CN108460486A - A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network - Google Patents
A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network Download PDFInfo
- Publication number
- CN108460486A CN108460486A CN201810178916.4A CN201810178916A CN108460486A CN 108460486 A CN108460486 A CN 108460486A CN 201810178916 A CN201810178916 A CN 201810178916A CN 108460486 A CN108460486 A CN 108460486A
- Authority
- CN
- China
- Prior art keywords
- data
- voltage deviation
- meteorological
- moment
- measured
- 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
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The present invention relates to a kind of based on the voltage deviation prediction technique for improving clustering algorithm and neural network, including：1, the voltage deviation data and meteorological data of acquisition power grid historical time section different moments obtain the meteorological data at moment to be measured by weather forecast as historical data；2, dimension-reduction treatment is carried out to the meteorological data of historical time section and moment to be measured by PCA, chooses several principal components as comprehensive meteorological variables；3, by combining the K means clustering algorithms of AP clustering algorithms to cluster comprehensive meteorological variables data, multiple class clusters are obtained；4, the historical data with moment meteorological data same class cluster to be measured is chosen as training sample data collection, is trained by BP neural network and is obtained the voltage deviation prediction result at moment to be measured.Compared with prior art, the present invention considers influence of the meteorologic factor to voltage deviation, extracts training sample using clustering algorithm is improved, reduces information interference, improve precision of prediction.
Description
Technical field
The present invention relates to power quality fields, more particularly, to a kind of based on the voltage for improving clustering algorithm and neural network
Deflection forecast method.
Background technology
As novel electric power electric disguises the continuous access set, power quality problem getting worse.More and more substations
Upgrade electric energy quality monitoring system, obtains the real-time electric energy quality monitoring data of magnanimity.Electric energy quality monitoring data are deeply excavated,
It makes prediction to the variation tendency of power quality and early warning, the necessity for becoming safe and stable, the economic conveying of guarantee electric power is arranged
It applies.
Power quality problem is divided into stationary power quality problem and transient power quality problem.Stationary power quality problem packet
Include voltage deviation, frequency departure, voltage fluctuation and flicker, harmonic wave and tri-phase unbalance factor.Wherein, the harm of voltage deviation is the most
Obviously.Overtension causes equipment overvoltage, threatens insulation, reduces service life；Brownout prevents user equipment from normal
It uses.Therefore, predicted voltage change of error trend and take certain measure for electric system safe normal operation have very
High application value.It, cannot be five Index For Steady-states with one since different power quality index data have different characteristics
A prediction model predicts, no the problem of it will cause low precisions, can be to voltage deviation Index Establishment prediction model.
Common prediction technique having time serial method, grey method, SVM prediction method etc..But time sequence
Row method focuses on the fitting of data, and the uncertain factor to influencing power quality considers insufficient；Grey method to data from
Scattered degree requires, and when data discrete degree is larger, precision of prediction is poor；Support vector machines is stronger in processing stochastic volatility
Data when, precision is poor, and when data set scale is excessive, and the training time is long, and speed is slow.
Invention content
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind based on improvement cluster
The voltage deviation prediction technique of algorithm and neural network.
The purpose of the present invention can be achieved through the following technical solutions：
A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network, includes the following steps：
S1, the voltage deviation data for acquiring power grid historical time section different moments and meteorological data as historical data,
The meteorological data at moment to be measured is obtained by weather forecast, the meteorological data includes a variety of meteorologic factor variables；
S2, by Principal Component Analysis (PCA) to the meteorological data of the obtained historical time sections and moment to be measured of step S1
Dimension-reduction treatment is carried out, chooses several principal components as comprehensive meteorological variables；
S3, by combine AP (Affinity Propagation) clustering algorithm K-means clustering algorithms to step S2
Synthesis meteorological variables data clustered, obtain multiple class clusters；
S4, the historical data with the meteorological data same class cluster at moment to be measured is chosen as training sample data collection, pass through
BP neural network is trained and obtains the voltage deviation prediction result at moment to be measured.
Preferably, the step S3 is specifically included：
S31, the synthesis meteorological variables data of step S2 are clustered by AP clustering algorithms, obtains cluster numbers K and K
Cluster centre；
S32, K cluster centre for obtaining step S31 are as the initial cluster center of K-means clustering algorithms, to step
The synthesis meteorological variables data of rapid S2 carry out K-means clusters, obtain K class cluster.
Preferably, the meteorologic factor variable of the meteorological data includes：Temperature, dew point, humidity, air pressure, wind direction, wind speed and
Weather conditions.
