgment basis of the recognition result, so that the improvement effect of the method on the probability of a correct result can be embodied, and the problem that the electric appliance cannot be recognized due to a small initial probability can be solved; and finally, a novel inspection mode is provided, wherein the result is substituted into a classifier with higher recall ratio R for observation.
Compared with the prior art, the invention has the following advantages and effects:
(1) the method provided by the invention considers that a plurality of data dimensions participate in identification together, can integrate the advantages of a plurality of dimension classification models, and solves the problem that a single dimension classification model cannot classify two similar electrical appliances.
(2) The invention takes the ratio of the number of each electric appliance to the number of all electric appliances in reality as the initial probability, takes the probability of each electric appliance in the actual identification process into consideration, and has more practicability.
(3) The method takes the magnification of the probability as the judgment basis of the recognition result, can reflect the improvement effect of the method on the probability of the correct result, and can solve the problem that the initial probability of certain electrical appliances is small so that the electrical appliances cannot be recognized.
(4) The invention provides a novel inspection mode for substituting the result into a classifier with higher recall ratio R for observation, and the result is further confirmed by utilizing a model with high recall ratio, so that the error recognition rate is reduced.
Drawings
Fig. 1 is a flowchart of a power fingerprint identification method based on multidimensional integration according to this embodiment.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited to these examples.
The electric power fingerprint identification method based on multi-dimensional integration provided by the invention is based on the Bayesian principle, can solve the comprehensive problem of a plurality of dimensional feature identification results, and has certain flexibility and expansibility; meanwhile, the magnification factor of the probability is used as the judgment basis of the recognition result, so that the improvement effect of the method on the probability of a correct result can be embodied, and the problem that the electric appliance cannot be recognized due to a small initial probability can be solved; and finally, a novel inspection mode is provided, wherein the result is substituted into a classifier with higher recall ratio R for observation. The method comprises the following steps:
step S1, obtaining data through the household smart meter and the household smart socket, including: electric appliance category set A containing N electric appliance categories and initial probability P corresponding to the electric appliance categoriesi 0And a time dimension data set D for all appliance categories in the set A1Transient dimensional data set D2Steady state dimensional data set D3。
A ═ 1,2,3, { fan, computer, hair dryer }, and1i.e. a fan, the same applies. Initial probability P of fan1 0Initial probability of computer P of 0.52 00.2, initial probability of hair dryer P3 0＝0.3。
Time dimension data set D1～D3Is a data set containing a fan, a computer, a hair dryer label, D1Is characterized by the time of opening the appliance, the duration of the appliance, the daily opening frequency of the appliance, D2Is characterized by a transient waveform S at the turn-on time of the electric appliance1And transient waveform S at the closing time of the electric appliance2Transient duration T of the electrical appliance2Transient overcurrent multiple β, D3Is characterized by active power P, reactive power Q, apparent power S and power factor when the electric appliance is in stable operation
Voltage harmonic HVCurrent harmonic HCWherein the voltage harmonic and the current harmonic are in phasor form, and the harmonic frequency is 2-11.
Step S2, time dimension data set D obtained in step S11Training classifier C1And calculating the probability matrix M of each electrical appliance1；
Step S3, the transient dimension data set D obtained in the step S12Training classifier C2And calculating the probability matrix M of each electrical appliance2：
Step S4, the steady dimensional data set D obtained in the step S13Training classifier C3And calculating the probability matrix M of each electrical appliance3：
Step S5, classifying the data input of the electric appliance to be identifiedDevice C1～C3In turn, the classification result R is obtained1～R3；
The data of the electrical appliance to be identified are as follows:
the time dimension is as follows:
d1the time of opening is 12h, the duration of the electric appliance is 2h, and the daily opening frequency of the electric appliance is 1
Transient dimension:
d2open transient waveform s1Turn off transient waveform s2Transient duration 0.05s, overcurrent multiple 2.5 steady-state dimension:
d3active power is 40W, reactive power is 30W, apparent power is 50W, and power factor is 0.8;
will d1Input classifier C1Output the result R1＝A1(ii) a Will d2Input classifier C2Output the result R2＝A2(ii) a Will d3Input classifier C3Output the result R3＝A2。
Step S6, inquiring a probability matrix M according to the classification result obtained in the step S51、M2、M3Corresponding data, correcting the initial probability P of each electric appliance category i in turni 0Obtaining the posterior probability P of each electrical appliancei 1
Calculated to obtain P1 1Calculate P by analogy with 85.263%2 1＝0.271％，P3 1＝0.051％
Step S7. Step S7 obtaining posterior probability P of each electric appliance according to step S5i 1Comparing the initial probability Pi 0Selecting a magnification αiMaximum appliance type AiAs a recognition result R.
