CN104268576A - Electric system transient stability classification method based on TNN-SVM - Google Patents
Electric system transient stability classification method based on TNN-SVM Download PDFInfo
- Publication number
- CN104268576A CN104268576A CN201410531234.9A CN201410531234A CN104268576A CN 104268576 A CN104268576 A CN 104268576A CN 201410531234 A CN201410531234 A CN 201410531234A CN 104268576 A CN104268576 A CN 104268576A
- Authority
- CN
- China
- Prior art keywords
- point
- sample
- nearest neighbor
- tnn
- classification
- 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
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an electric system transient stability classification method based on a TNN-SVM. The two factors of the class and the distance are comprehensively considered in the electric system transient stability classification process so that a sample set can be pruned, based on a traditional NN or KNN, a class affiliation concept is introduced, classification judgment of the nearest neighbor is conducted by the adoption of a quantized thought, and a nearest neighbor algorithm is improved based on the combination of the nearest neighbor algorithm and a support vector machine algorithm. Whether sample points are kept or not is determined in the aspect of the numbers of same nearest neighbor sample points and different nearest neighbor sample points and in combination with the distance between the points and the sample points, the limitations of the NN and the KNN are overcome, more effective sample set pruning can be provided, and the accuracy and the efficiency of transient stability classification are improved.
Description
Technical field
The present invention relates to a kind of electric power system transient stability sorting technique based on TNN-SVM.
Background technology
Along with the expanding day of electrical network scale, the structural complexity of electrical network also grows with each passing day, and dynamic perfromance is more complicated and changeable, Transient Instability occurs and to affect the possibility of grid stability also increasing.Transient stability is the ability of electric system generator synchronous operation in maintenance system when being subject to large interference (such as: power transmission line short circuit, cut machine, removal of load etc.).When there is large disturbances, comprise generator amature angle, Line Flow, node voltage system variable all will there is on a large scale change, if disturbed system can maintain all generators in synchronous operation state, and be transitioned into a new stable equilibrium point progressively, otherwise system chambers transient state instability.Transient stability evaluation in power system (Transient Stability Assessment, TSA) is a very important problem in Operation of Electric Systems.
The fast and stable assessment being combined into large-scale power system of artificial intelligence and mode identification technology provides a kind of new method for solving.Owing to not needing the mathematical model setting up system, directly can seek the mapping relations between state parameter and stability or stability index from sample, solving stability index does not need to repeat to sound out, and it is less that calculated amount receives system scale impact, and total evaluation speed is very fast.But lack the implementation method of standard, precision is to sample/rule interestingness, and the factor such as to choose of input variable is very responsive, and when relating to improper, the reliability and stability of assessment are poor.The mode identification method that appears as of SVM brings new vitality, and it makes it have tempting application prospect in TSA field with its intrinsic high execution speed.
It is suppose when linear separability that support vector machine (Support Vector Machine, SVM) theory carries out classifying, and finds optimizing decision face.Support vector machine is a kind of new small sample Statistical Learning Theory, can be applied to pattern recognition problem.The output of support vector machine is a serial number, reflects the distance of sample point to Optimal Separating Hyperplane.Adopt support vector machine structural classification device to carry out transient stability evaluation in power system, and discuss the impact that different parameters exports support vector machine, achieve better effects, but still have misclassification to exist.It has following advantage: it can obtain good classification capacity when Finite Samples; Not too responsive to the probability distribution of the dimension of the input space, training set size, sample.
The classificating thought of support vector machine is illustrated in fig. 1 shown below.For the bidimensional situation of such as Fig. 1, solid dot and hollow dots represent positive and negative two class samples respectively, straight line H is sorting track, straight line H1, H2 are parallel with straight line H, and cross the sample point that positive and negative two class middle distance sorting track H are nearest respectively, so, the distance of H1 to H2 is exactly class interval (margin).The optimal classification line that we will find is exactly not only two classes correctly can be separated, and will make class interval to reach maximum, this makes it possible to make the universality of classifying reach maximum.