Preferably, the accumulative variance contribution ratio for the principal component chosen in the step S2 is not less than 80%.
Preferably, the date of historical time section is identical as the season belonging to the date at moment to be measured in the step S1.
Preferably, the voltage deviation prediction result at moment to be measured is by multiple voltage deviation predicted values in the step S4
It takes and is worth to.
Compared with prior art, the present invention has the following advantages：
1, consider that voltage deviation is predicted in influence of the meteorologic factor to voltage deviation, it is single to make history data set no longer
One history voltage deviation data, improve the accuracy of prediction result.
2, Data Dimensionality Reduction processing is carried out using Principal Component Analysis, multiple meteorologic factor variables of meteorological data is pre-processed
One group of principal component for including enough information again independently of each other is obtained, effective meteorological data information is extracted, reduces calculating
Complexity, can also avoid information interference and information repeat.
3, training sample is extracted using improved clustering algorithm, AP clustering algorithms and K-means clustering algorithms is combined,
The mutual supplement with each other's advantages for realizing two class algorithms, compared with single clustering algorithm, the error sum of squares for improving clustering algorithm is minimum, makes acquisition
Class cluster between dispersibility it is high, the compactedness in class cluster is good, improves the accuracy of training sample selection.
4, it is had the following advantages that using AP clustering procedures：Cluster number need not be specified in advance；The result being performed a plurality of times
It is duplicate, initial value need not be randomly selected；The error sum of squares of clustering method more single than other is all low；Pass through the input phase
Carry out starting algorithm like degree matrix, therefore the symmetry for matrix of adjusting the distance does not require, the data scope of application is big.
Description of the drawings
Fig. 1 is the flow chart based on the voltage deviation prediction technique for improving clustering algorithm and neural network in embodiment one；
Fig. 2 is the flow chart of AP clustering algorithms in embodiment one；
Fig. 3 is the iterative process schematic diagram that AP is clustered in embodiment two；
Fig. 4 is on April 27th, 2014 and prediction result figure on the 28th in embodiment two；
Fig. 5 is the prediction error comparison diagram on April 27th, 2014 and 28 days in embodiment two；
Fig. 6 is the prediction result figure on May 1st, 2014 in embodiment two；
Fig. 7 is the prediction error comparison diagram on May 1st, 2014 in embodiment two；
Fig. 8 is the prediction result figure in April 28 and May 1 in 2014 in embodiment two；
Fig. 9 is the prediction error map of traditional BP neural network in embodiment two；
Figure 10 is the prediction error map of PCA+BP neural networks in embodiment two；
Figure 11 is the prediction error map of the method for the present invention in embodiment two.
Specific implementation mode
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, give detailed embodiment and specific operating process, but protection scope of the present invention is not limited to
Following embodiments.
Embodiment one
As shown in figure, a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network, including it is following
Step：
S1, the voltage deviation data for acquiring power grid historical time section different moments and meteorological data as historical data,
The meteorological data at moment to be measured is obtained by weather forecast, which includes a variety of meteorologic factor variables；
S2, the obtained historical time sections of step S1 and the meteorological data at moment to be measured are dropped by Principal Component Analysis
Dimension processing chooses several principal components as comprehensive meteorological variables；
S3, the synthesis meteorological variables data of step S2 are carried out by the K-means clustering algorithms of combination AP clustering algorithms
Cluster, obtains multiple class clusters；
S4, the historical data with the meteorological data same class cluster at moment to be measured is chosen as training sample data collection, pass through
BP neural network is trained and obtains the voltage deviation prediction result at moment to be measured.
Step S3 is specifically included：
S31, the synthesis meteorological variables data of step S2 are clustered by AP clustering algorithms, obtains cluster numbers K and K
Cluster centre μk(k=1,2 ..., K)；
S32, K cluster centre for obtaining step S31 are as the initial cluster center of K-means clustering algorithms, to step
The synthesis meteorological variables data of rapid S2 carry out K-means clusters, obtain K class cluster.
The date for the historical time section chosen in step S1 is identical as the season belonging to the date at moment to be measured, preferably
Historical time section chooses a period of time for including the date to be measured in former years.
In the present embodiment, the meteorologic factor variable of meteorological data includes：Temperature, dew point, humidity, air pressure, wind direction, wind speed and
Weather conditions, wherein weather conditions refer to the weather phenomena such as fine day, rainy day, cloudy.Table 1 show meteorologic factor and voltage deviation
Strength of correlation.
