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
- frequency sequence
- photovoltaic
- photovoltaic output
- sequence
- low
- 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
- 238000000034 method Methods 0.000 title claims abstract description 67
- 230000008859 change Effects 0.000 title claims abstract description 26
- 230000035945 sensitivity Effects 0.000 claims abstract description 31
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 45
- 238000000354 decomposition reaction Methods 0.000 claims description 28
- 238000012935 Averaging Methods 0.000 claims description 5
- 238000010219 correlation analysis Methods 0.000 abstract description 5
- 238000010206 sensitivity analysis Methods 0.000 abstract description 4
- 238000010248 power generation Methods 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
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 technical field of wind power, photovoltaic and other intermittent renewable energy source prediction, and particularly relates to a short-term photovoltaic decomposition prediction method (SSA-MF method for short) considering Meteorological factor change.
Background
With the development of high-proportion renewable energy, intermittent renewable energy sources such as wind power and photovoltaic are increasingly popularized and applied. But intermittent renewable energy sources such as wind power, photovoltaic and the like have strong randomness and volatility, so that the safe stability and economic operation of a power system face important challenges. Therefore, how to accurately predict intermittent renewable energy sources such as wind power, photovoltaic and the like has important practical guiding significance for dispatching and running of the power system.
At present, three types of methods, such as a physical method, a statistical method, a combination method of the methods and the like, are mainly used for photovoltaic output prediction. The physical method is that a physical model is established according to factors such as the detailed geographic position of the photovoltaic module, the photoelectric conversion efficiency and the like, and meteorological data is directly used as input for prediction according to the power generation principle of the photovoltaic system. The effectiveness depends on the degree of mastering the internal structure and the following rule of the research object and the precision of the model parameters, and the method has the advantages of multiple involved links, complex process and difficult parameter solving. The statistical method is established by analyzing historical photovoltaic output data by using a certain statistical method, searching for an internal rule in the data and using the internal rule for prediction. The method mainly comprises a time series method, a regression analysis method, a gray prediction method, a meta-heuristic series method and the like. The essence of the meta-heuristic method is to simulate the work and rest rules of the living being, and train sample data by adopting a certain algorithm to obtain the relationship between the prediction condition and the quantity to be predicted. The meta-heuristic method mainly comprises a neural network, a support vector machine, a genetic algorithm, a fuzzy system and the like. The neural network method has strong nonlinear fitting capacity, can map any complex nonlinear relation, is very similar to the characteristics of a photovoltaic power generation system, and is very suitable for short-term prediction of the output of a photovoltaic power station. However, a single neural network cannot adapt to variable weather types and has poor prediction effect. In addition, the traditional BP neural network training adopts a gradient descent method, so that the traditional BP neural network training is easy to fall into a local minimum value, and the convergence rate is low. When the fuzzy system predicts the photovoltaic output, a large amount of historical data and sufficient expert experience are needed for establishing the fuzzy inference rule. The combination method utilizes the information provided by different models and exerts respective advantages, and selects a proper mode for combination so as to improve the prediction effect. Compared with the former two methods, the combined method is more complex to model than the single method, and the realization process is more difficult.
In summary, when the method is used for predicting the photovoltaic output, modeling prediction is carried out after historical data are processed, photovoltaic output characteristics reflected by subsequences after data decomposition are not considered, and some implicit information and internal rules of the photovoltaic output are not mined, so that a better prediction effect is difficult to achieve.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a short-term photovoltaic decomposition prediction method considering meteorological factor change, and aims to solve the problem of low prediction precision in the prior art.
The invention provides a short-term photovoltaic decomposition prediction method considering meteorological factor change, which comprises the following steps of:
s1: decomposing the photovoltaic output time sequence by a singular spectrum analysis method to obtain a low-frequency sequence, a high-frequency sequence and a noise sequence, and removing the noise sequence;
s2: determining main meteorological factors influencing photovoltaic output by adopting a Pearson correlation coefficient method, and analyzing the sensitivity of the main meteorological factors to the photovoltaic output;
s3: aiming at the low-frequency sequence and the high-frequency sequence and combining the sensitivity, respectively establishing a prediction model of the high-frequency sequence and a prediction model of the low-frequency sequence which take meteorological factors into consideration;
s4: obtaining a high-frequency sequence predicted value according to the high-frequency sequence prediction model, and obtaining a low-frequency sequence predicted value according to the low-frequency sequence prediction model; and obtaining a photovoltaic output predicted value according to the low-frequency sequence predicted value and the high-frequency sequence predicted value.
Further, step S1 is specifically:
s11: converting the photovoltaic output time sequence into a matrix form, and decomposing the matrix into the sum of d sub-matrices equivalent to the matrix;
s12: d sub-matrixes obtained by decomposition are usedGrouping to obtain a low-frequency matrix ZlowHigh frequency matrix ZhighAnd a noise matrix ZnoiseAnd combining the low frequency matrix ZlowHigh frequency matrix ZhighAnd a noise matrix ZnoiseRespectively obtaining the low-frequency sequence P after diagonally averaging and reducing the low-frequency sequence P into a reconstructed sequence in the form of an original sequencelThe high frequency sequence PhAnd said noise sequence Pn。
Further, in step S2, the Pearson correlation coefficient method is used to determine that the main meteorological factors affecting the photovoltaic output are specifically:
s21: selecting temperature, irradiation, wind speed and rainfall as meteorological factors;
s22: according to the formulaRespectively calculating Pearson correlation coefficients between photovoltaic output and temperature, irradiation, wind speed or rainfall;
s23: determining main meteorological factors influencing photovoltaic output according to the Pearson correlation coefficient;
wherein,pearson correlation coefficientAbsolute value of (2)Closer to 1 indicates a higher degree of linear correlation between the two variables.
Further, the step S3 of establishing the prediction model considering the high-frequency sequence of the meteorological factors specifically includes:
(1) selecting reference day and reference value of the high-frequency sequence:
taking the day before the day to be predicted as a high-frequency sequence reference day, and taking the photovoltaic output high-frequency sequence of the reference day as a reference value of the high-frequency sequence of the day to be predicted;
(2) taking Pearson correlation coefficients between different meteorological factors and photovoltaic output as weight coefficients of the meteorological factors influencing photovoltaic output change;
(3) according to the sensitivity of meteorological factors to the photovoltaic output change, the temperature difference and the irradiation difference between the day to be predicted and the reference day and according to a formula Phigh＝P'high+α1ΔP1+α2ΔP2High-frequency sequence P for photovoltaic outputhighCorrecting;
wherein, PhighIs the photovoltaic-output high-frequency sequence of the day to be predicted, P'highFor reference day photovoltaic output high frequency series, Δ P1Is the photovoltaic output high-frequency sequence variable quantity, delta P, caused by temperature change2α is the photovoltaic output high-frequency sequence variable quantity caused by irradiation change1Weighting factors for temperature-influenced photovoltaic power high-frequency sequence variation, α2And the weight coefficient is the weight coefficient of irradiation influence on the photovoltaic output high-frequency sequence change.
Further, when the day temperature to be predicted and the reference day temperature are in the same sensitivity range, Δ P1＝St(t-t'); when the day temperature to be predicted and the reference day temperature are in two different sensitivity intervals,
wherein t is the daily temperature value to be predicted, t' is the reference daily temperature value, StIs the sensitivity, S ', of the interval in which the daily temperature is to be predicted'tTo refer to the sensitivity of the interval in which the daily temperature is present,representing the temperature value at the common end point of the two intervals.
Further, in step S4, the low frequency sequence prediction value P is predicted fromlowAnd a high frequency sequence predictor PhighObtaining a photovoltaic output predicted value P ═ Plow+Phigh。
The photovoltaic output is decomposed into different subsequences by a singular spectrum analysis method, and the characteristics of each sequence can be analyzed independently; through correlation analysis and sensitivity analysis, the influence degree of unit variation of different meteorological factors on photovoltaic output can be obtained, so that the photovoltaic output can be predicted more accurately, favorable data reference is provided for scheduling decision-making personnel, and impact on a power system caused by photovoltaic output access is reduced.
