CN109324250A - A kind of Power Quality Disturbance recognition methods - Google Patents
A kind of Power Quality Disturbance recognition methods Download PDFInfo
- Publication number
- CN109324250A CN109324250A CN201811447626.1A CN201811447626A CN109324250A CN 109324250 A CN109324250 A CN 109324250A CN 201811447626 A CN201811447626 A CN 201811447626A CN 109324250 A CN109324250 A CN 109324250A
- Authority
- CN
- China
- Prior art keywords
- power quality
- quality disturbance
- vector
- feature vector
- cluster centre
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
Abstract
The present invention provides a kind of Power Quality Disturbance recognition methods.A kind of Power Quality Disturbance recognition methods, wherein include the following steps: that S1. obtains known Power Quality Disturbance;S2. wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extracts feature vector;S3. the feature vector of the Power Quality Disturbance based on step S2 classifies to Power Quality Disturbance based on fuzzy clustering, and obtains each cluster centre vector;S4. Power Quality Disturbance to be identified is extracted into feature vector by step S2, and calculates separately the distance between each cluster centre vector obtained with step S3, and then judge the type of Power Quality Disturbance to be identified.The advantage of method of the invention based on fuzzy clustering, can be improved accuracy of identification.
Description
Technical field
The present invention relates to power quality technical fields, more particularly, to a kind of Power Quality Disturbance recognition methods.
Background technique
In recent years, with the extensive use of power electronic equipment in power grid and photovoltaic, wind-powered electricity generation distributed power supply it is big
Amount access, caused by power quality problem become increasingly conspicuous.In Precision Machining and people's life is required to power quality day
Under conditions of becoming stringent, power quality need to be interfered and carry out precise classification, so analyze the reason of causing electrical energy power quality disturbance and
Take corresponding control measures.
Classification of Power Quality Disturbances includes feature extraction and classifier two steps of classification.In terms of classifier selection, with
BP neural network is that the neural network classifier of representative is used than wide, but it is easy to converge on local pole in the training process
Small value, and accuracy of identification is not high when sample size is fewer.
Summary of the invention
The present invention is to overcome defect at least one of in the prior art, provides a kind of Power Quality Disturbance identification side
Method.The advantage of method of the invention based on fuzzy clustering, can be improved accuracy of identification.
In order to solve the above technical problems, the technical solution adopted by the present invention is that: a kind of Power Quality Disturbance identification side
Method, wherein include the following steps:
S1. known Power Quality Disturbance is obtained;
S2. wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extracts feature vector;
S3. the feature vector of the Power Quality Disturbance based on step S2, based on fuzzy clustering to electrical energy power quality disturbance
Signal is classified, and obtains each cluster centre vector;
S4. Power Quality Disturbance to be identified is extracted into feature vector by step S2, and calculated separately and step S3
The distance between obtained each cluster centre vector, and then judge the type of Power Quality Disturbance to be identified.
Further, in the step S1, the known Power Quality Disturbance includes that voltage swells signal, voltage are rapid
Signal, voltage interrupt signal, transient state pulse signal, transient oscillation signal, harmonic signal and voltage flicker signal drop.
