CN113673579A - Power load classification algorithm based on small samples - Google Patents
Power load classification algorithm based on small samples Download PDFInfo
- Publication number
- CN113673579A CN113673579A CN202110853474.0A CN202110853474A CN113673579A CN 113673579 A CN113673579 A CN 113673579A CN 202110853474 A CN202110853474 A CN 202110853474A CN 113673579 A CN113673579 A CN 113673579A
- Authority
- CN
- China
- Prior art keywords
- samples
- algorithm
- sample
- classification
- training
- 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
- 238000007635 classification algorithm Methods 0.000 title claims abstract description 11
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 48
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000012360 testing method Methods 0.000 claims abstract description 13
- 238000012706 support-vector machine Methods 0.000 claims abstract description 11
- 238000003066 decision tree Methods 0.000 claims abstract description 9
- 230000008859 change Effects 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000005457 optimization Methods 0.000 claims abstract description 4
- 238000011156 evaluation Methods 0.000 claims description 14
- 238000000034 method Methods 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000007547 defect Effects 0.000 claims description 3
- 238000005315 distribution function Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000004445 quantitative analysis Methods 0.000 claims description 3
- 238000004904 shortening Methods 0.000 claims description 2
- 230000001747 exhibiting effect Effects 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000035622 drinking Effects 0.000 description 1
- 238000010348 incorporation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000011218 segmentation Effects 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
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- 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
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Water Supply & Treatment (AREA)
- Public Health (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)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a small sample-based power load classification algorithm, which comprises the steps of firstly, extracting the characteristics of different non-steady change load samples and preprocessing the samples, then expanding the samples through network training and evaluating the samples, then carrying out mixed training on simplified data by adopting a K-nearest neighbor algorithm, a support vector machine and decision tree mixed training, finally adjusting the weight of each algorithm on the classification result precision in a weighting optimization mode and evaluating the weights, and when the classification precision of a mixed model meets the condition, the algorithm can be used for actual sample testing. Has the following advantages: 1. the original sample has better representativeness and universal applicability; 2. expanding the consistency of the samples before and after expansion; 3. the mixed classifier and weighting mode is adopted, so that the limitation of a single algorithm can be overcome.
Description
Technical Field
The invention relates to an electric load classification algorithm, in particular to an electric load classification algorithm based on small samples.
Background
With the large-scale incorporation of clean energy such as wind power and photovoltaic and the large-scale use of electric vehicles in an electric power system, the electric power load components of the electric power system become more complex, the analysis of the electric power load components is beneficial to power grid personnel to formulate reasonable requirements and implement corresponding measures according to the electric power load rule, once the classification of the electric power load is unclear, the load prediction of an electric power department is wrong, the dispatching of the electric power system is disordered, the electric power marketing market rule is damaged seriously, and the economic loss of huge economy is caused. Therefore, it is very important to develop a classification study of power load in a power system.
At present, the electric load classification adopts the data drive's mode to analyze, and it has characteristics such as high efficiency, easy operation, reduction operation cost, nevertheless has the unbalanced, the low scheduling problem of analysis result precision of data classification, and the concrete expression is: 1. in an actual power grid, a normal sample of the power load is far larger than a sample of power load fluctuation, so that negative influence on the analysis performance of data driving is brought by the brought unbalanced data; 2. the analysis algorithm is mainly based on a single algorithm and an improved algorithm thereof, and has remarkable property of not being similar to that of the other algorithms, so that the analysis algorithm is easy to fall into a local optimal solution, and the reliability of an analysis result is greatly reduced.
Disclosure of Invention
The invention mainly solves the technical problems existing in the prior art: the method provides a sample expansion algorithm, expands a small number of power load samples which show non-steady change, and solves the problem of unbalanced sample distribution.
The invention also aims to solve the technical problems existing in the prior art: the state classification algorithm is provided, load data are classified through mixed analysis and comprehensive decision of various intelligent algorithms, the limitation of a single algorithm is overcome, and efficient recognition of various power load states is achieved.
The technical problem of the invention is mainly solved by the following technical scheme:
a small sample-based power load classification algorithm comprises the following steps:
step 1, sample expansion: the method comprises the steps of extracting the characteristics of different non-steadily changing load samples, preprocessing the samples, expanding the processed samples in a network training mode, judging the generated samples according to evaluation indexes, outputting the samples when the generated samples meet conditions, and forming a database of a state classification link together with other few non-steadily changing power load samples and steadily changing power load samples.