Drawings
FIG. 1 is a diagram of SSA-MF method concept provided by an embodiment of the present invention;
FIG. 2 is a flow diagram of a singular spectral analysis technique provided by an embodiment of the present invention;
FIG. 3 is a photovoltaic output breakdown sequence of 4 months 2014; (a) historical photovoltaic output is obtained; (b) a historical photovoltaic output low-frequency sequence is obtained; (c) a historical photovoltaic output high-frequency sequence is obtained; (d) is a historical photovoltaic output noise sequence;
FIG. 4 is a graph of actual and predicted data for 12 solar-volt output at month 5 (SSA-MF); (a) the low-frequency sequence predicted value is 5 months and 12 solar photovoltaic output; (b) the predicted value of the high-frequency sequence of the 12 solar photovoltaic output in 5 months; (c) the predicted value is 12 solar photovoltaic output in 5 months.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The conventional prediction model and method are difficult to adapt to intermittent changes of photovoltaic output, the decomposed subsequences are not deeply mined and analyzed, and the meteorological factors related to the photovoltaic output are processed more complicatedly, so that the realization is difficult or the prediction accuracy is not high. The invention aims to overcome the following two limitations of the conventional photovoltaic output prediction method: (1) when meteorological factors are not considered, the prediction precision is not high, and an ideal prediction effect is difficult to achieve; (2) when meteorological factors are considered, the processing method and the processing process of the meteorological factors are complex, and the realization difficulty in practical application is high. Therefore, the invention provides a short-term photovoltaic decomposition prediction method considering meteorological factor change; according to the method, the processing process of the influence of the complex meteorological factors on the photovoltaic output can be simplified through correlation analysis and sensitivity analysis, historical photovoltaic output data can be decomposed, and the decomposed subsequences are analyzed and predicted respectively, so that the prediction result has high precision, and the safe and stable operation of a power system after photovoltaic access is guaranteed.
Fig. 1 shows an SSA-MF method concept diagram, and the singular spectrum analysis method for short-term photovoltaic output prediction considering meteorological factors provided by the invention includes the following steps:
(1) and decomposing the photovoltaic output time sequence by a singular spectrum analysis technology to obtain a low-frequency sequence, a high-frequency sequence and a noise sequence.
The Singular Value Decomposition (SVD) method is a technique for time series analysis and prediction, and performs Singular Value Decomposition (SVD) on a signal to obtain a trend characteristic, a period characteristic, a white noise characteristic, and the like of an original signal, thereby facilitating analysis of the original signal. FIG. 2 is a flow chart of the singular spectrum analysis technique, which mainly includes two complementary stages of decomposition and reconstruction. The SSA decomposition is to convert the original time sequence P into a matrix form and then use SVD decomposition to obtain d sub-matrices equivalent to the original matrix. And the SSA reconstruction is to firstly group the sub-matrixes after the SVD decomposition to obtain low-frequency, high-frequency and noise matrixes, and then average the diagonal angles of the low-frequency, high-frequency and noise matrixes to obtain low-frequency, high-frequency and noise sequences.
The step (1) is specifically as follows:
(11): and (3) SSA decomposition: the basic idea of the SSA decomposition method is to transform the original time series into a matrix form and then decompose the matrix into a sum of a plurality of sub-matrices equivalent thereto. The SSA Decomposition is mainly divided into two steps of Embedding operation (Embedding) and SVD Decomposition (Singular Value Decomposition).
(111): embedding operation: the embedding operation is to change the original one-dimensional time sequence photovoltaic output P with the length of N (N > 2) to (P)1,P2,L,PN) Converting into a multi-dimensional time sequence photovoltaic output matrix Z ═ Z1,Z2,L,ZK]I.e. P ═ P (P)1,P2,L,PN)→Z＝[Z1,Z2,L,ZK]… … (1); wherein Z isi(i ═ 1,2, L, K) is a column of matrix Z, Zi＝(Pi,Pi+1,L,Pi+L-1)T∈RLThe total dimension is L, L is an embedding dimension (L is more than or equal to 2 and less than or equal to N), and K is N-L + 1. In general, L is preferably selected to not exceed 1/3 for the entire sequence length. The Matrix Z is called a track Matrix (track Matrix), i.e.:and at this point, the conversion from the one-dimensional photovoltaic output sequence to the multi-dimensional photovoltaic output matrix is completed.