The strength of correlation of table 1 meteorologic factor and voltage deviation
| Related coefficient | Temperature | Dew point | Humidity | Air pressure | Wind direction | Wind speed | Weather conditions | Voltage deviation |
| Temperature | 1.0000 | 0.3536 | -0.4269 | -0.5393 | 0.1255 | 0.1704 | -0.1311 | -0.4626 |
| Dew point | 0.3536 | 1.0000 | 0.6761 | -0.6720 | 0.0789 | -0.0021 | 0.4354 | -0.1645 |
| Humidity | -0.4269 | 0.6761 | 1.0000 | -0.2521 | -0.0374 | -0.1477 | 0.5688 | 0.2126 |
| Air pressure | -0.5393 | -0.6720 | -0.2521 | 1.0000 | -0.0624 | -0.0475 | -0.2226 | 0.1173 |
| Wind direction | 0.1255 | 0.0789 | -0.0374 | -0.0624 | 1.0000 | 0.3200 | -0.1155 | -0.0238 |
| Wind speed | 0.1704 | -0.0021 | -0.1477 | -0.0475 | 0.3200 | 1.0000 | 0.0486 | -0.1943 |
| Weather conditions | -0.1311 | 0.4354 | 0.5688 | -0.2226 | -0.1155 | 0.0486 | 1.0000 | -0.0583 |
| Voltage deviation | -0.4626 | -0.1645 | 0.2126 | 0.1173 | -0.0238 | -0.1943 | -0.0583 | 1.0000 |
As shown in Table 1, gas epidemic disaster, wind speed, the correlation of dew point and voltage deviation are higher in meteorologic factor, but simultaneously
Correlation between inside meteorologic factor is also higher, if only extraction and the stronger meteorologic factor of voltage deviation correlation, will appear
Information polyisomenism can be obtained one group independently of each other and information content so meteorological data is carried out principal component analysis dimension-reduction treatment
Big principal component.Principal component analysis dimension-reduction algorithm can be described as, by original numerous variables with correlation, being reassembled into one
The new generalized variable being independent of each other is organized to replace primal variable, these generalized variables are principal component, and calculation formula is as follows：
V=eigvec [cov (X)]
Wherein：X is normalized data matrix；Cov (X) is covariance matrix；Eigvec indicates that the column vector for calculating V is
The Orthogonal Units feature vector of cov (X).By X be down to L dimension after data matrix be：
P=XVL
Wherein VLIt is arranged for the preceding L of matrix V.
When step S2 carries out principal component analysis, principal component number, the present embodiment are chosen according to characteristic value and variance contribution ratio
In, the accumulative variance contribution ratio of the principal component of selection is not less than 80%, can also adjust according to actual needs.
The flow chart of AP clustering algorithms is as shown in figure 3, AP clustering procedures can be described as the process of information transmission, with attraction information
Matrix R and attaching information matrix A exchange information between data point, and continuous iteration updates two information matrixs, until iteration knot
Beam, formula are as follows：
In formula, r (i, j) and a (i, j) is respectively the attraction information matrix element and attaching information matrix between i points and j points
Element；Similarities of the s (i, j) between i points and j points.R (i, j) and a (i, j) is stronger, then possibility of the j points as cluster centre
Property is bigger, and i points be under the jurisdiction of it is also bigger using j points as the possibility of the class cluster of cluster centre.
Since AP clustering algorithm iterative process easy tos produce concussion, so iteration all adds damped coefficient a λ, λ every time
∈ (0,1), has：
Wherein,WithIt indicates not considering the calculated r of damped coefficient when iterationi+1(i, j) and ai+1(i,
j)。
K-means cluster detailed process be：
(1) first from n data object X={ x1,x2,…xnIn, K object of random selection is as initial cluster center
μk(k=1,2 ..., K)；
(2) for unchecked other objects, then the Euclidean distance at the center selected according to them and step (1), respectively
They are allocated to the classification most like with it, forms K class cluster C={ ck, k=1,2 ... K }；
(3) all kinds of cluster data objects are calculated to class cluster cluster centre μ where itkTotal square distance and J (C) are until most
It is small, the mean value of all objects in class cluster is calculated as new cluster centre, wherein
In formula, xlFirst of data object is indicated, with matrix D=(dkl)n×KIt indicates to divide data set X, if wherein xl∈
cl,dkl=1；Ifdkl=0.