Further, in the step S2, wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extract
Feature vector specifically comprises the following steps:
S21. 3 layers of wavelet package transforms are carried out to Power Quality Disturbance, obtains the wavelet packet coefficient on each sub-band, remembered
For Wd(k), d=1,2,3 ..., D, D=2^3 are sub-band number, and k=1,2,3 ..., K, K are wavelet packet coefficient length;
S22. for the wavelet packet coefficient W on sub-bandd(k), its ENERGY E is calculated as followsd:
S23. for the wavelet packet coefficient W on sub-bandd(k), its energy percentage Δ E is calculated as followsd:
S24. for Power Quality Disturbance, feature vector λ are as follows:
λ=(Δ E1,ΔE2,....,ΔED)
Further, in the step S3, the feature vector of the Power Quality Disturbance based on step S2, based on fuzzy
Cluster classifies to Power Quality Disturbance, and obtains each cluster centre vector, specifically comprises the following steps:
S31. objective function J:
In formula, N is Power Quality Disturbance number, and C is Power Quality Disturbance type number, i.e. C=7, uijFor
The element of subordinated-degree matrix U, x are Weighted Index, λjFor the feature vector of j-th of Power Quality Disturbance, viIt is poly- for i-th
Class center vector, (λj-vi)·(λj-vi) it is vector (λj-vi) and vector (λj-vi) inner product;
S32. Weighted Index x and cluster centre vector v are initializedi;
S33. in the t times iteration, subordinated-degree matrix U and cluster centre vector V is updated according to the following formula;
In formula, uij (t)And vi (t)U after respectively the t times iterationijAnd viValue, vi (t-1)For v after the t-1 times iterationi's
Value, dijFor vectorAnd vectorBetween inner product;Equally, drjFor vectorAnd vectorBetween inner product;
S34. u is utilizedij (t)And vi (t)Calculate J(t)If met | J(t)-J(t-1)|≤ε, ε are required precision, then stop changing
In generation, otherwise return step S33 continues iteration;J(t)And J(t-1)Objective function after respectively the t times iteration and the t-1 times iteration
Value;
S35. to subordinated-degree matrix U=[uij]C×N, λ is judged using maximum membership grade principlejClassification, maximum membership degree is former
It then indicates are as follows: Ai(λj)=max { u1j,u2j,....,uCj}
In formula, Ai(λj) indicate λjIt is under the jurisdiction of ith cluster center;
S36. the cluster centre vector after iteration is denoted as
Further, in the step S4, it is as follows to judge that the type of Power Quality Disturbance to be identified specifically includes
Step:
S41. Power Quality Disturbance to be identified is extracted into feature vector by step S2, be denoted as
S42. feature vector is calculated separatelyWith cluster centre vectorDistance Si:
S43. distance S is found outiMinimum value is correspondingThen Power Quality Disturbance to be identified belongs to y-th of cluster
Center.
Compared with prior art, beneficial effects of the present invention:
A kind of Power Quality Disturbance recognition methods provided by the invention, the advantage based on fuzzy clustering can be improved
The accuracy of identification of Power Quality Disturbance.
Detailed description of the invention
Fig. 1 is overall flow schematic diagram of the invention.
Specific embodiment
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent;In order to better illustrate this embodiment, attached
Scheme certain components to have omission, zoom in or out, does not represent the size of actual product;To those skilled in the art,
The omitting of some known structures and their instructions in the attached drawings are understandable.Being given for example only property of positional relationship is described in attached drawing
Illustrate, should not be understood as the limitation to this patent.
As shown in Figure 1, a kind of Power Quality Disturbance recognition methods, wherein include the following steps:
S1. known Power Quality Disturbance is obtained;Specifically, known Power Quality Disturbance includes voltage swells letter
Number, voltage dip signal, voltage interrupt signal, transient state pulse signal, transient oscillation signal, harmonic signal and voltage flicker letter
Number.
S2. wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extracts feature vector;Specifically include as
Lower step:
S21. 3 layers of wavelet package transforms are carried out to Power Quality Disturbance, obtains the wavelet packet coefficient on each sub-band, remembered
For Wd(k), d=1,2,3 ..., D, D=2^3 are sub-band number, and k=1,2,3 ..., K, K are wavelet packet coefficient length;
S22. for the wavelet packet coefficient W on sub-bandd(k), its ENERGY E is calculated as followsd:
S23. for the wavelet packet coefficient W on sub-bandd(k), its energy percentage Δ E is calculated as followsd:
S24. for Power Quality Disturbance, feature vector λ are as follows:
λ=(Δ E1,ΔE2,....,ΔED)
S3. the feature vector of the Power Quality Disturbance based on step S2, based on fuzzy clustering to electrical energy power quality disturbance
Signal is classified, and obtains each cluster centre vector;Specifically comprise the following steps:
S31. objective function J:
In formula, N is Power Quality Disturbance number, and C is Power Quality Disturbance type number, i.e. C=7, uijFor
The element of subordinated-degree matrix U, x are Weighted Index, λjFor the feature vector of j-th of Power Quality Disturbance, viIt is poly- for i-th
Class center vector, (λj-vi)·(λj-vi) it is vector (λj-vi) and vector (λj-vi) inner product;
S32. Weighted Index x and cluster centre vector v are initializedi;
S33. in the t times iteration, subordinated-degree matrix U and cluster centre vector V is updated according to the following formula;
In formula, uij (t)And vi (t)U after respectively the t times iterationijAnd viValue, vi (t-1)For v after the t-1 times iterationi's
Value, dijFor vectorAnd vectorBetween inner product;Equally, drjFor vectorAnd vectorBetween inner product;
S34. ui is utilizedj (t)And vi (t)Calculate J(t)If met | J(t)-J(t-1)|≤ε, ε are required precision, then stop changing
In generation, otherwise return step S33 continues iteration;J(t)And J(t-1)Objective function after respectively the t times iteration and the t-1 times iteration
Value;
S35. to subordinated-degree matrix U=[uij]C×N, λ is judged using maximum membership grade principlejClassification, maximum membership degree is former
It then indicates are as follows: Ai(λj)=max { u1j,u2j,....,uCj}
In formula, Ai(λj) indicate λjIt is under the jurisdiction of ith cluster center;
S36. the cluster centre vector after iteration is denoted as
S4. Power Quality Disturbance to be identified is extracted into feature vector by step S2, and calculated separately and step S3
The distance between obtained each cluster centre vector, and then judge the type of Power Quality Disturbance to be identified.Specific packet
Include following steps:
S41. Power Quality Disturbance to be identified is extracted into feature vector by step S2, be denoted as
S42. feature vector is calculated separatelyWith cluster centre vectorDistance Si:
S43. distance S is found outiMinimum value is correspondingThen Power Quality Disturbance to be identified belongs to y-th of cluster
Center.
Obviously, the above embodiment of the present invention is just for the sake of clearly demonstrating examples made by the present invention, and is not
Restriction to embodiments of the present invention.For those of ordinary skill in the art, on the basis of the above description also
It can make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all
Made any modifications, equivalent replacements, and improvements etc. within the spirit and principles in the present invention should be included in right of the present invention and want
Within the protection scope asked.
Claims (5)
1. a kind of Power Quality Disturbance recognition methods, which comprises the steps of:
S1. known Power Quality Disturbance is obtained;
S2. wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extracts feature vector;
S3. the feature vector of the Power Quality Disturbance based on step S2, based on fuzzy clustering to Power Quality Disturbance
Classify, and obtains each cluster centre vector;
S4. Power Quality Disturbance to be identified is extracted into feature vector by step S2, and calculates separately and is obtained with step S3
The distance between each cluster centre vector, and then judge the type of Power Quality Disturbance to be identified.
2. a kind of Power Quality Disturbance recognition methods according to claim 1, which is characterized in that the step S1
In, the known Power Quality Disturbance includes voltage swells signal, voltage dip signal, voltage interrupt signal, transient state arteries and veins
Rush signal, transient oscillation signal, harmonic signal and voltage flicker signal.
3. a kind of Power Quality Disturbance recognition methods according to claim 1, which is characterized in that the step S2
In, wavelet package transforms are carried out to the Power Quality Disturbance of step S1, and extract feature vector, specifically comprised the following steps:
S21. 3 layers of wavelet package transforms are carried out to Power Quality Disturbance, obtains the wavelet packet coefficient on each sub-band, is denoted as Wd
(k), d=1,2,3 ..., D, D=2^3 are sub-band number, and k=1,2,3 ..., K, K are wavelet packet coefficient length;