On the basis of support vector machine, in order to training speed can be increased substantially, improve the generalization ability of support vector machine, in conjunction with nearest neighbor method (Nearest Neighbor, NN) NN-SVM sorter is proposed, it utilizes arest neighbors thought to select Margin Vector, it is first pruned training set, determines that it is accepted or rejected according to each sample and its arest neighbors class target similarities and differences, and then obtains sorter with SVM training.Nearest neighbor method (Nearest Neighbor, NN) be all regard all learning samples as independent representative point, by calculating the distance investigated between sample and other all samples, find and investigate the shortest namely nearest sample of sample distance, the classification of this sample is just by the classification as investigation sample.This method belongs to the category of pattern-recognition, and does not need parameter, not by the restriction of parameter.
Although SVM compares with other mode identification method have good recognition capability, but also there are some problems in actual applications, as not high to challenge nicety of grading, to select permeability of different application sum functions parameter etc. and NN-SVM in linearly inseparable situation, more support vector can be lost, cause the classification capacity of SVM to decline.For these situations, propose a kind of NN-SVM sorting algorithm of improvement, i.e. KNN-SVM.KNN is the popularization of NN, first selects K the nearest samples point investigating sample point exactly, then see which kind of the most sample points in this K neighbour's sample point belong to, and which kind of this sample point just belongs to when carrying out classification and judging.
Mainly there are following three problems in the research of current SVM, NN-SVM, KNN-SVM sorter:
1. divide the position situation of sample from mistake during classification, the situation of SVM classifier misclassification is all appear near interphase in certain area, so we should strengthen sample point analysis near interphase to improve the indexs such as classification accuracy.
2. if by SVM and NN combine, SVM is regarded as the NN sorter that every class only has a representative point, a point is only got for every class support vector, just there will be the situation of by mistake deleting.
3. still, KNN-SVM deletes for sample set the problem still existing and delete by mistake, because KNN-SVM finds its K arest neighbors for the investigation point in sample set, if similar, retains, if foreign peoples, deletes, carry out classification judgement with this.If this rule runs into an investigation point there is a large amount of foreign peoples's points thick and fast around, adopt KNN-SVM just a large amount of foreign peoples may be put like this and remove, cause and delete by mistake, the mistake having carried out being unprofitable to classification effectiveness is deleted.This rule also has unaccommodated situation to exist, if in fact majority is foreign peoples's point in the K near sample point point, but these foreign peoples point is all positioned at position far away, the similar point that number is few is positioned at the periphery of sample point on the contrary thick and fast, although consider K arest neighbors like this, still cause and delete by mistake.
So should consider classification and distance two because usually carrying out sample set pruning, the concept introducing herein class degree of membership is come comprehensive number and distance factor and is differentiated and investigate some classification ownership.
Summary of the invention
Devise a kind of TNN-SVM sorter herein, for transient stability evaluation in power system, effectively can avoid misclassification, improve the reliability of Stability Assessment.
In the process of electric power system transient stability classification, consider classification and distance two because usually carrying out sample set pruning, the concept introducing class degree of membership is come comprehensive number and distance factor and is differentiated and investigate some classification ownership.On the basis of traditional NN or KNN, introduce the concept of class degree of membership, the classification adopting the thought quantized to carry out arest neighbors judges.By arest neighbors (NN) algorithm on the basis that support vector machine (SVM) algorithm is combined, improve arest neighbors (NN, KNN) algorithm: not only from arest neighbors with foreign peoples's sample point quantitative aspects, and combine the factor of distance between these point and sample points, determine retaining of this sample point.Overcome the limitation of NN and KNN, more effective sample set can be provided to prune, improve precision and the efficiency of transient stability classification.
Based on an electric power system transient stability sorting technique of TNN-SVM, comprise the steps:
Step one: pre-service is carried out to data;
Step 2: find support vector machines decision surface;
Step 3: adopt TNN algorithm to prune sample point in threshold range certain near support vector machines decision surface;
Step 4: judge whether wrong report, if wrong report, returns step one; Otherwise, system responses.