(4) judge whether cluster centre and J (C) value change, if variation goes to step (2), if constant, cluster knot
Beam.
BP neural network is a kind of multilayer feedforward neural network, which is mainly characterized by before signal to transmission, error
Backpropagation has very strong learning ability, can preferably adapt to variation and the data relationship of various complexity of data space, extensively
It is general to be applied to Classification and Identification, the fields such as approach, return, compressing.BP neural network is suitble to solve the problem of internal mechanism complexity, but
It is initial weight sensitivity, so the voltage deviation prediction result at moment to be measured is by taking multiple voltage deviations to predict in step S4
Value is worth to.
Embodiment two
In the present embodiment, somewhere on April 10th, 2012 to May 15 is chosen, on April 10th, 2013 to May 15,
On April 1st, 2014 to April 26, totally 2352 whole time point meteorological datas and voltage deviation data were as historical data, under docking
Carry out on April 27th, 2014, the voltage deviation of April 28 and May festivals or holidays whole time point on the 1st is predicted.
When carrying out principal component analysis dimension-reduction treatment to the meteorological data of historical time section and moment to be measured, obtain in principal component
The accumulative variance contribution ratio of preceding 4 principal components is up to 83.6%, if preceding 5 principal components of choosing, although up to 91%, dimension
Also corresponding to increase, it increases difficulty in computation and calculates the time, so select preceding 4 principal components as comprehensive meteorological variables, it is true
It proves, the requirement of precision can be met.
Obtained synthesis meteorological variables data are clustered by AP clustering algorithms, in the present embodiment, data point biAnd bj
Between similarity be s (i, j)=- (bi-bj)2, damped coefficient λ=0.5, Fig. 3 are set and show iterative process schematic diagram, from
It is found that when AP cluster iteration about 10 times tends to stablize in figure, it is 21 to obtain cluster numbers K, and obtains the cluster of 21 high quality
Center.
Initial cluster center of 21 cluster centres that AP is clustered as K-means clustering algorithms, to comprehensive meteorological
Variable data carries out K-means clusters, obtains 21 class clusters.
In the present embodiment, the iterations of BP neural network are set as 200 times, and learning rate 0.3, mean square error is set as
0.000004, input vector is the historical data of moment meteorological data same class cluster to be measured, therefore is 4 dimensional vectors, output
Vector is the voltage deviation at integral point moment, is 1 dimensional vector.Take the mean value of 10 prediction results as final prediction result.This reality
It applies in example and carries out prediction result with the prediction technique of traditional BP neural net prediction method, BP neural network combination PCA dimensionality reductions
Comparison obtains the prediction result comparison diagram in 27 days April on working day, April 28 as shown in figure 4, Fig. 5 compares for corresponding error
Figure.The prediction result comparison diagram in May 1 festivals or holidays is as shown in fig. 6, Fig. 7 is corresponding error comparison diagram.
It can be seen that the prediction technique of the application proposition is more pre- than traditional BP neural network, PCA+BP neural networks in Fig. 4~7
The average relative error for surveying result is all low, illustrates that PCA dimensionality reductions and AP+K-means two steps of cluster can carry in this method
High precision of prediction.Fig. 8 show the prediction result comparison diagram in April 28 and May 1, although it can be seen from the figure that working day
With festivals or holidays since larger fluctuation can occur for the difference of people's social activities, power quality, this prediction technique has remained to higher
Precision of prediction, it is seen that this method still has higher stability when external environment has certain variation.
Fig. 9~11 are respectively the prediction error distribution histogram of three kinds of prediction techniques in the present embodiment, it can be seen that Figure 11
Shown in the relative error of prediction technique that proposes of the application more concentrated near 0.The error comparison of three kinds of distinct methods is such as
Shown in table 2, the average relative error of this method is 2.987%, and 9.069%, and tradition are reduced than traditional BP neural network
Probability of the prediction result relative error of the BP neural network and PCA+BP neural networks control within 3% be respectively 16.67%,
23.61%, and probability of this method prediction result relative error control within 3% reaches 59.72%, extreme relative error is only
It is small probability event.In conclusion the method for the application proposition effect in voltage deviation prediction is more preferable.
2 prediction technique error of table compares
Claims (6)
1. a kind of based on the voltage deviation prediction technique for improving clustering algorithm and neural network, which is characterized in that including following step
Suddenly：
S1, the voltage deviation data for acquiring power grid historical time section different moments and meteorological data pass through as historical data
Weather forecast obtains the meteorological data at moment to be measured, and the meteorological data includes a variety of meteorologic factor variables；
S2, the obtained historical time sections of step S1 and the meteorological data at moment to be measured are carried out at dimensionality reduction by Principal Component Analysis
Reason chooses several principal components as comprehensive meteorological variables；
S3, the synthesis meteorological variables data of step S2 are clustered by the K-means clustering algorithms of combination AP clustering algorithms,
Obtain multiple class clusters；
S4, it chooses with the historical data of the meteorological data same class cluster at moment to be measured as training sample data collection, passes through BP god
It is trained through network and obtains the voltage deviation prediction result at moment to be measured.
2. a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network according to claim 1,
It is characterized in that, the step S3 is specifically included：
S31, the synthesis meteorological variables data of step S2 are clustered by AP clustering algorithms, obtains cluster numbers K and K cluster
Center；
S32, K cluster centre for obtaining step S31 are as the initial cluster center of K-means clustering algorithms, to step S2
Synthesis meteorological variables data carry out K-means clusters, obtain K class cluster.
3. a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network according to claim 1,
It is characterized in that, the meteorologic factor variable of the meteorological data includes：Temperature, dew point, humidity, air pressure, wind direction, wind speed and day are vaporous
Condition.
4. a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network according to claim 1,
It is characterized in that, the accumulative variance contribution ratio for the principal component chosen in the step S2 is not less than 80%.
5. a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network according to claim 1,
It is characterized in that, the date of historical time section is identical as the season belonging to the date at moment to be measured in the step S1.
6. a kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network according to claim 1,
It is characterized in that, the voltage deviation prediction result at moment to be measured is to take mean value by multiple voltage deviation predicted values in the step S4
It obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810178916.4A CN108460486A (en) | 2018-03-05 | 2018-03-05 | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810178916.4A CN108460486A (en) | 2018-03-05 | 2018-03-05 | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108460486A true CN108460486A (en) | 2018-08-28 |
Family
ID=63217297
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810178916.4A Pending CN108460486A (en) | 2018-03-05 | 2018-03-05 | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460486A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145997A (en) * | 2018-09-04 | 2019-01-04 | 格尔木美满新能源科技有限公司 | A kind of abandoning optical quantum prediction technique and device based on typical abandoning light field scape |
| CN109871976A (en) * | 2018-12-20 | 2019-06-11 | 浙江工业大学 | A kind of prediction technique of power quality containing distributed power distribution network based on cluster and neural network |
| CN109948695A (en) * | 2019-03-18 | 2019-06-28 | 云南电网有限责任公司 | A kind of power grid fragility node automatic identifying method based on neighbour's propagation clustering algorithm |
| CN110414788A (en) * | 2019-06-25 | 2019-11-05 | 国网上海市电力公司 | A kind of power quality prediction technique based on similar day and improvement LSTM |
| CN110570012A (en) * | 2019-08-05 | 2019-12-13 | 华中科技大学 | Storm-based power plant production equipment fault early warning method and system |
| CN110866074A (en) * | 2019-07-02 | 2020-03-06 | 黑龙江省电工仪器仪表工程技术研究中心有限公司 | Electric energy meter improved K-means classification method based on regional characteristics |
| CN115580526A (en) * | 2022-09-30 | 2023-01-06 | 中国人民解放军陆军工程大学 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| US9857778B1 (en) * | 2016-10-07 | 2018-01-02 | International Business Machines Corporation | Forecasting solar power generation using real-time power data, weather data, and complexity-based similarity factors |
-
2018
- 2018-03-05 CN CN201810178916.4A patent/CN108460486A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| US9857778B1 (en) * | 2016-10-07 | 2018-01-02 | International Business Machines Corporation | Forecasting solar power generation using real-time power data, weather data, and complexity-based similarity factors |
Non-Patent Citations (3)
| Title |
|---|