S22. for the wavelet packet coefficient W on sub-bandd(k), its ENERGY E is calculated as followsd:
S23. for the wavelet packet coefficient W on sub-bandd(k), its energy percentage Δ E is calculated as followsd:
S24. for Power Quality Disturbance, feature vector λ are as follows:
λ=(Δ E1,ΔE2,....,ΔED)
4. a kind of Power Quality Disturbance recognition methods according to claim 1, which is characterized in that the step S3
In, the feature vector of the Power Quality Disturbance based on step S2 carries out Power Quality Disturbance based on fuzzy clustering
Classification, and each cluster centre vector is obtained, specifically comprise the following steps:
S31. objective function J:
In formula, N is Power Quality Disturbance number, and C is Power Quality Disturbance type number, i.e. C=7, uijTo be subordinate to
The element of matrix U is spent, x is Weighted Index, λjFor the feature vector of j-th of Power Quality Disturbance, viFor in ith cluster
Heart vector, (λj-vi)·(λj-vi) it is vector (λj-vi) and vector (λj-vi) inner product;
S32. Weighted Index x and cluster centre vector v are initializedi;
S33. in the t times iteration, subordinated-degree matrix U and cluster centre vector V is updated according to the following formula;
In formula, uij (t)And vi (t)U after respectively the t times iterationijAnd viValue, vi (t-1)For v after the t-1 times iterationiValue, dij
For vectorAnd vectorBetween inner product;Equally, drjFor vectorAnd vectorIt
Between inner product;
S34. ui is utilizedj (t)And vi (t)Calculate J(t)If met | J(t)-J(t-1)|≤ε, ε are required precision, then stop iteration, no
Then return step S33 continues iteration;J(t)And J(t-1)Target function value after respectively the t times iteration and the t-1 times iteration;
S35. to subordinated-degree matrix U=[uij]C×N, λ is judged using maximum membership grade principlejClassification, maximum membership grade principle table
It is shown as: Ai(λj)=max { u1j,u2j,....,uCj}
In formula, Ai(λj) indicate λjIt is under the jurisdiction of ith cluster center;
S36. the cluster centre vector after iteration is denoted as
5. a kind of Power Quality Disturbance recognition methods according to claim 1, which is characterized in that the step S4
In, judge that the type of Power Quality Disturbance to be identified specifically comprises the following steps:
S41. Power Quality Disturbance to be identified is extracted into feature vector by step S2, be denoted as
S42. feature vector is calculated separatelyWith cluster centre vectorDistance Si:
S43. distance S is found outiMinimum value is correspondingThen Power Quality Disturbance to be identified belongs to y-th of cluster centre.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811447626.1A CN109324250A (en) | 2018-11-29 | 2018-11-29 | A kind of Power Quality Disturbance recognition methods |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811447626.1A CN109324250A (en) | 2018-11-29 | 2018-11-29 | A kind of Power Quality Disturbance recognition methods |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109324250A true CN109324250A (en) | 2019-02-12 |
Family
ID=65257651
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811447626.1A Pending CN109324250A (en) | 2018-11-29 | 2018-11-29 | A kind of Power Quality Disturbance recognition methods |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109324250A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110338821A (en) * | 2019-08-12 | 2019-10-18 | 深圳市发掘科技有限公司 | A kind of driver's driving condition recognition methods |
CN111474410A (en) * | 2020-04-07 | 2020-07-31 | 宁波市重强电器有限公司 | Detection method of electric connector |
CN111639583A (en) * | 2020-05-26 | 2020-09-08 | 国网湖南省电力有限公司 | Method and system for identifying power quality disturbance of power grid |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016007217A1 (en) * | 2014-07-11 | 2016-01-14 | Abb Inc. | Decision support system for outage management and automated crew dispatch |
CN106156485A (en) * | 2016-06-16 | 2016-11-23 | 广州供电局有限公司 | Method for diagnosing fault of power transformer and device |
CN106778923A (en) * | 2017-03-23 | 2017-05-31 | 广东电网有限责任公司珠海供电局 | A kind of Power Quality Disturbance sorting technique and device |
CN107064759A (en) * | 2017-04-25 | 2017-08-18 | 国网安徽省电力公司电力科学研究院 | A kind of extra-high voltage equipment insulation defect type judgement method and system |
CN107153155A (en) * | 2017-06-26 | 2017-09-12 | 广东电网有限责任公司珠海供电局 | A kind of cable local discharge signal characteristic vector extracting method |
CN108132428A (en) * | 2017-11-24 | 2018-06-08 | 国网北京市电力公司 | The method, apparatus of shelf depreciation multi-source separation |
CN108776276A (en) * | 2018-06-29 | 2018-11-09 | 合肥工业大学 | Multiplexing electric abnormality detection method and system |
-
2018
- 2018-11-29 CN CN201811447626.1A patent/CN109324250A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016007217A1 (en) * | 2014-07-11 | 2016-01-14 | Abb Inc. | Decision support system for outage management and automated crew dispatch |
CN106156485A (en) * | 2016-06-16 | 2016-11-23 | 广州供电局有限公司 | Method for diagnosing fault of power transformer and device |
CN106778923A (en) * | 2017-03-23 | 2017-05-31 | 广东电网有限责任公司珠海供电局 | A kind of Power Quality Disturbance sorting technique and device |
CN107064759A (en) * | 2017-04-25 | 2017-08-18 | 国网安徽省电力公司电力科学研究院 | A kind of extra-high voltage equipment insulation defect type judgement method and system |
CN107153155A (en) * | 2017-06-26 | 2017-09-12 | 广东电网有限责任公司珠海供电局 | A kind of cable local discharge signal characteristic vector extracting method |
CN108132428A (en) * | 2017-11-24 | 2018-06-08 | 国网北京市电力公司 | The method, apparatus of shelf depreciation multi-source separation |
CN108776276A (en) * | 2018-06-29 | 2018-11-09 | 合肥工业大学 | Multiplexing electric abnormality detection method and system |
Non-Patent Citations (4)
Title |
---|
常炳双 等: "小波包和最小二乘支持向量机的电能质量扰动识别", 《现代电子技术》 * |
王忠玉 等: "《模糊数据统计学》", 31 March 2008, 哈尔滨工业大学 * |
胡昌华 等: "《基于MATLAB 6.x 的系统分析与设计——小波分析》", 31 January 2004, 西安电子科技大学出版社 * |
韩玉环 等: "基于FCM 的暂态电能质量扰动识别", 《电力系统保护与控制》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110338821A (en) * | 2019-08-12 | 2019-10-18 | 深圳市发掘科技有限公司 | A kind of driver's driving condition recognition methods |
CN111474410A (en) * | 2020-04-07 | 2020-07-31 | 宁波市重强电器有限公司 | Detection method of electric connector |
CN111639583A (en) * | 2020-05-26 | 2020-09-08 | 国网湖南省电力有限公司 | Method and system for identifying power quality disturbance of power grid |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109871976A (en) | A kind of prediction technique of power quality containing distributed power distribution network based on cluster and neural network | |
CN109324250A (en) | A kind of Power Quality Disturbance recognition methods | |
CN105550709B (en) | A kind of remote sensing image power transmission line corridor wood land extracting method | |
Li et al. | A new flood forecasting model based on SVM and boosting learning algorithms | |
CN109063911A (en) | A kind of Load aggregation body regrouping prediction method based on gating cycle unit networks | |
CN108832619A (en) | Transient stability evaluation in power system method based on convolutional neural networks | |
CN106327357B (en) | A kind of load identification method based on improvement probabilistic neural network | |
CN103323749A (en) | Multi-classifier information fusion partial discharge diagnostic method | |
CN103646354A (en) | Effective index FCM and RBF neural network-based substation load characteristic categorization method | |
Wang et al. | The load characteristics classification and synthesis of substations in large area power grid | |
CN108875912A (en) | A kind of neural network model for image recognition | |
CN109330613A (en) | Human body Emotion identification method based on real-time brain electricity | |
CN110389663A (en) | A kind of sEMG gesture identification method based on small wave width learning system | |
CN111652478B (en) | Umbrella algorithm-based power system voltage stability evaluation misclassification constraint method | |
Peng et al. | A very short term wind power prediction approach based on Multilayer Restricted Boltzmann Machine | |
Wang et al. | Wind process pattern forecasting based ultra-short-term wind speed hybrid prediction | |
CN114997217A (en) | Transformer fault diagnosis method based on RFID sensing tag and IGWO-ELM | |
CN110321555A (en) | A kind of power network signal classification method based on Recognition with Recurrent Neural Network model | |
Yang et al. | Event detection, localization, and classification based on semi-supervised learning in power grids | |
CN109558467A (en) | The recognition methods of electricity consumption class of subscriber and system | |
Chang et al. | Forecasting wind power generation by a new type of radial basis function-based neural network | |
CN108038518A (en) | A kind of photovoltaic generation power based on meteorological data determines method and system | |
Lu et al. | A deep belief network based model for urban haze prediction | |
CN109492651A (en) | A kind of intelligent identification Method of device signal lamp | |
CN109034453A (en) | A kind of Short-Term Load Forecasting Method based on multiple labeling neural network |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190212 |
|
RJ01 | Rejection of invention patent application after publication |