Calculated α1＝1.7053，α2＝0.0135，α20.00170, then R is equal to AiAn electric fan.
Step S8, step S8 obtaining result A according to step S7iSubstituting into result AiIs RE ofiHighest classifier Cl(1, 2,3), the result R 'is output, and if the recognition result R matches the result of the output node R', the result passes the test.
Results A1Has a recall ratio of C1Classifiers, i.e. R ═ R1＝A1And R ═ AiAnd comparing the recognition result R with the output node R', wherein the result is consistent, and the output result is the electric fan.
Through the steps, the initial probability of the electric appliance can be solved, the advantages of each dimension classification model are combined with the data of multiple dimensions, the initial probability of each electric appliance is corrected to obtain the posterior probability, and the most possible electric appliance category is deduced to be used as the recognition result. The magnification factor of the probability is used as the judgment basis of the recognition result, so that the improvement effect of the method on the probability of a correct result can be embodied, and the problem that the electric appliance cannot be recognized due to a small initial probability can be solved; and finally, a novel inspection mode is provided, wherein the result is substituted into a classifier with higher recall ratio R for observation.
The above examples are merely illustrative of the embodiments of the present invention, and the description thereof is more specific and detailed, but not to be construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the appended claims.
Claims (10)
1. A multi-dimensional integration-based power fingerprint identification method is characterized by comprising the following steps:
step S1, acquiring power fingerprint data of the electrical appliance;
step S2, time dimension data set D obtained in step S11Training classifier C1And calculating the probability matrix M of each electrical appliance1；
Step S3, the transient dimension data set D obtained in the step S12Training classifier C2And calculating the probability matrix M of each electrical appliance2；
Step S4, the steady dimensional data set D obtained in the step S13Training classifier C3And calculating the probability matrix M of each electrical appliance3；
Step S5, inputting the data of the electric appliance to be identified into a classifier C1Classifier C2Classifier C3In turn, the classification result R is obtained1、R2、R3；
Step S6, inquiring a probability matrix M according to the classification result obtained in the step S51、M2、M3Corresponding data, correcting the initial probability P of each electric appliance category i in turni 0Obtaining the posterior probability P of each electrical appliancei 1；
Step S7, P obtained according to step S5i 1Compared with the initial probability Pi 0Selecting the maximum magnification α as the recognition result;
and step S8, checking the result obtained in step S7, if the result passes the check, finishing the identification, otherwise, the result does not pass the check.
2. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: the power fingerprint data acquired at step S1 includes: electric appliance category set A containing N electric appliance categories and electric applianceInitial probability P corresponding to categoryi 0And a time dimension data set D for all appliance categories in the set A1Transient dimensional data set D2Steady state dimensional data set D3。
3. The multi-dimensional integration-based power fingerprint identification method according to claim 2, wherein: the time dimension data set D1Including but not limited to: type A of electric applianceiThe starting time T of the electrical appliance and the duration time T of the electrical appliance1Daily opening frequency f of the appliance, where T and T1In an arbitrary time format, AiExpressed as the ith category;
the transient dimensional data set D2Including but not limited to: type A of electric applianceiAnd transient waveform S at the time of opening electric appliance1And transient waveform S at the closing time of the electric appliance2Transient duration T of the electrical appliance2A transient overcurrent multiple β, wherein the transient waveform S is at the moment of turning on the electrical appliance1And the transient waveform S at the closing time of the electric appliance2Collecting at any sampling frequency;
the steady dimensional data set D3Including but not limited to: type A of electric applianceiActive power P, reactive power Q, apparent power S and power factors during stable operation of electric appliance
Voltage harmonic HVCurrent harmonic HCWherein the voltage harmonic and the current harmonic are in phasor form, and the harmonic frequency is 2-11 or more.
4. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: classifier C of step S21Through a time dimension data set D1Training to obtain, classifier C1The specific form of the method is a Bayesian classifier, a BP (back propagation) neural network or a decision tree; probability matrix M1Is an N × N matrix, wherein the jth row isThe elements of k columns represent the classifier C in the training process1The recognition result for the training data is j, but actually the ratio of k.
5. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: classifier C of step S32By transient dimensional data set D2Training to obtain, classifier C2The specific form is a convolution neural network or a BP neural network; probability matrix M2Is an N × N matrix, in which the j row and k column elements represent the classifier C in the training process2The recognition result for the training data is j, but actually the ratio of k.
6. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: classifier C of step S43Through a time dimension data set D3Training to obtain, classifier C3The specific form is a deep neural network, a random deep forest, a long and short memory network or a support vector machine; probability matrix M3Is an N × N matrix, in which the j row and k column elements represent the classifier C in the training process3The recognition result for the training data is j, but actually the ratio of k.
7. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: step S5 inputs data of the electric appliance to be recognized into the classifier C1～C3Obtaining classification results R in sequence1～R3Wherein the classification result R1、R2、R3Are the categories in the appliance category set a.
8. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: step S6 is to obtain the classification result and the probability matrix M according to the step S51、M2、M3Correcting the initial probability P of each appliance class ii 0To obtain each electrical applianceA posteriori probability P ofi 1：
Wherein: pi'、PiIs a passing probability matrix M1Probability matrix M2Posterior probability of appliance class i after correction, an intermediate quantity in the correction process, M1(R1,i)、M2(R2,i)、M3(R3I) are respectively probability matrices M1、M2、M3Rn row and i column elements, n ═ 1,2, 3.
9. The multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: step S7 obtaining P from step S5i 1Comparing the initial probability Pi 0The maximum magnification α is selected as the recognition result R:
10. the multi-dimensional integration-based power fingerprint identification method according to claim 1, wherein: step S8 substituting the result obtained in step S7 into recall REiHighest classifier ClIf the result of the recognition result R is identical to the result of the recognition result R', the result is checked to obtain the recall ratio REiThe calculation is as follows:
wherein M is1(j,i)、M2(i,i)、M3(i, i) are respectively probability matrix M1、M2、M3Row jth and column ith elements.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010367254.2A CN111582357B (en) | 2020-04-30 | 2020-04-30 | Electric fingerprint identification method based on multidimensional integration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010367254.2A CN111582357B (en) | 2020-04-30 | 2020-04-30 | Electric fingerprint identification method based on multidimensional integration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111582357A true CN111582357A (en) | 2020-08-25 |
| CN111582357B CN111582357B (en) | 2023-05-26 |
Family
ID=72124639
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010367254.2A Active CN111582357B (en) | 2020-04-30 | 2020-04-30 | Electric fingerprint identification method based on multidimensional integration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111582357B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113033633A (en) * | 2021-03-12 | 2021-06-25 | 贵州电网有限责任公司 | Equipment type identification method combining power fingerprint knowledge and neural network |
| CN113884782A (en) * | 2021-08-24 | 2022-01-04 | 国网天津市电力公司营销服务中心 | Method and system for identifying starting characteristic of microwave oven, computer equipment and storage medium |
| WO2022141330A1 (en) * | 2020-12-31 | 2022-07-07 | Guizhou Power Grid Company Limited | A non-intrusive load identification method based on fingerprint characteristics of load power |
| CN116861316A (en) * | 2023-09-04 | 2023-10-10 | 国网浙江省电力有限公司余姚市供电公司 | Electrical appliance monitoring method and device |
| CN117132843A (en) * | 2023-10-26 | 2023-11-28 | 长春中医药大学 | Wild ginseng, under-forest mountain ginseng and garden ginseng in-situ identification method, system and related equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN110287124A (en) * | 2019-07-03 | 2019-09-27 | 大连海事大学 | A kind of automatic marker software error reporting simultaneously carries out seriousness and knows method for distinguishing |
| CN110544177A (en) * | 2019-08-15 | 2019-12-06 | 广州水沐青华科技有限公司 | Load identification method based on power fingerprint and computer readable storage medium |
-
2020
- 2020-04-30 CN CN202010367254.2A patent/CN111582357B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN110287124A (en) * | 2019-07-03 | 2019-09-27 | 大连海事大学 | A kind of automatic marker software error reporting simultaneously carries out seriousness and knows method for distinguishing |
| CN110544177A (en) * | 2019-08-15 | 2019-12-06 | 广州水沐青华科技有限公司 | Load identification method based on power fingerprint and computer readable storage medium |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022141330A1 (en) * | 2020-12-31 | 2022-07-07 | Guizhou Power Grid Company Limited | A non-intrusive load identification method based on fingerprint characteristics of load power |
| CN113033633A (en) * | 2021-03-12 | 2021-06-25 | 贵州电网有限责任公司 | Equipment type identification method combining power fingerprint knowledge and neural network |
| CN113884782A (en) * | 2021-08-24 | 2022-01-04 | 国网天津市电力公司营销服务中心 | Method and system for identifying starting characteristic of microwave oven, computer equipment and storage medium |
| CN116861316A (en) * | 2023-09-04 | 2023-10-10 | 国网浙江省电力有限公司余姚市供电公司 | Electrical appliance monitoring method and device |
| CN116861316B (en) * | 2023-09-04 | 2023-12-15 | 国网浙江省电力有限公司余姚市供电公司 | Electrical appliance monitoring method and device |
| CN117132843A (en) * | 2023-10-26 | 2023-11-28 | 长春中医药大学 | Wild ginseng, under-forest mountain ginseng and garden ginseng in-situ identification method, system and related equipment |
| CN117132843B (en) * | 2023-10-26 | 2024-04-09 | 长春中医药大学 | Wild ginseng, under-forest mountain ginseng and garden ginseng in-situ identification method, system and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111582357B (en) | 2023-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111582357A (en) | Electric power fingerprint identification method based on multi-dimensional integration | |
| Zhang et al. | Dependency analysis and improved parameter estimation for dynamic composite load modeling | |
| CN112699913A (en) | Transformer area household variable relation abnormity diagnosis method and device | |
| US8756181B2 (en) | System and method employing a self-organizing map load feature database to identify electric load types of different electric loads | |
| CN111368904B (en) | Electrical equipment identification method based on electric power fingerprint | |
| US20130138669A1 (en) | System and method employing a hierarchical load feature database to identify electric load types of different electric loads | |
| Miranian et al. | Nonlinear power system load identification using local model networks | |
| CN113408341B (en) | Load identification method and device, computer equipment and storage medium | |
| CN108875783A (en) | A kind of extreme learning machine Diagnosis Method of Transformer Faults towards unbalanced dataset | |
| CN110220602A (en) | A kind of switchgear overheating fault recognition methods | |
| CN112085111A (en) | Load identification method and device | |
| CN117272116A (en) | Transformer fault diagnosis method based on LORAS balance data set | |
| Ren et al. | Application of RBF neural network optimized based on K-means cluster algorithm in fault diagnosis | |
| CN115169735A (en) | Electroceramic quality prediction method and system based on sample expansion | |
| CN114722915A (en) | Fault diagnosis method and system based on ADASYN algorithm and random forest algorithm | |
| Jamadi et al. | New method for accurate parameter estimation of induction motors based on artificial bee colony algorithm | |
| CN111242146B (en) | POI information classification based on convolutional neural network | |
| Zhou et al. | Transient stability assessment of large-scale AC/DC hybrid power grid based on separation feature and deep belief networks | |
| Chen et al. | Consumer-transformer relationship identification based on UMAP and mini batch K-means | |
| CN105913010A (en) | Electric appliance type determination device | |
| Ibrahem et al. | Genetic programming based face recognition | |
| CN110569331A (en) | Context-based relevance prediction method and device and storage equipment | |
| Yin et al. | Novel sparse representation classification with a closed-form solution | |
| CN116415510B (en) | Breaker temperature rise prediction method and system based on phase space reconstruction and neural network | |
| CN114972914B (en) | Power system dominant instability mode identification method based on Markov transfer field |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |