CN107563565A - A kind of short-term photovoltaic for considering Meteorology Factor Change decomposes Forecasting Methodology - Google Patents
A kind of short-term photovoltaic for considering Meteorology Factor Change decomposes Forecasting Methodology Download PDFInfo
- Publication number
- CN107563565A CN107563565A CN201710828824.1A CN201710828824A CN107563565A CN 107563565 A CN107563565 A CN 107563565A CN 201710828824 A CN201710828824 A CN 201710828824A CN 107563565 A CN107563565 A CN 107563565A
- Authority
- CN
- China
- Prior art keywords
- photovoltaic
- high frequency
- day
- series
- sequence
- 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
Landscapes
- Photovoltaic Devices (AREA)
Abstract
The invention discloses a kind of short-term photovoltaic for considering Meteorology Factor Change to decompose Forecasting Methodology, including：S1 is decomposed by singular spectrum analysis method to photovoltaic output time series, obtains low frequency sequence, high frequency series and noise sequence；S2 determines to influence the main weather factor of photovoltaic output using Pearson correlation coefficient method, and analyzes its sensitivity contributed to photovoltaic；S3 establishes the forecast model of consideration meteorologic factor for low frequency sequence and high frequency series and with reference to sensitivity respectively；S4 obtains low frequency sequence prediction value and high frequency series predicted value according to forecast model, and obtains photovoltaic power generation output forecasting value according to low frequency sequence prediction value and high frequency series predicted value.The present invention, which is contributed photovoltaic by singular spectrum analysis method, is decomposed into the feature that different subsequences individually analyzes each sequence；The influence degree contributed by the unit change amount of correlation analysis and the different meteorologic factors of sensitivity analysis acquisition to photovoltaic, more precisely to predict that photovoltaic is contributed.
Description
Technical field
The invention belongs to the batch (-type) regenerative resource electric powder prediction such as wind-powered electricity generation, photovoltaic, examined more particularly, to one kind
The short-term photovoltaic for considering Meteorology Factor Change decomposes Forecasting Methodology (Singular Spectrum Analysis Method
Considering Meteorological Factors, abbreviation SSA-MF method).
Background technology
Increasingly be promoted with the batch (-type) regenerative resource such as development, wind-powered electricity generation, photovoltaic of regenerative resource at high proportion and
Using.But the batch (-type) such as wind-powered electricity generation, photovoltaic regenerative resource has stronger randomness and fluctuation so that the peace of power system
Complete stable and economical operation faces significant challenge.Therefore, how more accurately to the batch (-type) regenerative resource such as wind-powered electricity generation, photovoltaic
It is predicted, there is important practical guided significance for the management and running of power system.
At present, mainly there is the combination side of physics class method, statistics class method and the above method for photovoltaic power generation output forecasting
The class of method etc. three.Physical method is that the factors such as detailed geographical position and photoelectric transformation efficiency according to where photovoltaic module establish physics
Model, the electricity generating principle according to photovoltaic system are directly predicted meteorological data as input.Its validity is depended on to grinding
Study carefully object immanent structure and its assurance degree of follow regularity and the precision of model parameter, be related to that link is more, process is complicated, parameter
Solve difficult.Statistics class method is built upon to go out force data to history photovoltaic using certain statistical method and analyze, and finds number
Inherent law in and for predicting.It mainly includes time series method, regression analysis, grey method and member and inspired
Formula series methods etc..The essence of meta-heuristic method is that the work and rest rule of biology is simulated, using certain algorithm to sample
Data are trained and obtain the relation between predicted condition and amount to be predicted.Meta-heuristic method mainly include neutral net,
SVMs, genetic algorithm, fuzzy system etc..Wherein neural network has very strong nonlinear fitting ability, Ke Yiying
Arbitrarily complicated non-linear relation is penetrated, the characteristics of this is with photovoltaic generating system is quite similar, so being well suited for photovoltaic plant
Power short-term forecast.But single neutral net can not adapt to that unsettled type is general, and prediction effect is bad.It is in addition, traditional
BP neural network training uses gradient descent method, is easily trapped into local minimum, convergence rate is slow.And fuzzy system goes out to photovoltaic
When power is predicted, the foundation of fuzzy inference rule needs substantial amounts of historical data and sufficient expertise.Combined method profit
With different models provide information and display one's respective advantages, select suitable mode to be combined, to improve prediction effect.Compared with
For first two method, combined method modeling is more complicated than single method, and implementation process is more difficult.
Described in synthesis, the above method models pre- after when contributing and being predicted to photovoltaic being handled using historical data
Survey, do not consider the photovoltaic power producing characteristics that the subsequence after data decomposition is reflected, do not excavate some implicit letters of photovoltaic output
Breath and inherent law, so wanting to reach preferable prediction effect relatively difficult to achieve.
The content of the invention
The defects of for prior art, it is an object of the invention to provide a kind of short-term photovoltaic for considering Meteorology Factor Change
Decompose Forecasting Methodology, it is intended to solve the problems, such as that precision of prediction is low in the prior art.
The invention provides a kind of short-term photovoltaic for considering Meteorology Factor Change to decompose Forecasting Methodology, comprises the steps：
S1：Photovoltaic output time series is decomposed by singular spectrum analysis method, obtains low frequency sequence, high frequency series
And noise sequence, and noise sequence is rejected；
S2：Determine to influence the main weather factor of photovoltaic output using Pearson correlation coefficient method, and analyze described main
The sensitivity that meteorologic factor is contributed to photovoltaic；
S3：Consideration meteorologic factor is established for the low frequency sequence and the high frequency series and respectively with reference to the sensitivity
The forecast model of high frequency series and the forecast model of low frequency sequence；
S4：High frequency series predicted value is obtained according to the forecast model of the high frequency series, according to the pre- of the low frequency sequence
Survey model and obtain low frequency sequence prediction value；And photovoltaic is obtained according to the low frequency sequence prediction value and the high frequency series predicted value
Output predicted value.
Further, step S1 is specially：
S11：Photovoltaic output time series is transformed into matrix form, and is d son of equal value therewith by the matrix decomposition
Matrix sum；
S12：D obtained submatrix will be decomposedLow frequency matrices Z is obtained after being groupedlow, it is high
Frequency matrix ZhighWith noise matrix Znoise, and by the low frequency matrices Zlow, high frequency matrix ZhighWith noise matrix ZnoiseIt is right respectively
Angle equalization obtains the low frequency sequence P after being reduced into the reproducing sequence of original series forml, the high frequency series PhWith it is described
Noise sequence Pn。
Further, in step S2 using Pearson correlation coefficient method determine influence photovoltaic contribute main meteorological because
Element is specially：
S21：Temperature, irradiation, wind speed and rainfall are selected as meteorologic factor；
S22：According to formulaRespectively calculate photovoltaic contribute with temperature, irradiation, wind speed or
Pearson correlation coefficient between rainfall；
S23：Determined to influence the main weather factor of photovoltaic output according to the size of Pearson correlation coefficient；
Wherein,Pearson correlation coefficientAbsolute valueCloser to 1, show two changes
It is higher to measure linearly related degree.
Further, foundation considers that the forecast model of the high frequency series of meteorologic factor is specially in step S3：
(1) reference day and a reference value of high frequency series are chosen：
Using the previous day of day to be predicted as high frequency series with reference to day, and using the photovoltaic output high frequency series with reference to day as
The a reference value of day high frequency series to be predicted；
(2) Pearson correlation coefficient using between different meteorologic factors and photovoltaic output is used as the meteorological factor influence light
The weight coefficient of volt output change；
(3) sensitivity changed, day to be predicted and the temperature difference with reference to day and irradiation of being contributed according to meteorologic factor to photovoltaic are poor,
And according to formula Phigh=P'high+α1ΔP1+α2ΔP2To photovoltaic output high frequency series PhighIt is modified；
Wherein, PhighFor the photovoltaic output high frequency series of day to be predicted, P'highFor with reference to day photovoltaic output high frequency series,
ΔP1For the photovoltaic output high frequency series variable quantity caused by temperature change, Δ P2Contributed for the photovoltaic caused by irradiating change high
Frequency sequence variable quantity；α1The weight coefficient of photovoltaic output high frequency series change, α are influenceed for temperature2It is high that photovoltaic output is influenceed for irradiation
The weight coefficient of frequency sequence change.
Further, when degree/day to be predicted with reference to degree/day with being in same sensitivity section, Δ P1=St(t-
t')；When degree/day to be predicted with reference to degree/day from being in two different sensitivity sections,
Wherein, t is day temperature angle value to be predicted, and t' is with reference to day temperature angle value, StThe spirit in section where degree/day to be predicted
Sensitivity, S'tFor the sensitivity with reference to section where degree/day,Represent the temperature value of two section public points.
Further, in step S4, according to low frequency sequence prediction value PlowWith high frequency series predicted value PhighObtain photovoltaic
Output predicted value P=Plow+Phigh。
Photovoltaic is contributed and is decomposed into different subsequences by the present invention by singular spectrum analysis method, can individually analyze each sequence
The feature of row；By correlation analysis and sensitivity analysis, the unit change amount that can obtain different meteorologic factors goes out to photovoltaic
The influence degree of power, more precisely to predict that photovoltaic is contributed, favourable data reference is provided for scheduling decision personnel, so as to
Reduce the impact that photovoltaic output access is brought to power system.