Step 2, state classification: the method comprises the steps of obtaining training data from a database, extracting features of different data, simplifying and shortening a classification process, adopting K-nearest algorithm, support vector machine and decision tree mixed training to make up for defects of a single algorithm, adjusting the weight of each algorithm on classification result precision in a weighting optimization mode, providing classification accuracy, combining evaluation indexes to judge, and being used for actual sample testing when the classification precision of a mixed model meets conditions.
In the above algorithm for classifying electrical loads based on small samples, in step 1, the specific operation method of sample expansion is as follows:
step 1.1, obtaining samples under different non-steady change electric load conditions in a certain time period in different time periods of 0-24 hours a day.
And 1.2, preprocessing samples with different characteristics, and eliminating influences caused by different dimensions, measurement intervals and the like in a normalization mode.
And 1.3, adding noise into the preprocessed sample set, establishing a linear model based on a normal distribution function, and considering that a proper sample can be generated when constraint conditions are met.
And step 1.4, introducing an evaluation index to carry out quantitative analysis on the extended sample set, and considering that the extended sample has higher reliability when the condition is met.
In the above algorithm for classifying electrical loads based on small samples, in step 2, the specific operation method for state classification is as follows:
and 2.1, acquiring data from the database, dividing the data into training data and test data according to a proportion, and extracting entropy characteristic values of the data.
And 2.2, performing mixed training by adopting KNN, SVM and DT according to the entropy characteristic value of the sample, and acquiring a training error of the single algorithm when the training precision of the single algorithm meets the requirement.
And 2.3, distributing the weights of the algorithms according to the training parameters of the hybrid classifier, wherein the hybrid training result is determined by the weighted result of each algorithm.
And 2.4, integrating the training data and the test data, quantitatively analyzing the mixed classification precision according to the evaluation index, and if the mixed classification precision meets the condition, testing the actual sample.
Therefore, the invention has the following advantages: 1. the diversity of non-stationary samples is considered, common types of power users are simplified, and the samples have good representativeness and universal applicability and are closer to the actual use condition; 2. the small samples are expanded, preprocessing, noise adding and evaluation are adopted, and when the characteristics of different small samples are considered, the difficulty of expanding the large samples by the small samples is reduced, and the consistency of the samples before and after expansion is ensured; 3. the load data is classified by adopting a hybrid classifier and a weighting mode and through hybrid analysis and comprehensive decision of various intelligent algorithms, the limitation of a single algorithm is overcome, and the high-efficiency recognition of various power load states is realized.
Drawings
FIG. 1 is a flow chart of the main implementation of the present invention.
FIG. 2 is a main flow chart of the present invention for power load sample expansion.
Fig. 3 is a main flow chart of the classification of the state of the electrical load according to the present invention.
Detailed Description
The technical scheme of the invention is further specifically described by the following embodiments and the accompanying drawings.
As shown in the attached figure 1, the method sequentially adopts a sample expansion mode and a state classification mode to realize accurate classification of the power load based on small samples, wherein the sample expansion expands a small number of power load samples which show non-steady change, the problem of unbalanced sample distribution is solved, the state classification classifies the load data through mixed analysis and comprehensive decision of various intelligent algorithms, the limitation of a single algorithm is overcome, and the high-efficiency identification of various power load states is realized.
As shown in figure 2, in the sample expansion link, the invention extracts the characteristics of different non-steadily changing load samples, preprocesses the samples, expands the processed samples in a network training mode, judges the generated samples according to the evaluation indexes, outputs the samples when the conditions are met, and forms a database of the state classification link together with other few non-steadily changing power load samples and steadily changing power load samples. The main steps of the link are as follows:
step 1.1, obtaining samples under different non-steady change electric load conditions in a certain time period in different time periods of 0-24 hours a day.