Pruning should be carried out based on the electric power system transient stability sorting technique described employing TNN algorithm also comprised in step 3 of TNN-SVM to sample point in threshold range certain near support vector machines decision surface to comprise:
Step 1: a given training set
,
...,
,
,
,
.Wherein
represent sample point,
represent the class mark of this point:
Step 2: for sample
, find its T nearest neighbor point;
Step 3: investigate each nearest neighbor point from 1 to T
if, classification with
identical, then
=1, otherwise
=-1;
Step 4: calculate each nearest neighbor point from 1 to T
with sample
distance
;
Step 5: then
it is exactly nearest neighbor point
to sample point
classification ownership factor of influence;
Step 6: find
m the nearest neighbor point of=1, then
the nearest neighbor point of=-1 is exactly T-m;
Step 7: classification is on average sued for peace, and obtains the class degree of membership of this sample point;
Step 8: by the class degree of membership of this sample point
with given threshold value
compare, if
<
then show that the classification ownership degree of this sample point to around T nearest neighbor point is low, this sample point is deleted;
Step 9: return step 1 is right
carry out the judgement of class degree of membership, to determine the choice of this point;
Step 10: after one by one n sample point of sample set having been pruned from 1 to n, then input support vector machines and classify.
Accompanying drawing explanation
Fig. 1 shows SVM interphase;
Fig. 2 shows the electric power system transient stability sorting technique that the present invention is based on TNN-SVM.
Embodiment
Fig. 1 is the circuit theory diagrams of one embodiment of the invention based on the electric power system transient stability sorting technique of TNN-SVM, and it comprises the steps:
Step one: pre-service is carried out to data;
Step 2: find support vector machines decision surface;
Step 3: adopt TNN algorithm to prune sample point in threshold range certain near support vector machines decision surface;
Step 4: judge whether wrong report, if wrong report, returns step one; Otherwise, system responses.
The concept introducing class degree of membership below weighs the influence degree of same foreign peoples point to sample point classification.
Using the class mark (1 or-1) of each nearest neighbor point of sample point as molecule, with the distance of sample point as denominator, in this, as the classification ownership factor of influence of each nearest neighbor point to this sample point, afterwards average for all factor of influence classification rear summation is judged the classification ownership of this sample point, decide the choice of sample point with this.
In one embodiment, the class comprising N number of point is supposed
, wherein
i-th sample point of class K, and each sample point
it is all m dimension.
Class degree of membership: for each sample
ask T nearest with it sample, if investigate point apart from this T sample point be
,
,
,
, Euclidean distance is as follows, namely
And represent the factor of influence of i-th point to the sample class ownership investigated with 1/Di.Define in the following several ways: if this T sample is all similar with investigation sample, then class degree of membership is:
If this T sample is not similar with investigation sample, then class degree of membership is:
If this T sample has m to be all similarly (might as well suppose that distance is with investigating sample
,
...,
), and remaining T-m is not that class (might as well suppose that distance is with investigation sample
,
...,
), the classification ownership degree of sample is described with class degree of membership, and class degree of membership is:
According to the theoretical foundation of improvement and the definition of class degree of membership concept, adopt TNN algorithm to carry out pruning to sample point in threshold range certain near support vector machines decision surface and comprise the steps:
Step 1: a given training set
,
...,
,
,
,
.Wherein
represent sample point,
represent the class mark of this point:
Step 2: for sample
, find its T nearest neighbor point;
Step 3: investigate each nearest neighbor point from 1 to T
if, classification with
identical, then
=1, otherwise
=-1;
Step 4: calculate each nearest neighbor point from 1 to T
with sample
distance
;
Step 5: then
it is exactly nearest neighbor point
to sample point
classification ownership factor of influence;
Step 6: find
m the nearest neighbor point of=1, then
the nearest neighbor point of=-1 is exactly T-m;
Step 7: classification is on average sued for peace, and obtains the class degree of membership of this sample point:
Step 8: by the class degree of membership of this sample point
with given threshold value
compare, if
<
then show that the classification ownership degree of this sample point to around T nearest neighbor point is low, this sample point is deleted;
Step 9: return step 1 is right
carry out the judgement of class degree of membership, to determine the choice of this point;
Step 10: after one by one n sample point of sample set having been pruned from 1 to n, then input support vector machines and classify.
Test macro adopts New England 10 machine 39 node system, adopts the DianKeYuan exploitation PSASP generally used to carry out Load flow calculation and stability analysis.Boundary condition: synchro generator model does not consider the effect of speed regulator, excitation system and pressure regulator, adopts classical model; Load model is pressed constant impedance and is calculated.Failure mode: three-phase shortcircuit; Fault clearing time: 0.2 s; Topological structure: circuit automatic reclosing after failure removal, namely before and after fault, power network topology does not change; Abort situation: 20.Contingency set: before building training sample set, fault scanning is carried out to system every bar bus (as 120%) under heavier load level and draws.Load level is arranged: 80%, 90%, 100%, 110% and 120%, under each load level, arrange 5 kinds of generator outputs at random by certain constraint.After single transient stability completes, then revise load level or amendment generator output.
Obtain 400 effective samples by emulation, originally, all the other 100 form test sets to random selecting 300 composition training set.According to simulation result and expertise, altogether have selected the input space X collection that 8 proper vectors are formed:
1) the maximum initial acceleration of generator;
2) during failure removal, the maximum rotor kinetic energy of generator;
3) there is the initial rotor angle of peak acceleration generator;
4), during failure removal, there is the corner of maximum kinetic energy generator;
5), during failure removal, there is the kinetic energy of hard-over generator;
6) " the gross energy adjustment " of system;
7) the minimum initial acceleration of generator;
8) mean square deviation of the initial acceleration of all generators.
According to simulation result, all samples are divided into stable and unstable two classes, represent Stabilized with 1, represent unstable class with 0, composition output region Y collection.
Compared with traditional BP neural network, the Transient Stability Evaluation sorter based on support vector machine in this paper, not only decreases the training time, and can obtain a classification boundaries district, improves the reliability of Transient Stability Evaluation.The validity of method in this paper to Transient Stability Evaluation is shown to the emulation that New England 10 machine 39 node system is done.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.
Claims (4)
1., based on an electric power system transient stability sorting technique of TNN-SVM, comprise the steps:
Step one: pre-service is carried out to data;
Step 2: find support vector machines decision surface;
Step 3: adopt TNN algorithm to prune sample point in threshold range certain near support vector machines decision surface;
Step 4: judge whether wrong report, if wrong report, returns step one; Otherwise, system responses.
2. the electric power system transient stability sorting technique based on TNN-SVM according to claim 1, is characterized in that, the described employing TNN algorithm in step 3 carries out pruning to sample point in threshold range certain near support vector machines decision surface and comprises:
Step 1: a given training set
,
...,
,
,
,
, wherein
represent sample point,
represent the class mark of this point:
Step 2: for sample
, find its T nearest neighbor point;
Step 3: investigate each nearest neighbor point from 1 to T
if, classification with
identical, then
=1, otherwise
=-1;
Step 4: calculate each nearest neighbor point from 1 to T
with sample
distance
;
Step 5: then
it is exactly nearest neighbor point
to sample point
classification ownership factor of influence;
Step 6: find
m the nearest neighbor point of=1, then
the nearest neighbor point of=-1 is exactly T-m;
Step 7: classification is on average sued for peace, and obtains the class degree of membership of this sample point;
Step 8: by the class degree of membership of this sample point
with given threshold value
compare, if
<
then show that the classification ownership degree of this sample point to around T nearest neighbor point is low, this sample point is deleted;
Step 9: return step 1 is right
carry out the judgement of class degree of membership, to determine the choice of this point;
Step 10: after one by one n sample point of sample set having been pruned from 1 to n, then input support vector machines and classify.