Brief description of the drawings
Fig. 1 is SSA-MF methods figure provided in an embodiment of the present invention；
Fig. 2 is singular spectrum analysis techniqueflow chart provided in an embodiment of the present invention；
Fig. 3 is in April, 2014 photovoltaic output Decomposition Sequence；(a) contributed for history photovoltaic；(b) contributed for history photovoltaic
Low frequency sequence；(c) it is history photovoltaic output high frequency series；(d) it is history photovoltaic output noise sequence；
Fig. 4 be May 12 photovoltaic output real data with prediction data curve (SSA-MF)；(a) go out for photovoltaic on May 12
Power low frequency sequence prediction value；(b) it is photovoltaic output high frequency series predicted value on May 12；(c) it is photovoltaic power generation output forecasting on May 12
Value.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Conventional forecast model and method is often difficult in adapt to the intermittent change of photovoltaic output, to the subsequence after decomposition
Do not carry out going deep into mining analysis, the processing for related meteorologic factor of contributing to photovoltaic is also complex yet so that realize it is difficult or
Person's precision of prediction is not high.It is an object of the invention to overcome conventional photovoltaic to go out two following aspect limitations of force prediction method：(1)
When not considering meteorologic factor, precision of prediction is not high, it is difficult to reaches preferable prediction effect；(2) when considering meteorologic factor, to meteorology
The processing method and process of factor are complex, realize that difficulty is big in practice.Therefore, the present invention provides a kind of consideration gas
The short-term photovoltaic changed as factor decomposes Forecasting Methodology；Method in the present invention can be with by correlation analysis and sensitivity analysis
Simplify the processing procedure that complicated meteorology factor influences on photovoltaic output, history photovoltaic can be gone out force data and be decomposed, to dividing
Subsequence after solution is analyzed and predicted respectively so that prediction result has higher precision, ensures electric power after photovoltaic access
The safe and stable operation of system.
Fig. 1 shows SSA-MF method figures, and a kind of short-term photovoltaic for considering meteorologic factor provided by the invention is contributed
The singular spectrum analysis method of prediction, comprises the following steps：
(1) photovoltaic output time series is decomposed by singular spectrum analysis technology, obtains low frequency sequence, high frequency series
And noise sequence.
Singular spectrum analysis (SSA) method is a kind of technology for time series analysis and prediction, signal is carried out unusual
Value, which decomposes (Singular Value Decomposition, abbreviation SVD decomposition), can obtain trend characteristic, the week of primary signal
Phase characteristic and white noise character etc., be advantageous to the analysis to primary signal.Fig. 2 is singular spectrum analysis techniqueflow chart, mainly
Including decomposing and reconstructing two complementary stages.It is that original time series P first is changed into a matrix form that SSA, which is decomposed, is recycled
SVD decomposes to obtain the d submatrix of equal value with original matrix.SSA reconstruct is that the submatrix after first SVD is decomposed is grouped, and is obtained
Low frequency, high frequency and noise sequence are obtained to low frequency, high frequency and noise matrix, then to its diagonal equalization.
Step (1) is specially：
(11)：SSA is decomposed：The basic thought of SSA decomposition methods is that original time series is transformed into a matrix form, then
It is multiple submatrix sums of equal value therewith by the matrix decomposition.SSA decompose be broadly divided into embedding operation (Embedding) and
SVD decomposes (Singular Value Decomposition) two steps.
(111)：Embedding operation：Embedding operation is the original one-dimensional sequential photovoltaic output P=(P for N (N ＞ 2) by length1,
P2,L,PN) it is converted into multi-dimensional time sequence photovoltaic output matrix Z=[Z1,Z2,L,ZK] a kind of map operation, i.e. P=(P1,P2,L,
PN) → Z=[Z1,Z2,L,ZK]……(1)；Wherein, Zi(i=1,2, L, K) be matrix Z a certain row, Zi=(Pi,Pi+1,L,
Pi+L-1)T∈RL, L dimensions are shared, L is Embedded dimensions (2≤L≤N), K=N-L+1.Generally, L selection is no more than whole sequence
The 1/3 of length.Matrix Z is referred to as track matrix (Trajectory Matrix), i.e.,：So far, one-dimensional photovoltaic output sequence is completed to multidimensional photovoltaic
The conversion of output matrix.
(112)：SVD is decomposed：SVD decomposes is decomposed into d submatrix by the track matrix Z of formula (2)D is square
Battle array Z order, and cause d submatrix sum to be equal to matrix Z, i.e.,： Calculated by formula (4)In formula, λ1,λ2,L,λL(λ1≥λ2
≥L≥λL>=0) it is S=ZZTCharacteristic value, U1,U2,L,ULIt is characterized vectorial normal orthogonal system.
VjTried to achieve by formula (5)Wherein, UjAnd VjTrack square is represented respectively
Battle array Z left and right characteristic vector,For track matrix Z singular value, setFor matrix Z singular spectrum,It is collectively forming a feature ringSo far, complete SVD to decompose, obtain corresponding to matrix Z singular spectrum
D submatrix.
(12)：SSA is reconstructed：SSA reconstruct is that SVD first is decomposed into obtained d submatrixDivided
Group, low frequency/high frequency/noise matrix is obtained, is designated as Z respectivelylow、ZhighAnd Znoise；It is again that low frequency/high frequency/noise matrix is right respectively
Angle equalizes the reproducing sequence for being reduced into original series form, i.e. low frequency sequence Pl, high frequency series PhWith noise sequence Pn.SSA weights
Structure mainly includes packet (Grouping) and diagonal equalization (Diagonal Averaging) two steps.
(121)：Packet：Contribution rate η according to preceding r (r >=0) individual singular value to matrix Z singular value sum, and it is unusual
The situation of large jump occurs for value, is grouped.For example, it is assumed that λ ' and λ " is matrix S two unequal characteristic values, and λ '
And λ " much smaller than its previous eigenvalue λ '0And λ "0.Work as λjDuring >=λ ', λjCorresponding matrix ZjIt is considered as low frequency submatrix；λ'
＞ λjDuring ＞ λ ", λjCorresponding matrix ZjIt is considered as high frequency submatrix；λjDuring≤λ ', λjCorresponding matrix ZjIt is considered as the sub- square of noise
Battle array.Specifically it is grouped depending on actual conditions.The d submatrix that formula (3) can be obtained is divided into low frequency/height as shown in formula (6)
Frequently/noise matrix.Contribution rate η calculation formula such as formula
(7) shown in：
(122)：Diagonal equalization：Step S121 is further grouped to the low frequency/high frequency/noise matrix conversion growth determined
Low frequency/high frequency/noise sequence for N is spent, below with high frequency matrix ZhighExemplified by illustrate.
It is assumed that ZhighFor a × b matrix, ZijFor ZhighEither element, remember a*=min (a, b), b*=max (a, b), N
=a+b-1, and as a ＜ b,Otherwise,Reproducing sequence RC=(rc corresponding to above-mentioned packet matrix1,
rc2,…,rcN) can be tried to achieve by following formula：
By formula (8), you can obtain low frequency
Sequence Ph, can similarly try to achieve low frequency sequence PlAnd noise sequence Pn。
(2) determine to influence the main weather factor of photovoltaic output using Pearson correlation coefficient method.
Concrete methods of realizing is as follows：Correlation analysis is carried out to photovoltaic output time series and different meteorologic factors.This hair
The bright Pearson correlation coefficient method using as shown in formula (9).1) meteorologic factor is selected, the present invention is to temperature, irradiation, wind speed, drop
The different meteorologic factors such as rainfall are studied；2) photovoltaic is calculated respectively according to formula (9) to contribute and temperature, irradiation, wind speed, rainfall
It is as shown in table 1 etc. the Pearson correlation coefficient between meteorologic factor, its result of calculation；3) according to the size of coefficient correlation in 2)
It is determined that influence the main weather factor that photovoltaic is contributed.
Wherein,Phase
Relation numberAbsolute valueCloser to 1, show that two linear variable displacement degrees of correlation are higher.
The meteorological data of table 1 goes out the relative coefficient of force data with light
(3) main weather factor determined according to step (2), analyzes the sensitivity that each main meteorological is contributed to photovoltaic.
Sensitivity of the main weather factor to photovoltaic output change is analyzed, what the sensitivity that photovoltaic is contributed to meteorologic factor referred to
It is the variable quantity that photovoltaic is contributed under unit Meteorology Factor Change.
Sensitivity of the main weather factor to photovoltaic output change is analyzed, what the sensitivity that photovoltaic is contributed to meteorologic factor referred to
It is the variable quantity that photovoltaic is contributed under unit Meteorology Factor Change.Specific operation process is as follows：According to the main meteorological determined in 2 because
Plain (being illustrated below by taking temperature and irradiation intensity as an example), photovoltaic is analyzed respectively and is contributed to the sensitive of temperature and irradiation intensity
Degree, its solving result is as shown in table 2, table 3.
The sensitivity that the photovoltaic of table 2 is contributed to temperature Change
| Temperature section/DEG C | Sensitivity/kW | Temperature section/DEG C | Sensitivity/kW |
| - | - | [26,27) | 13.996 |
| [16,17) | 15.186 | [27,28) | 13.877 |
| [17,18) | 15.067 | [28,29) | 13.758 |
| [18,19) | 14.948 | [29,30) | 13.640 |
| [19,20) | 14.829 | [30,31) | 13.521 |
| [20,21) | 14.710 | [31,32) | 13.402 |
| [21,22) | 14.591 | [32,33) | 13.283 |
| [22,23) | 14.472 | [33,34) | 13.164 |
| [23,24) | 14.353 | [34,35) | 13.045 |
| [24,25) | 14.234 | [35,36) | 12.926 |
| [25,26) | 14.115 | - | - |
Sensitivity of the photovoltaic of table 3 output to irradiation intensity change
(4) low frequency sequence and high frequency series are directed to, establish the forecast model for considering meteorologic factor respectively.
The step of in view of to low frequency/high frequency series modeling and forecasting, is identical, below by taking the prediction process of high frequency series as an example
Illustrate.
Specific practice is as follows：
A. reference day and a reference value of high frequency series are chosen, the reference day using the previous day of day to be predicted as high frequency series,
And a reference value of day high frequency series to be predicted is used as using the photovoltaic output high frequency series with reference to day.
B. the Pearson correlation coefficient using between different meteorologic factors and photovoltaic output is used as the meteorological factor influence photovoltaic
The weight coefficient of output change.
C. the sensitivity changed, day to be predicted and the temperature difference with reference to day and irradiation of being contributed according to meteorologic factor to photovoltaic are poor,
According to formula (10) to photovoltaic output high frequency series PhighIt is modified.
Phigh=P'high+α1ΔP1+α2ΔP2……(10)；In formula：Phigh,P'high,ΔP1With Δ P2It is respectively to be predicted
The photovoltaic output high frequency series of day, the photovoltaic output high frequency series with reference to day, the photovoltaic output high frequency sequence caused by temperature change
Row variable quantity and the photovoltaic output high frequency series variable quantity caused by irradiating change.α1And α2Respectively temperature and irradiation influences
The weight coefficient of photovoltaic output high frequency series change.
From formula (10) as can be seen that the amount of amendment includes Δ P1With Δ P2, it is of the invention to determine both take using following methods
Value, below with Δ P1Exemplified by illustrate.
(a) when degree/day to be predicted with reference to degree/day with being in same sensitivity section：
ΔP1=St(t-t')……(11)；
(b) it is such as adjacent with two when degree/day to be predicted with reference to degree/day from being in two different sensitivity sections
Exemplified by section, thenIn formula：T and t' represents degree/day to be predicted and reference respectively
Day temperature angle value；StAnd S'tRepresent respectively degree/day to be predicted with reference to degree/day each place section sensitivity；Represent two
The temperature value of section public point.
(5) step (4) is obtained into low frequency sequence prediction value and high frequency series predicted value to be overlapped according to formula (13), obtained
Photovoltaic power generation output forecasting value.
P=Plow+Phigh……(13)；Wherein, Plow、PhighRespectively low frequency sequence, high frequency series predicted value.
Case study on implementation of the present invention is described in further detail below in conjunction with accompanying drawing.
The present invention proposes a kind of method suitable for the prediction of the batch (-type) regenerative resources such as wind-powered electricity generation, photovoltaic；It is specific to provide
A kind of singular spectrum analysis method for the short-term photovoltaic power generation output forecasting for considering meteorologic factor.This method can pass through singular spectrum analysis
Photovoltaic is contributed and is decomposed into different subsequences, Ke Yidan by technology (Singular Spectrum Analysis, abbreviation SSA method)
Solely analyze the feature of each sequence；By correlation analysis and sensitivity analysis, the unit change of different meteorologic factors can be obtained
The influence degree contributed to photovoltaic is measured, more precisely to predict that photovoltaic is contributed, favourable number is provided for scheduling decision personnel
According to reference, so as to reduce the impact that photovoltaic output access is brought to power system.
Implementation steps 1：Obtain photovoltaic output, temperature, irradiation, wind speed, the drop on May 30,1 day to 2014 May in 2013
The data such as the temperature in rainfall and in May, 2014, irradiation.The present invention is gone out using following 1 day photovoltaic of the data prediction of 1 year in the past
Power, using the data on May 30th, 1 day 1 May in 2013 as forecast sample in invention, the data conduct in May, 2014
Test sample.
Implementation steps 2：Photovoltaic output time series is decomposed using SSA technologies, obtain photovoltaic output low frequency sequence,
High frequency series and noise sequence are as shown in Figure 1.It is right because noise sequence is formed by the submatrix reconstruct of characteristic value accounting very little
The influence of initial data is little, considers to reject noise sequence.Therefore, emphasis of the present invention is carried out to low frequency sequence and high frequency series
Prediction.
Implementation steps 3：Determine to influence the main weather factor of photovoltaic output change, root using Pearson correlation coefficient method
Know according to knot is calculated in table 1, temperature and irradiation intensity are to influence the main weather factor that photovoltaic is contributed.
Implementation steps 4：Calculate main weather factor to be temperature and irradiation intensity contribute to photovoltaic the sensitivity of change, such as table
2 and table 3 shown in.
Implementation steps 5：According to the result of Calculation of Sensitivity, prediction is modeled respectively to low frequency sequence and high frequency series.
Low frequency sequence and the prediction result of high frequency series are obtained, as shown in Fig. 4 (a), 4 (b).
Implementation steps 6：Low frequency sequence and high frequency series that step 5 obtains are overlapped, obtain Fig. 4 (c) photovoltaics output
Prediction result.
As it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention, not to
The limitation present invention, all any modification, equivalent and improvement made within the spirit and principles of the invention etc., all should be included
Within protection scope of the present invention.
Claims (6)
1. a kind of short-term photovoltaic for considering Meteorology Factor Change decomposes Forecasting Methodology, it is characterised in that comprises the steps：
S1：Photovoltaic output time series is decomposed by singular spectrum analysis method, low frequency sequence, high frequency series is obtained and makes an uproar
Sound sequence, while cancelling noise sequence；
S2：Determine to influence the main weather factor of photovoltaic output using Pearson correlation coefficient method, and analyze the main meteorological
The sensitivity that factor is contributed to photovoltaic；
S3：Establish the height for considering meteorologic factor respectively for the low frequency sequence and the high frequency series and with reference to the sensitivity
The forecast model of frequency sequence and the forecast model of low frequency sequence；
S4：High frequency series predicted value is obtained according to the forecast model of the high frequency series, according to the prediction mould of the low frequency sequence
Type obtains low frequency sequence prediction value；And photovoltaic is obtained according to the low frequency sequence prediction value and the high frequency series predicted value and contributed
Predicted value.
2. photovoltaic as claimed in claim 1 short-term decomposes Forecasting Methodology, it is characterised in that step S1 is specially：
S11：Photovoltaic output time series is transformed into matrix form, and is d submatrix of equal value therewith by the matrix decomposition
Sum；
S12：D obtained submatrix will be decomposedLow frequency matrices Z is obtained after being groupedlow, high frequency matrix
ZhighWith noise matrix Znoise, and by the low frequency matrices Zlow, high frequency matrix ZhighWith noise matrix ZnoiseIt is diagonal average respectively
Change obtains the low frequency sequence P after being reduced into the reproducing sequence of original series forml, the high frequency series PhWith the noise sequence
Arrange Pn。
3. short-term photovoltaic as claimed in claim 1 or 2 decomposes Forecasting Methodology, it is characterised in that Pearson is used in step S2
Correlation coefficient process determines that influenceing the main weather factor that photovoltaic is contributed is specially：
S21：Temperature, irradiation, wind speed and rainfall are selected as meteorologic factor；
S22：According to formulaPhotovoltaic is calculated respectively to contribute and temperature, irradiation, wind speed or rainfall
Pearson correlation coefficient between amount；
S23：Determined to influence the main weather factor of photovoltaic output according to the size of Pearson correlation coefficient；
Wherein,Pearson correlation coefficientAbsolute valueCloser to 1, show two variable lines
Property degree of correlation is higher.
4. the short-term photovoltaic as described in claim any one of 1-3 decomposes Forecasting Methodology, it is characterised in that establishes and examines in step S3
The forecast model of high frequency series for considering meteorologic factor is specially：
(1) reference day and a reference value of high frequency series are chosen：
The reference day using the previous day of day to be predicted as high frequency series, and it is pre- using the photovoltaic output high frequency series with reference to day as treating
Survey a reference value of day high frequency series；
(2) Pearson correlation coefficient using between different meteorologic factors and photovoltaic output goes out as the meteorological factor influence photovoltaic
The weight coefficient of power change；
(3) sensitivity changed, day to be predicted and the temperature difference with reference to day and irradiation of being contributed according to meteorologic factor to photovoltaic are poor, and root
According to formula Phigh=Ph'igh+α1ΔP1+α2ΔP2To photovoltaic output high frequency series PhighIt is modified；
Wherein, PhighFor the photovoltaic output high frequency series of day to be predicted, Ph'ighFor the photovoltaic output high frequency series with reference to day, Δ P1
For the photovoltaic output high frequency series variable quantity caused by temperature change, Δ P2For the photovoltaic output high frequency sequence caused by irradiating change
Row variable quantity；α1The weight coefficient of photovoltaic output high frequency series change, α are influenceed for temperature2Photovoltaic output high frequency sequence is influenceed for irradiation
Arrange the weight coefficient of change.
5. short-term photovoltaic as claimed in claim 4 decomposes Forecasting Methodology, it is characterised in that when degree/day to be predicted and with reference to day
When temperature is in same sensitivity section, Δ P1=St(t-t')；When degree/day to be predicted with reference to degree/day with being in two differences
Sensitivity section when,
Wherein, t is day temperature angle value to be predicted, and t' is with reference to day temperature angle value, StThe sensitivity in section where degree/day to be predicted,
St' it is sensitivity with reference to section where degree/day,Represent the temperature value of two section public points.
6. the short-term photovoltaic as described in claim any one of 1-5 decomposes Forecasting Methodology, it is characterised in that in step S4, according to
Low frequency sequence prediction value PlowWith high frequency series predicted value PhighObtain photovoltaic power generation output forecasting value P=Plow+Phigh。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710828824.1A CN107563565B (en) | 2017-09-14 | 2017-09-14 | A kind of short-term photovoltaic decomposition prediction technique considering Meteorology Factor Change |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710828824.1A CN107563565B (en) | 2017-09-14 | 2017-09-14 | A kind of short-term photovoltaic decomposition prediction technique considering Meteorology Factor Change |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107563565A true CN107563565A (en) | 2018-01-09 |
| CN107563565B CN107563565B (en) | 2019-04-02 |
Family
ID=60979951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710828824.1A Active CN107563565B (en) | 2017-09-14 | 2017-09-14 | A kind of short-term photovoltaic decomposition prediction technique considering Meteorology Factor Change |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107563565B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108518804A (en) * | 2018-03-21 | 2018-09-11 | 武汉物联远科技有限公司 | A kind of machine room humiture environmental forecasting method and system |
| CN109102102A (en) * | 2018-06-25 | 2018-12-28 | 南京邮电大学 | Based on the photovoltaic of multivariate phase space reconstruction and SVR power output short term prediction method |
| CN109767043A (en) * | 2019-01-17 | 2019-05-17 | 中南大学 | A kind of electric load time series big data intelligent modeling and prediction technique |
| CN109919353A (en) * | 2019-01-11 | 2019-06-21 | 杭州电子科技大学 | A kind of distributed photovoltaic prediction technique of the ARIMA model based on spatial coherence |
| CN111340288A (en) * | 2020-02-25 | 2020-06-26 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering space-time correlation |
| CN111539550A (en) * | 2020-03-13 | 2020-08-14 | 远景智能国际私人投资有限公司 | Method, device and equipment for determining working state of photovoltaic array and storage medium |
| CN111738154A (en) * | 2020-06-23 | 2020-10-02 | 广西大学 | RNN-based large civil engineering acceleration response time-frequency domain decomposition method |
| CN112132328A (en) * | 2020-09-04 | 2020-12-25 | 国网上海市电力公司 | Photovoltaic output power ultra-short-term local emotion reconstruction neural network prediction method |
| CN112561178A (en) * | 2020-12-21 | 2021-03-26 | 国网甘肃省电力公司电力科学研究院 | Distributed photovoltaic output prediction method considering microclimate factors |
| CN113779861A (en) * | 2021-07-23 | 2021-12-10 | 国网河北省电力有限公司电力科学研究院 | Photovoltaic power prediction method and terminal equipment |
| CN114188989A (en) * | 2021-12-14 | 2022-03-15 | 福州大学 | Distributed photovoltaic power station output prediction method considering correlation |
| CN117709556A (en) * | 2024-02-06 | 2024-03-15 | 山东建筑大学 | Photovoltaic power generation short-term prediction method, system, medium and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110060703A1 (en) * | 2009-09-04 | 2011-03-10 | Alex Alaniz | Method and system for detecting correlation in data sets |
| CN103136598A (en) * | 2013-02-26 | 2013-06-05 | 福建省电力有限公司 | Monthly electrical load computer forecasting method based on wavelet analysis |
| CN105373855A (en) * | 2015-11-19 | 2016-03-02 | 国网河南省电力公司电力科学研究院 | Electric power system daily load prediction method used for special weather and special dates |
| CN106503851A (en) * | 2016-10-25 | 2017-03-15 | 国家电网公司 | A kind of improved Short-Term Load Forecasting Method based on wavelet analysises |
-
2017
- 2017-09-14 CN CN201710828824.1A patent/CN107563565B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110060703A1 (en) * | 2009-09-04 | 2011-03-10 | Alex Alaniz | Method and system for detecting correlation in data sets |
| CN103136598A (en) * | 2013-02-26 | 2013-06-05 | 福建省电力有限公司 | Monthly electrical load computer forecasting method based on wavelet analysis |
| CN105373855A (en) * | 2015-11-19 | 2016-03-02 | 国网河南省电力公司电力科学研究院 | Electric power system daily load prediction method used for special weather and special dates |
| CN106503851A (en) * | 2016-10-25 | 2017-03-15 | 国家电网公司 | A kind of improved Short-Term Load Forecasting Method based on wavelet analysises |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108518804B (en) * | 2018-03-21 | 2021-06-25 | 武汉物联远科技有限公司 | Method and system for predicting temperature and humidity environment of machine room |
| CN108518804A (en) * | 2018-03-21 | 2018-09-11 | 武汉物联远科技有限公司 | A kind of machine room humiture environmental forecasting method and system |
| CN109102102A (en) * | 2018-06-25 | 2018-12-28 | 南京邮电大学 | Based on the photovoltaic of multivariate phase space reconstruction and SVR power output short term prediction method |
| CN109919353A (en) * | 2019-01-11 | 2019-06-21 | 杭州电子科技大学 | A kind of distributed photovoltaic prediction technique of the ARIMA model based on spatial coherence |