(112): SVD decomposition: SVD decomposition decomposes the trajectory matrix Z of equation (2) into d sub-matricesd is the rank of matrix Z and makes the sum of d sub-matrices equal to matrix Z, i.e.: according to equation (4) to calculateIn the formula, λ1,λ2,L,λL(λ1≥λ2≥L≥λLNot less than 0) is S ═ ZZTCharacteristic value of (U)1,U2,L,ULIs a feature vector orthonormal system.
VjObtained from the formula (5)Wherein, UjAnd VjRespectively representing the left and right eigenvectors of the trajectory matrix Z,set as singular values of the trajectory matrix ZIs a singular spectrum of the matrix Z,together forming a characteristic ringAnd at this point, completing SVD decomposition to obtain d sub-matrixes corresponding to the singular spectrum of the matrix Z.
(12): SSA reconstruction: SSA reconstruction is d sub-matrixes obtained by decomposing SVD firstlyGrouping to obtain low frequency/high frequency/noise matrixes respectively marked as Zlow、ZhighAnd Znoise(ii) a Then respectively diagonally averaging and reducing the low-frequency/high-frequency/noise matrix into a reconstructed sequence in the form of an original sequence, namely a low-frequency sequence PlHigh frequency sequence PhAnd a noise sequence Pn. SSA reconstruction mainly includes two steps of Grouping (Grouping) and Diagonal Averaging (diagonalaveraging).
(121) Grouping according to the contribution η of the first r (r ≧ 0) singular values to the sum of the singular values of the matrix Z, and the case of a large jump in the singular values.0And λ "0. When lambda isjWhen λ' is not less than λ, λjCorresponding matrix ZjRegarding the low-frequency submatrix; lambda' > lambdajλ > λjCorresponding matrix ZjRegarding the high-frequency sub-matrix; lambda [ alpha ]jWhen the lambda is less than or equal to lambda', the lambdajCorresponding matrix ZjConsidered as a noise submatrix. The specific grouping is contingent on the actual situation. The d sub-matrices obtained in equation (3) can be divided into low/high frequency/noise matrices as shown in equation (6).The calculation formula of the contribution η is shown in equation (7):
(122): diagonal averaging: further converting the low/high frequency/noise matrix determined in the step S121 into a low/high frequency/noise sequence of length N, hereinafter referred to as high frequency matrix ZhighThe description is given for the sake of example.
Suppose ZhighIs a matrix of a x b, ZijIs ZhighAny one of the elements of (1), note a*＝min(a,b)，b*Max (a, b), N ═ a + b-1, and when a < b,if not, then,the reconstruction sequence RC corresponding to the above-mentioned grouping matrix is (RC)1,rc2,…,rcN) Can be obtained by the following formula:
from the formula (8), the low frequency sequence P can be obtainedhSimilarly, the low frequency sequence P can be obtainedlAnd a noise sequence Pn。
(2) And determining main meteorological factors influencing photovoltaic output by adopting a Pearson correlation coefficient method.
The specific implementation method comprises the following steps: and carrying out correlation analysis on the photovoltaic output time sequence and different meteorological factors. The invention adopts Pearson correlation coefficient method as shown in formula (9). 1) Selecting meteorological factors, and researching different meteorological factors such as temperature, irradiation, wind speed, rainfall and the like; 2) respectively calculating Pearson correlation coefficients between photovoltaic output and meteorological factors such as temperature, irradiation, wind speed and rainfall according to the formula (9), wherein the calculation results are shown in Table 1; 3) determining main meteorological factors influencing photovoltaic output according to the magnitude of the correlation coefficient in the step 2).
Wherein,correlation coefficientAbsolute value of (2)Closer to 1 indicates a higher degree of linear correlation between the two variables.
TABLE 1 correlation coefficient of meteorological data and luminous output data
(3) And (3) analyzing the sensitivity of each main weather to the photovoltaic output according to the main weather factors determined in the step (2).
And analyzing the sensitivity of the main meteorological factors to the photovoltaic output change, wherein the sensitivity of the photovoltaic output to the meteorological factors refers to the variable quantity of the photovoltaic output under the change of unit meteorological factors.