The invention divides the sample set of the user load of the power system into four categories: community residents ([ I)1,U1,P1,T1]、[I2,U2,P2,T2]、[I3,U3,P3,T3]) Individual merchant ([ I)4,U4,P4,T4]、[I5,U5,P5,T5]、[I6,U6,P6,T6]) Business unit ([ I)7,U7,P7,T7]、[I8,U8,P8,T8]、[I9,U9,P9,T9]) And large-scale enterprise ([ I)10,U10,P10,T10]、[I11,U11,P11,T11]、[I12,U12,P12,T12]) Wherein [ I ]1,U1,P1,T1]、[I4,U4,P4,T4]、[I7,U7,P7,T7]、[I10,U10,P10,T10]Representing community residents, individual merchants, public institutions and large enterprises in time period T1、T4、T7、T10Internal normal electrical load, I1、I4、I7、I10A set of currents, U, representing the user during the time period1、U4、U7、U10A set of voltages, P, representing the user during the time period1、P4、P7、P10Represents the set of powers of the users in the time period, [ I ]2,U2,P2,T2]、[I5,U5,P5,T5]、[I8,U8,P8,T8]、[I11,U11,P11,T11]Representing community residents, individual merchants, public institutions and large enterprises in time period T2、T5、T8、T11Electricity load at peak in the interior, I2、I5、I8、I11A set of currents, U, representing the user during the time period2、U5、U8、U11A set of voltages, P, representing the user during the time period2、P5、P8、P11Represents the set of powers of the users in the time period, [ I ]3,U3,P3,T3]、[I6,U6,P6,T6]、[I9,U9,P9,T9]、[I12,U12,P12,T12]Representing community residents, individual merchants, public institutions and large enterprises in time period T3、T6、T9、T12Electrical load in the interior valley, I3、I6、I9、I12A set of currents, U, representing the user during the time period3、U6、U9、U12A set of voltages, P, representing the user during the time period3、P6、P9、P12Representing the set of powers for the user during that time period.
And 1.2, preprocessing samples with different characteristics, and eliminating influences caused by different dimensions, measurement intervals and the like in a normalization mode.
B represents a normalized sample set, AiRepresenting a set of user loads, (A)i)maxRepresents the maximum load in the set (A)i)minRepresenting the minimum load in the set.
And 1.3, adding noise into the preprocessed sample set, establishing a linear model based on a normal distribution function, and considering that a proper sample can be generated when constraint conditions are met.
B′＝B+α (3)
B 'represents a set of samples after noise mixing, alpha represents a set of noise, f (B') represents a linear model, B1Representing coefficients of a Gaussian function, b2Represents the function offset, B represents the dimension of the sample set B,
represents the mean, δ, of the sample set BB'Represents the variance of the sample set B ', and f' (B ') represents the differentiation of the sample set B'.
In order to reduce the negative influence caused by the sample set and the noise set magnitude interval, when the constraint condition of the equation (6) is satisfied, the samples having consistency with the input samples are considered to be generated. When the constraint condition of the formula (6) is not satisfied, returning the formulas (4) and (5) to adjust b1、b2And (4) function coefficients until the constraint condition of the formula (6) is met. In particular, the present invention is not limited to a particular mode of function coefficient adjustment.
On the basis of acquiring different characteristic sample sets in step 1.1, the invention adopts a noise mixing mode to expand samples with fluctuation, establishes a linear model based on a Gaussian function in view of the universality of the Gaussian function in probability density distribution, forms a new sample in a reverse solving mode, and considers that a new sample consistent with an input sample is generated when a constraint condition is met, and is used for evaluating and judging whether the condition is met or not in step 1.4.
And step 1.4, introducing an evaluation index to carry out quantitative analysis on the extended sample set, and considering that the extended sample has higher reliability when the condition is met.
C3＝||f'(B')||2 (9)
C1Representing the degree of overlap of the set of samples f '(B') over the entire range, C4Represents the ratio of the mean of the sample set f '(B') to its minimum over the entire range, C5Represents the ratio of the mean of the sample set f '(B') to its maximum over the entire range, C2Represents the contour width, C, of the sample set f' (B3Represents the sum of squared distances of the sample set f '(B').