3. the electric power system transient stability sorting technique based on TNN-SVM according to claim 2, is characterized in that, step: calculate each nearest neighbor point from 1 to T in 4
with sample
distance
computing formula be:
。
4. the electric power system transient stability sorting technique based on TNN-SVM according to claim 2, it is characterized in that, the classification in step 7 is on average sued for peace, and obtains the class degree of membership of this sample point according to following formula:
。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410531234.9A CN104268576A (en) | 2014-10-11 | 2014-10-11 | Electric system transient stability classification method based on TNN-SVM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410531234.9A CN104268576A (en) | 2014-10-11 | 2014-10-11 | Electric system transient stability classification method based on TNN-SVM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104268576A true CN104268576A (en) | 2015-01-07 |
Family
ID=52160096
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410531234.9A Pending CN104268576A (en) | 2014-10-11 | 2014-10-11 | Electric system transient stability classification method based on TNN-SVM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104268576A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104881741A (en) * | 2015-05-25 | 2015-09-02 | 清华大学 | Power system transient stability determination method based on support vector machine |
| CN105512799A (en) * | 2015-11-26 | 2016-04-20 | 中国电力科学研究院 | Mass online historical data-based power system transient stability evaluation method |
| CN106055883A (en) * | 2016-05-25 | 2016-10-26 | 中国电力科学研究院 | Transient stability assessment input characteristic validity analysis method based on improved Sammon mapping |
| CN106503279A (en) * | 2015-09-06 | 2017-03-15 | 中国电力科学研究院 | A kind of modeling method for transient stability evaluation in power system |
| CN109459759A (en) * | 2018-11-13 | 2019-03-12 | 中国科学院合肥物质科学研究院 | City Terrain three-dimensional rebuilding method based on quadrotor drone laser radar system |
| CN109711450A (en) * | 2018-12-20 | 2019-05-03 | 北京科东电力控制***有限责任公司 | A kind of power grid forecast failure collection prediction technique, device, electronic equipment and storage medium |
| CN110533265A (en) * | 2019-09-20 | 2019-12-03 | 云南电网有限责任公司电力科学研究院 | A kind of power distribution network lightning stroke method for early warning and power distribution network are struck by lightning prior-warning device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100239170A1 (en) * | 2009-03-18 | 2010-09-23 | Asnis Gary I | System and method for target separation of closely spaced targets in automatic target recognition |
| US20100322525A1 (en) * | 2009-06-19 | 2010-12-23 | Microsoft Corporation | Image Labeling Using Multi-Scale Processing |
| CN102819693A (en) * | 2012-08-17 | 2012-12-12 | 中国人民解放军第三军医大学第二附属医院 | Prediction method for protein subcellular site formed based on improved-period pseudo amino acid |
-
2014
- 2014-10-11 CN CN201410531234.9A patent/CN104268576A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100239170A1 (en) * | 2009-03-18 | 2010-09-23 | Asnis Gary I | System and method for target separation of closely spaced targets in automatic target recognition |
| US20100322525A1 (en) * | 2009-06-19 | 2010-12-23 | Microsoft Corporation | Image Labeling Using Multi-Scale Processing |
| CN102819693A (en) * | 2012-08-17 | 2012-12-12 | 中国人民解放军第三军医大学第二附属医院 | Prediction method for protein subcellular site formed based on improved-period pseudo amino acid |
Non-Patent Citations (2)
| Title |
|---|
| 于秋玲: "基于改进NN-SVM算法的网络入侵检测", 《***工程理论与实践》 * |
| 盛琪: "基于支持向量机的电力***暂态稳定评估研究", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104881741A (en) * | 2015-05-25 | 2015-09-02 | 清华大学 | Power system transient stability determination method based on support vector machine |
| CN104881741B (en) * | 2015-05-25 | 2017-12-12 | 国网浙江省电力公司 | Electric power system transient stability determination methods based on SVMs |
| CN106503279A (en) * | 2015-09-06 | 2017-03-15 | 中国电力科学研究院 | A kind of modeling method for transient stability evaluation in power system |
| CN106503279B (en) * | 2015-09-06 | 2019-07-09 | 中国电力科学研究院 | A kind of modeling method for transient stability evaluation in power system |