And analyzing the sensitivity of the main meteorological factors to the photovoltaic output change, wherein the sensitivity of the photovoltaic output to the meteorological factors refers to the variable quantity of the photovoltaic output under the change of unit meteorological factors. The specific operation process is as follows: the sensitivities of the photovoltaic output to the temperature and the irradiation intensity are analyzed according to the main meteorological factors (the temperature and the irradiation intensity are taken as examples in the following) determined in the step 2, and the solving results are shown in the tables 2 and 3.
TABLE 2 sensitivity of photovoltaic output to air temperature variation
| Temperature interval/. degree.C | sensitivity/kW | Temperature interval/. degree.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 | - | - |
TABLE 3 sensitivity of photovoltaic contribution to variation of irradiation intensity
(4) And respectively establishing a prediction model considering meteorological factors aiming at the low-frequency sequence and the high-frequency sequence.
In view of the same steps for modeling and predicting the low/high frequency sequences, the following description will take the prediction process of the high frequency sequences as an example.
The method comprises the following steps:
a. and selecting a reference day and a reference value of the high-frequency sequence, taking the day before the day to be predicted as a high-frequency sequence reference day, and taking the photovoltaic output high-frequency sequence of the reference day as the reference value of the high-frequency sequence of the day to be predicted.
b. And taking Pearson correlation coefficients between different meteorological factors and photovoltaic output as weight coefficients of the meteorological factors influencing photovoltaic output change.
c. According to the sensitivity of meteorological factors to the change of photovoltaic output, the temperature difference and the irradiation difference between the day to be predicted and the reference day, the photovoltaic output high-frequency sequence P is subjected to the equation (10)highAnd (6) correcting.
Phigh＝P'high+α1ΔP1+α2ΔP2… … (10); in the formula: phigh,P'high,ΔP1And Δ P2α are photovoltaic output high-frequency sequence of the day to be predicted, photovoltaic output high-frequency sequence of the reference day, photovoltaic output high-frequency sequence variation caused by temperature change and photovoltaic output high-frequency sequence variation caused by irradiation change1And α2The weight coefficients of the high-frequency sequence change of the photovoltaic output influenced by the temperature and the irradiation are respectively.
As can be seen from equation (10), the amount of correction includes Δ P1And Δ P2The invention adopts the following method to determine the values of the two, and the value is expressed by delta P1The description is given for the sake of example.
(a) When the day temperature to be predicted and the reference day temperature are in the same sensitivity interval:
ΔP1＝St(t-t')……(11)；
(b) when the daily temperature to be predicted and the reference daily temperature are in two different sensitivity intervals, such as two adjacent intervals as an example, thenIn the formula: t and t' respectively represent the daily temperature to be predicted and the reference daily temperature value; stAnd S'tRespectively representing the sensitivity of the interval in which the daily temperature to be predicted and the reference daily temperature are respectively located;representing the temperature value at the common end point of the two intervals.
(5) And (5) superposing the low-frequency sequence predicted value and the high-frequency sequence predicted value obtained in the step (4) according to an equation (13) to obtain a photovoltaic output predicted value.
P＝Plow+Phigh… … (13); wherein, Plow、PhighRespectively, low frequency sequence and high frequency sequence prediction values.
The embodiments of the present invention will be described in further detail with reference to the accompanying drawings.
The invention provides a method suitable for predicting intermittent renewable energy sources such as wind power, photovoltaic and the like; the singular spectrum analysis method for short-term photovoltaic output prediction considering meteorological factors is specifically provided. The method can decompose photovoltaic output into different subsequences by Singular Spectrum Analysis (SSA) and can analyze the characteristics of each sequence independently; through correlation analysis and sensitivity analysis, the influence degree of unit variation of different meteorological factors on photovoltaic output can be obtained, so that the photovoltaic output can be predicted more accurately, favorable data reference is provided for scheduling decision-making personnel, and impact on a power system caused by photovoltaic output access is reduced.
Step 1 is implemented: and acquiring data such as photovoltaic output, temperature, irradiation, wind speed and rainfall from 1 month and 5 days in 2013 to 30 months in 2014 and temperature and irradiation in 5 months in 2014. The photovoltaic output of 1 day in the future is predicted by adopting data of the past 1 year, data of 1 month and 1 day in 2013 to 30 months in 5 months in 2014 are used as prediction samples, and data of 5 months in 2014 are used as test samples.
Step 2 is implemented: the photovoltaic output time sequence is decomposed by utilizing the SSA technology, and a photovoltaic output low-frequency sequence, a photovoltaic output high-frequency sequence and a photovoltaic output noise sequence are obtained and are shown in figure 1. Because the noise sequence is reconstructed by the submatrix with a small ratio of the characteristic values, the influence on the original data is small, and the noise sequence is removed. Therefore, the invention focuses on predicting low-frequency sequences and high-frequency sequences.
Step 3 is implemented: the main meteorological factors influencing the photovoltaic output change are determined by utilizing a Pearson correlation coefficient method, and the temperature and the irradiation intensity are calculated and known as the main meteorological factors influencing the photovoltaic output according to the table 1.
And (4) implementing the step: the sensitivities of the main meteorological factors, i.e. temperature and irradiation intensity, to the change in photovoltaic output were calculated as shown in tables 2 and 3.
And 5, implementation step: and respectively carrying out modeling prediction on the low-frequency sequence and the high-frequency sequence according to the sensitivity calculation result. The prediction results of the low-frequency sequence and the high-frequency sequence are obtained as shown in fig. 4(a) and 4 (b).
And 6, implementation step: and (4) superposing the low-frequency sequence and the high-frequency sequence obtained in the step (5) to obtain a prediction result of the photovoltaic output of the graph (4 (c).
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (6)
1. A short-term photovoltaic decomposition prediction method considering meteorological factor change is characterized by comprising the following steps:
s1: decomposing the photovoltaic output time sequence by a singular spectrum analysis method to obtain a low-frequency sequence, a high-frequency sequence and a noise sequence, and simultaneously removing the noise sequence;
s2: determining main meteorological factors influencing photovoltaic output by adopting a Pearson correlation coefficient method, and analyzing the sensitivity of the main meteorological factors to the photovoltaic output;
s3: aiming at the low-frequency sequence and the high-frequency sequence and combining the sensitivity, respectively establishing a prediction model of the high-frequency sequence and a prediction model of the low-frequency sequence which take meteorological factors into consideration;
s4: obtaining a high-frequency sequence predicted value according to the high-frequency sequence prediction model, and obtaining a low-frequency sequence predicted value according to the low-frequency sequence prediction model; and obtaining a photovoltaic output predicted value according to the low-frequency sequence predicted value and the high-frequency sequence predicted value.
2. The short-term photovoltaic decomposition prediction method according to claim 1, wherein the step S1 specifically comprises:
s11: converting the photovoltaic output time sequence into a matrix form, and decomposing the matrix into the sum of d sub-matrices equivalent to the matrix;
s12: d sub-matrixes obtained by decomposition are usedGrouping to obtain a low-frequency matrix ZlowHigh frequency matrix ZhighAnd a noise matrix ZnoiseAnd combining the low frequency matrix ZlowHigh frequency matrix ZhighAnd a noise matrix ZnoiseRespectively obtaining the low-frequency sequence P after diagonally averaging and reducing the low-frequency sequence P into a reconstructed sequence in the form of an original sequencelThe high frequency sequence PhAnd said noise sequence Pn。
3. The short-term photovoltaic decomposition prediction method according to claim 1 or 2, wherein the step S2 of determining the main meteorological factors affecting the photovoltaic output by using Pearson correlation coefficient method specifically includes:
s21: selecting temperature, irradiation, wind speed and rainfall as meteorological factors;
s22: according to the formulaRespectively calculating Pearson correlation coefficients between photovoltaic output and temperature, irradiation, wind speed or rainfall;
s23: determining main meteorological factors influencing photovoltaic output according to the Pearson correlation coefficient;
wherein,pearson correlation coefficientAbsolute value of (2)Closer to 1 indicates a higher degree of linear correlation between the two variables.
4. The short-term photovoltaic decomposition prediction method according to any one of claims 1 to 3, wherein the step S3 of establishing a prediction model considering the weather factor high-frequency sequence specifically comprises:
(1) selecting reference day and reference value of the high-frequency sequence:
taking the day before the day to be predicted as a high-frequency sequence reference day, and taking the photovoltaic output high-frequency sequence of the reference day as a reference value of the high-frequency sequence of the day to be predicted;
(2) taking Pearson correlation coefficients between different meteorological factors and photovoltaic output as weight coefficients of the meteorological factors influencing photovoltaic output change;
(3) according to the sensitivity of meteorological factors to the photovoltaic output change, the temperature difference and the irradiation difference between the day to be predicted and the reference day and according to a formula Phigh＝Ph'igh+α1ΔP1+α2ΔP2High-frequency sequence P for photovoltaic outputhighCorrecting;
wherein, PhighPhotovoltaic output high frequency sequence, P, for the day to be predictedh'ighFor reference day photovoltaic output high frequency series, Δ P1Is the photovoltaic output high-frequency sequence variable quantity, delta P, caused by temperature change2α is the photovoltaic output high-frequency sequence variable quantity caused by irradiation change1High photovoltaic output for temperature influenceWeight coefficient of frequency sequence variation, α2And the weight coefficient is the weight coefficient of irradiation influence on the photovoltaic output high-frequency sequence change.
5. The short-term photovoltaic decomposition prediction method of claim 4, wherein Δ P is obtained when the temperature of the day to be predicted and the temperature of the reference day are in the same sensitivity range1＝St(t-t'); when the day temperature to be predicted and the reference day temperature are in two different sensitivity intervals,
wherein t is the daily temperature value to be predicted, t' is the reference daily temperature value, StSensitivity of the interval in which the daily temperature is to be predicted, St' is the sensitivity of the interval in which the reference daily temperature is located,representing the temperature value at the common end point of the two intervals.
6. The short-term photovoltaic decomposition prediction method according to any one of claims 1 to 5, wherein in step S4, the value P is predicted according to the low-frequency sequencelowAnd a high frequency sequence predictor PhighObtaining a photovoltaic output predicted 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 |
|---|---|---|
| CN107563565B (en) | A kind of short-term photovoltaic decomposition prediction technique considering Meteorology Factor Change | |
| Dong et al. | Hourly energy consumption prediction of an office building based on ensemble learning and energy consumption pattern classification | |
| CN110070226B (en) | Photovoltaic power prediction method and system based on convolutional neural network and meta-learning | |
| CN109524993B (en) | Wind power photovoltaic typical cycle output scene generation method for medium-long term optimization scheduling | |
| CN111080032A (en) | Load prediction method based on Transformer structure | |
| CN106779223B (en) | Photovoltaic system power generation real-time prediction method and device | |
| CN105184391A (en) | Method for predicting wind speed and power of wind farm based on wavelet decomposition and support vector machine | |
| CN108717579B (en) | Short-term wind power interval prediction method | |
| CN102479339A (en) | Method and system for forecasting short-term wind speed of wind farm based on hybrid neural network | |
| CN110070228B (en) | BP neural network wind speed prediction method for neuron branch evolution | |
| CN112668611B (en) | Kmeans and CEEMD-PE-LSTM-based short-term photovoltaic power generation power prediction method | |
| CN104601104B (en) | Ultra-short term photovoltaic prediction method with ARMA correction based on LSSVM | |
| CN108537359B (en) | Photovoltaic power generation power prediction based on wavelet analysis and BP neural network | |
| CN110648017A (en) | Short-term impact load prediction method based on two-layer decomposition technology | |
| CN106204332B (en) | A kind of photovoltaic plant efficiency decaying prediction technique | |
| CN114091766B (en) | CEEMDAN-LSTM-based space load prediction method | |
| CN105160423A (en) | Photovoltaic power generation prediction method based on Markov residual error correction | |
| CN115860797B (en) | Electric quantity demand prediction method suitable for new electricity price reform situation | |
| CN116307291B (en) | Distributed photovoltaic power generation prediction method and prediction terminal based on wavelet decomposition | |
| CN112561178A (en) | Distributed photovoltaic output prediction method considering microclimate factors | |
| CN115759389A (en) | Day-ahead photovoltaic power prediction method based on weather type similar day combination strategy | |
| CN113984198B (en) | Shortwave radiation prediction method and system based on convolutional neural network | |
| CN117251672A (en) | Regional industry multi-dimensional energy carbon emission calculation method and device based on electric power data | |
| CN112215410A (en) | Power load prediction method based on improved deep learning | |
| CN117076738A (en) | Medium-long term weather prediction method, system and medium based on variable grid model |
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 |