The invention proposes a global overlap C1Contour width C2Distance sum of squares C3As an evaluation index, among others, the global overlap C1The floating interval is [0,1 ]]Representing the similarity degree of the extended sample set and the original sample set, wherein the closer the value is to 1, the consistency between the extended sample generated by reverse solution and the original sample in a dimensional range is; width of contour C2The floating interval is [ -1,1 [)]Representing the closeness degree of the extended sample set curve and the original sample set curve, wherein the closer the value is to 1, the more the change trend of the load curve of the extended sample and the original sample is the same; sum of squares of distances C3The floating interval is [0, + ∞]And representing the tightness degree between the different feature samples after the expansion and the original drinking sample, wherein the closer the value is to 0, the tighter the different feature objects in the expansion sample are. In particular, the present invention is not limited to the criterion of qualification of a particular sample, i.e., making a decision with reference to actual conditions.
As shown in figure 3, in the state classification link, the method obtains training data from a database, extracts features of different data, simplifies and shortens the classification process, adopts mixed training of a K-nearest neighbor algorithm (KNN), a Support Vector Machine (SVM) and a Decision Tree (DT) to make up for the defects of a single algorithm, adjusts the weight of each algorithm on the precision of the classification result in a weighting optimization mode, provides the classification accuracy, combines an evaluation index for judgment, and can be used for testing an actual sample when the classification precision of a mixed model meets the condition. The main steps of the link are as follows:
and 2.1, acquiring data from the database, dividing the data into training data and test data according to a proportion, and extracting entropy characteristic values of the data.
D(B')＝∑f'(B')logf'(B') (10)
D (B') is the set of entropy feature values of the data.
Through the reverse solution in the step 1, the complexity of data information is greatly simplified, and the dimensional characteristics of the data can be further represented in an information entropy mode.
And 2.2, performing mixed training by adopting KNN, SVM and DT according to the entropy characteristic value of the sample, and acquiring a training error of the single algorithm when the training precision of the single algorithm meets the requirement.
ESVM＝β(xj,yj)+e1 (12)
EDT＝∑E1(xj,yj)＝e2 (13)
xj、yj∈D(B') (14)
Sample xj、yjAre derived from the entropy feature value set D (B'), EKNNRepresenting the sum of Euclidean distances of different sample points, ESVMA hyperplane segmentation curve representing the sample, beta represents the SVM kernel, e1Representing its bias term, as a constant, EDTEntropy decision function representing the sample, e2Representing the categories of different feature sample sets.
The invention provides a method for training entropy feature samples by using a hybrid classifier, namely, a KNN algorithm, an SVM algorithm and a DT algorithm are adopted to train the entropy feature samples, so that the reliability of the algorithm is improved.
And 2.3, distributing the weights of the algorithms according to the training parameters of the hybrid classifier, wherein the hybrid training result is determined by the weighted result of each algorithm.
χkRespectively corresponding to weight coefficients, delta, of KNN, SVM and DTkAnd R represents a voting result, when the voting result is 1, the detection result of KNN is taken as the standard, when the voting result is 2, the detection result of SVM is taken as the standard, and when the voting result is 3, the detection result of DT is taken as the standard.
The invention optimizes the data result by means of weighting adjustment on the basis of mixed classification, further improves the accuracy of the algorithm, and is particularly not limited to the error calculation mode of a single algorithm.
And 2.4, integrating the training data and the test data, quantitatively analyzing the mixed classification precision according to the evaluation index, and if the mixed classification precision meets the condition, testing the actual sample.
Having thus described the fundamental principles and advantages of the present invention, it will be appreciated by those skilled in the art that the present invention is not limited by the embodiments described above, which are merely illustrative of the structural relationships and principles of the present invention, and that various changes and modifications may be made without departing from the spirit and scope of the invention as hereinafter claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
1. A small sample-based power load classification algorithm is characterized by comprising the following steps:
step 1, sample expansion: extracting the characteristics of different non-steadily changing load samples, preprocessing the samples, expanding the processed samples in a network training mode, judging the generated samples according to evaluation indexes, outputting the samples when the generated samples meet the conditions, and forming a database of a state classification link together with other few non-steadily changing power load samples and steadily changing power load samples;
step 2, state classification: the method comprises the steps of obtaining training data from a database, extracting features of different data, simplifying and shortening a classification process, adopting K-nearest algorithm, support vector machine and decision tree mixed training to make up for defects of a single algorithm, adjusting the weight of each algorithm on classification result precision in a weighting optimization mode, providing classification accuracy, combining evaluation indexes to judge, and being used for actual sample testing when the classification precision of a mixed model meets conditions.