| CN105512799A (en) * | 2015-11-26 | 2016-04-20 | 中国电力科学研究院 | Mass online historical data-based power system transient stability evaluation method |
| CN105512799B (en) * | 2015-11-26 | 2022-12-30 | 中国电力科学研究院 | Power system transient stability evaluation method based on mass online historical data |
| CN106055883A (en) * | 2016-05-25 | 2016-10-26 | 中国电力科学研究院 | Transient stability assessment input characteristic validity analysis method based on improved Sammon mapping |
| CN106055883B (en) * | 2016-05-25 | 2022-09-02 | 中国电力科学研究院 | Transient stability evaluation input feature validity analysis method based on improved Sammon mapping |
| CN109459759A (en) * | 2018-11-13 | 2019-03-12 | 中国科学院合肥物质科学研究院 | City Terrain three-dimensional rebuilding method based on quadrotor drone laser radar system |
| CN109459759B (en) * | 2018-11-13 | 2020-06-30 | 中国科学院合肥物质科学研究院 | Urban terrain three-dimensional reconstruction method based on quad-rotor unmanned aerial vehicle laser radar system |
| CN109711450A (en) * | 2018-12-20 | 2019-05-03 | 北京科东电力控制***有限责任公司 | A kind of power grid forecast failure collection prediction technique, device, electronic equipment and storage medium |
| CN110533265A (en) * | 2019-09-20 | 2019-12-03 | 云南电网有限责任公司电力科学研究院 | A kind of power distribution network lightning stroke method for early warning and power distribution network are struck by lightning prior-warning device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104268576A (en) | Electric system transient stability classification method based on TNN-SVM | |
| Guo et al. | Online identification of power system dynamic signature using PMU measurements and data mining | |
| CN102074955B (en) | Method based on knowledge discovery technology for stability assessment and control of electric system | |
| CN109670213B (en) | Multi-machine equivalent modeling method and device for wind power plant | |
| CN110428005B (en) | Umbrella algorithm-based dynamic security misclassification constraint method for power system | |
| CN111523785A (en) | Power system dynamic security assessment method based on generation countermeasure network | |
| Badr et al. | Fault identification of photovoltaic array based on machine learning classifiers | |
| CN111652478B (en) | Umbrella algorithm-based power system voltage stability evaluation misclassification constraint method | |
| CN110609477B (en) | Electric power system transient stability discrimination system and method based on deep learning | |
| CN110795690A (en) | Wind power plant operation abnormal data detection method | |
| CN102063626A (en) | Power quality disturbance mode discrimination method | |
| CN106503279A (en) | A kind of modeling method for transient stability evaluation in power system | |
| CN113988558B (en) | Power grid dynamic security assessment method based on blind area identification and electric coordinate system expansion | |
| CN111651932A (en) | Online dynamic security assessment method for power system based on integrated classification model | |
| CN111400966A (en) | Static voltage stability evaluation method of power system based on improved AdaBoost | |
| CN111401792A (en) | Dynamic safety assessment method based on extreme gradient lifting decision tree | |
| Nizam | Kohonen neural network clustering for voltage control in power systems | |
| Wang et al. | Complex fault source identification method for high-voltage trip-offs of wind farms based on SU-MRMR and PSO-SVM | |
| CN116896093A (en) | Online analysis and optimization method for grid-connected oscillation stability of wind farm | |
| Guo et al. | Identification of power system dynamic signature using hierarchical clustering | |
| Elgamal et al. | Seamless Machine Learning Models to Detect Faulty Solar Panels | |
| CN116204771A (en) | Power system transient stability key feature selection method, device and product | |
| CN115693785A (en) | Method for judging transient power angle stability of power system and related device thereof | |
| He et al. | A method for transient stability assessment based on pattern recognition | |
| CN108647275A (en) | The recognition methods of isolated island detecting state and device, storage medium, processor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150107 |