| 张介等: "基于聚类分析方法的风电场日前功率预测研究", 《浙江电力》 * |
| 蒋浩等: "主成分分析结合神经网络的光伏发电量预测", 《电力系统及其自动化学报》 * |
| 魏小曼等: "基于Affinity propagation 和K⁃means 算法的电力大用户细分方法分析", 《电力需求侧管理》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145997A (en) * | 2018-09-04 | 2019-01-04 | 格尔木美满新能源科技有限公司 | A kind of abandoning optical quantum prediction technique and device based on typical abandoning light field scape |
| CN109145997B (en) * | 2018-09-04 | 2022-03-11 | 格尔木美满新能源科技有限公司 | Light abandoning electric quantity prediction method and device based on typical light abandoning scene |
| CN109871976B (en) * | 2018-12-20 | 2020-12-25 | 浙江工业大学 | Clustering and neural network-based power quality prediction method for power distribution network with distributed power supply |
| CN109871976A (en) * | 2018-12-20 | 2019-06-11 | 浙江工业大学 | A kind of prediction technique of power quality containing distributed power distribution network based on cluster and neural network |
| CN109948695A (en) * | 2019-03-18 | 2019-06-28 | 云南电网有限责任公司 | A kind of power grid fragility node automatic identifying method based on neighbour's propagation clustering algorithm |
| CN110414788A (en) * | 2019-06-25 | 2019-11-05 | 国网上海市电力公司 | A kind of power quality prediction technique based on similar day and improvement LSTM |
| CN110414788B (en) * | 2019-06-25 | 2023-12-08 | 国网上海市电力公司 | Electric energy quality prediction method based on similar days and improved LSTM |
| CN110866074A (en) * | 2019-07-02 | 2020-03-06 | 黑龙江省电工仪器仪表工程技术研究中心有限公司 | Electric energy meter improved K-means classification method based on regional characteristics |
| CN110866074B (en) * | 2019-07-02 | 2022-11-04 | 黑龙江省电工仪器仪表工程技术研究中心有限公司 | Electric energy meter improved K-means classification method based on regional characteristics |
| CN110570012A (en) * | 2019-08-05 | 2019-12-13 | 华中科技大学 | Storm-based power plant production equipment fault early warning method and system |
| CN110570012B (en) * | 2019-08-05 | 2022-05-20 | 华中科技大学 | Storm-based power plant production equipment fault early warning method and system |
| CN115580526A (en) * | 2022-09-30 | 2023-01-06 | 中国人民解放军陆军工程大学 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
| CN115580526B (en) * | 2022-09-30 | 2024-03-22 | 中国人民解放军陆军工程大学 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460486A (en) | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network | |
| CN103324980B (en) | A kind of method for forecasting | |
| CN110991786B (en) | 10kV static load model parameter identification method based on similar daily load curve | |
| CN111199016A (en) | DTW-based improved K-means daily load curve clustering method | |
| CN110766200A (en) | Method for predicting generating power of wind turbine generator based on K-means mean clustering | |
| CN108549907B (en) | Data verification method based on multi-source transfer learning | |
| CN105447572A (en) | Wind power prediction system and method based on neural network optimized by genetic algorithm | |
| CN111882114B (en) | Short-time traffic flow prediction model construction method and prediction method | |
| CN110322075A (en) | A kind of scenic spot passenger flow forecast method and system based on hybrid optimization RBF neural | |
| CN110399686A (en) | A kind of unrelated aircraft flight profiles clustering method of parameter based on silhouette coefficient | |
| CN111695288B (en) | Transformer fault diagnosis method based on Apriori-BP algorithm | |
| CN110766313A (en) | Cable tunnel comprehensive state evaluation method based on operation and maintenance system | |
| Habib et al. | Retracted: Forecasting model for wind power integrating least squares support vector machine, singular spectrum analysis, deep belief network, and locality‐sensitive hashing | |
| Kumar et al. | Comparative analysis of SOM neural network with K-means clustering algorithm | |
| CN110796303B (en) | Short-term power load prediction method based on EWT and ODBSCAN | |
| CN116169670A (en) | Short-term non-resident load prediction method and system based on improved neural network | |
| CN115983374A (en) | Cable partial discharge database sample expansion method based on optimized SA-CACGAN | |
| CN108446358A (en) | The Data Modeling Method of optimization method and petrochemical equipment based on MIV and correlation rule | |
| CN111611293B (en) | Outlier data mining method based on feature weighting and MapReduce | |
| Sharma et al. | Aberration detection in electricity consumption using clustering technique | |
| CN108509537B (en) | System and method for forecasting galloping probability of power transmission line | |
| CN115859594A (en) | Health evaluation algorithm of power transformation equipment based on hierarchical analysis | |
| CN108985563A (en) | A kind of dynamically labeled method of Mechatronic Systems military service based on self-organizing feature map | |
| CN115079052A (en) | Transformer fault diagnosis method and system | |
| CN114423013A (en) | 5G heterogeneous network base station deployment method facing power distribution 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: 20180828 |
|
| RJ01 | Rejection of invention patent application after publication |