2. The algorithm for classifying electrical loads according to claim 1, wherein the sample expansion expands a small number of electrical load samples exhibiting non-stationary variation to solve the problem of unbalanced distribution of samples.
3. The small sample-based electric load classification algorithm as claimed in claim 1, wherein the state classification classifies the load data through hybrid analysis and comprehensive decision of multiple intelligent algorithms, overcomes the limitation of a single algorithm, and realizes efficient identification of multiple electric load states.
4. The small sample-based power load classification algorithm according to claim 1, wherein the specific operation method of the sample expansion is as follows:
step 1.1, obtaining samples under different non-steady change electric load conditions in a certain time period in different time periods of 0-24 hours a day.
And 1.2, preprocessing samples with different characteristics, and eliminating influences caused by different dimensions, measurement intervals and the like in a normalization mode.
And 1.3, adding noise into the preprocessed sample set, establishing a linear model based on a normal distribution function, and considering that a proper sample can be generated when constraint conditions are met.
And step 1.4, introducing an evaluation index to carry out quantitative analysis on the extended sample set, and considering that the extended sample has higher reliability when the condition is met.
5. The small sample-based power load classification algorithm according to claim 1, wherein the specific operation method of the state classification is as follows:
and 2.1, acquiring data from the database, dividing the data into training data and test data according to a proportion, and extracting entropy characteristic values of the data.
And 2.2, performing mixed training by adopting KNN, SVM and DT according to the entropy characteristic value of the sample, and acquiring a training error of the single algorithm when the training precision of the single algorithm meets the requirement.
And 2.3, distributing the weights of the algorithms according to the training parameters of the hybrid classifier, wherein the hybrid training result is determined by the weighted result of each algorithm.
And 2.4, integrating the training data and the test data, quantitatively analyzing the mixed classification precision according to the evaluation index, and if the mixed classification precision meets the condition, testing the actual sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110853474.0A CN113673579B (en) | 2021-07-27 | 2021-07-27 | Small sample-based electricity load classification algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110853474.0A CN113673579B (en) | 2021-07-27 | 2021-07-27 | Small sample-based electricity load classification algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113673579A true CN113673579A (en) | 2021-11-19 |
| CN113673579B CN113673579B (en) | 2024-05-28 |
Family
ID=78540532
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110853474.0A Active CN113673579B (en) | 2021-07-27 | 2021-07-27 | Small sample-based electricity load classification algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113673579B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160328837A1 (en) * | 2015-05-08 | 2016-11-10 | Kla-Tencor Corporation | Method and System for Defect Classification |
| CN108416366A (en) * | 2018-02-06 | 2018-08-17 | 武汉大学 | A kind of power-system short-term load forecasting method of the weighting LS-SVM based on Meteorological Index |
| CN108537281A (en) * | 2018-04-13 | 2018-09-14 | 贵州电网有限责任公司 | A kind of power consumer feature recognition sorting technique based on random forest |
| WO2019033636A1 (en) * | 2017-08-16 | 2019-02-21 | 哈尔滨工业大学深圳研究生院 | Method of using minimized-loss learning to classify imbalanced samples |
| CN109447364A (en) * | 2018-11-08 | 2019-03-08 | 国网湖南省电力有限公司 | Power customer based on label complains prediction technique |
| CN110334661A (en) * | 2019-07-09 | 2019-10-15 | 国网江苏省电力有限公司扬州供电分公司 | Infrared power transmission and transformation abnormal heating point target detecting method based on deep learning |
| WO2019237492A1 (en) * | 2018-06-13 | 2019-12-19 | 山东科技大学 | Semi-supervised learning-based abnormal electricity utilization user detection method |
| CN111832615A (en) * | 2020-06-04 | 2020-10-27 | 中国科学院空天信息创新研究院 | Sample expansion method and system based on foreground and background feature fusion |
| CN112801123A (en) * | 2020-09-09 | 2021-05-14 | 华北电力大学 | Small sample user electricity consumption data expansion method with frequency domain distribution consistency |
-
2021
- 2021-07-27 CN CN202110853474.0A patent/CN113673579B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160328837A1 (en) * | 2015-05-08 | 2016-11-10 | Kla-Tencor Corporation | Method and System for Defect Classification |
| WO2019033636A1 (en) * | 2017-08-16 | 2019-02-21 | 哈尔滨工业大学深圳研究生院 | Method of using minimized-loss learning to classify imbalanced samples |
| CN108416366A (en) * | 2018-02-06 | 2018-08-17 | 武汉大学 | A kind of power-system short-term load forecasting method of the weighting LS-SVM based on Meteorological Index |
| CN108537281A (en) * | 2018-04-13 | 2018-09-14 | 贵州电网有限责任公司 | A kind of power consumer feature recognition sorting technique based on random forest |
| WO2019237492A1 (en) * | 2018-06-13 | 2019-12-19 | 山东科技大学 | Semi-supervised learning-based abnormal electricity utilization user detection method |
| CN109447364A (en) * | 2018-11-08 | 2019-03-08 | 国网湖南省电力有限公司 | Power customer based on label complains prediction technique |
| CN110334661A (en) * | 2019-07-09 | 2019-10-15 | 国网江苏省电力有限公司扬州供电分公司 | Infrared power transmission and transformation abnormal heating point target detecting method based on deep learning |
| CN111832615A (en) * | 2020-06-04 | 2020-10-27 | 中国科学院空天信息创新研究院 | Sample expansion method and system based on foreground and background feature fusion |
| CN112801123A (en) * | 2020-09-09 | 2021-05-14 | 华北电力大学 | Small sample user electricity consumption data expansion method with frequency domain distribution consistency |
Non-Patent Citations (2)
| Title |
|---|
| 李想;王鹏;刘洋;许立雄;: "考虑类别不平衡的海量负荷用电模式辨识方法", 中国电机工程学报, no. 01 * |
| 潘明明;刘连光;田世明;徐震;: "基于LS-SVM的重要用户供用电安全评价分析方法", 发电与空调, no. 03, 15 June 2016 (2016-06-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113673579B (en) | 2024-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106650767B (en) | Flood forecasting method based on cluster analysis and real-time correction | |
| CN106845717B (en) | Energy efficiency evaluation method based on multi-model fusion strategy | |
| CN112561156A (en) | Short-term power load prediction method based on user load mode classification | |
| CN109492026B (en) | Telecommunication fraud classification detection method based on improved active learning technology | |
| CN104573879A (en) | Photovoltaic power station output predicting method based on optimal similar day set | |
| CN110909977A (en) | Power grid fault diagnosis method based on ADASYN-DHSD-ET | |
| CN104463202A (en) | Multi-class image semi-supervised classifying method and system | |
| CN104657718A (en) | Face recognition method based on face image feature extreme learning machine | |
| CN113469253A (en) | Electricity stealing detection method based on triple twin network | |
| CN110738232A (en) | grid voltage out-of-limit cause diagnosis method based on data mining technology | |
| CN112418476A (en) | Ultra-short-term power load prediction method | |
| CN111723876A (en) | Load curve integrated spectrum clustering algorithm considering double-scale similarity | |
| CN115829105A (en) | Photovoltaic power prediction method based on historical data feature search | |
| CN115099296A (en) | Sea wave height prediction method based on deep learning algorithm | |
| CN112287980B (en) | Power battery screening method based on typical feature vector | |
| CN105678406A (en) | Short-term load prediction method based on cloud model | |
| CN113283491A (en) | Fault diagnosis method of electric vehicle alternating current charging pile based on optimized deep confidence network | |
| CN115099511A (en) | Photovoltaic power probability estimation method and system based on optimized copula | |
| CN115660182A (en) | Photovoltaic output prediction method based on maximum expected sample weighted neural network model | |
| Omar et al. | Seasonal clustering forecasting technique for intelligent hourly solar irradiance systems | |
| CN117094754B (en) | Macroscopic and microscopic combined medium-long term electric quantity prediction method | |
| CN117332288A (en) | Building energy saving technology evaluation system and method | |
| CN113673579A (en) | Power load classification algorithm based on small samples | |
| CN114841266A (en) | Voltage sag identification method based on triple prototype network under small sample | |
| Qin | Software reliability prediction model based on PSO and SVM |
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 |