| CN109919353B (en) * | 2019-01-11 | 2022-07-05 | 杭州电子科技大学 | Distributed photovoltaic prediction method of ARIMA model based on spatial correlation |
| CN109767043A (en) * | 2019-01-17 | 2019-05-17 | 中南大学 | A kind of electric load time series big data intelligent modeling and prediction technique |
| CN111340288A (en) * | 2020-02-25 | 2020-06-26 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering space-time correlation |
| CN111340288B (en) * | 2020-02-25 | 2024-04-05 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering time-space correlation |
| CN111539550A (en) * | 2020-03-13 | 2020-08-14 | 远景智能国际私人投资有限公司 | Method, device and equipment for determining working state of photovoltaic array and storage medium |
| CN111738154A (en) * | 2020-06-23 | 2020-10-02 | 广西大学 | RNN-based large civil engineering acceleration response time-frequency domain decomposition method |
| CN112132328A (en) * | 2020-09-04 | 2020-12-25 | 国网上海市电力公司 | Photovoltaic output power ultra-short-term local emotion reconstruction neural network prediction method |
| CN112561178A (en) * | 2020-12-21 | 2021-03-26 | 国网甘肃省电力公司电力科学研究院 | Distributed photovoltaic output prediction method considering microclimate factors |
| CN113779861A (en) * | 2021-07-23 | 2021-12-10 | 国网河北省电力有限公司电力科学研究院 | Photovoltaic power prediction method and terminal equipment |
| CN113779861B (en) * | 2021-07-23 | 2023-08-22 | 国网河北省电力有限公司电力科学研究院 | Photovoltaic Power Prediction Method and Terminal Equipment |
| CN114188989A (en) * | 2021-12-14 | 2022-03-15 | 福州大学 | Distributed photovoltaic power station output prediction method considering correlation |
| CN114188989B (en) * | 2021-12-14 | 2024-04-12 | 福州大学 | Correlation-considered distributed photovoltaic power station output prediction method |
| CN117709556A (en) * | 2024-02-06 | 2024-03-15 | 山东建筑大学 | Photovoltaic power generation short-term prediction method, system, medium and equipment |
| CN117709556B (en) * | 2024-02-06 | 2024-05-31 | 山东建筑大学 | Photovoltaic power generation short-term prediction method, system, medium and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107563565B (en) | 2019-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107563565A (en) | A kind of short-term photovoltaic for considering Meteorology Factor Change decomposes Forecasting Methodology | |
| Zhao et al. | Research and application based on the swarm intelligence algorithm and artificial intelligence for wind farm decision system | |
| Zhou et al. | Prediction of photovoltaic power output based on similar day analysis, genetic algorithm and extreme learning machine | |
| Lv et al. | A VMD and LSTM based hybrid model of load forecasting for power grid security | |
| Dong et al. | Hourly energy consumption prediction of an office building based on ensemble learning and energy consumption pattern classification | |
| Davò et al. | Post-processing techniques and principal component analysis for regional wind power and solar irradiance forecasting | |
| Zhang et al. | Probabilistic solar irradiation forecasting based on variational Bayesian inference with secure federated learning | |
| De Giorgi et al. | Error analysis of hybrid photovoltaic power forecasting models: A case study of mediterranean climate | |
| CN109902874A (en) | A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning | |
| CN110070226A (en) | Photovoltaic power prediction technique and system based on convolutional neural networks and meta learning | |
| Succetti et al. | Deep neural networks for multivariate prediction of photovoltaic power time series | |
| Voyant et al. | Twenty four hours ahead global irradiation forecasting using multi‐layer perceptron | |
| Moayyed et al. | A Cyber-Secure generalized supermodel for wind power forecasting based on deep federated learning and image processing | |
| CN104573879A (en) | Photovoltaic power station output predicting method based on optimal similar day set | |
| CN110309603A (en) | A kind of short-term wind speed forecasting method and system based on wind speed characteristics | |
| CN106251027B (en) | Electric load probability density Forecasting Methodology based on fuzzy support vector quantile estimate | |
| CN101793907A (en) | Short-term wind speed forecasting method of wind farm | |
| Lu et al. | A hybrid approach for day-ahead forecast of PV power generation | |
| CN108549960A (en) | A kind of 24 hours Methods of electric load forecasting | |
| Abedinia et al. | Net demand prediction for power systems by a new neural network‐based forecasting engine | |
| CN103514457A (en) | Multi-target improvement differential evolution model for land use structure optimization under ecological constraint conditions | |
| CN110555515A (en) | Short-term wind speed prediction method based on EEMD and LSTM | |
| Cococcioni et al. | One day-ahead forecasting of energy production in solar photovoltaic installations: An empirical study | |
| Kaplan et al. | A novel method based on Weibull distribution for short-term wind speed prediction | |
| Nhita | A rainfall forecasting using fuzzy system based on genetic algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |