CN115907195A - Photovoltaic power generation power prediction method, system, electronic device and medium - Google Patents
Photovoltaic power generation power prediction method, system, electronic device and medium Download PDFInfo
- Publication number
- CN115907195A CN115907195A CN202211574793.9A CN202211574793A CN115907195A CN 115907195 A CN115907195 A CN 115907195A CN 202211574793 A CN202211574793 A CN 202211574793A CN 115907195 A CN115907195 A CN 115907195A
- Authority
- CN
- China
- Prior art keywords
- power generation
- historical data
- predicted
- photovoltaic power
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010248 power generation Methods 0.000 title claims abstract description 96
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000013528 artificial neural network Methods 0.000 claims abstract description 10
- 230000002068 genetic effect Effects 0.000 claims abstract description 10
- 239000011159 matrix material Substances 0.000 claims description 31
- 230000006870 function Effects 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 3
- 238000013461 design Methods 0.000 description 4
- 230000005855 radiation Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 1
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a photovoltaic power generation power prediction method, a photovoltaic power generation power prediction system, electronic equipment and a medium, and particularly relates to the technical field of power generation prediction of a power system. The method comprises the steps of obtaining a historical data set of a photovoltaic power station in a target area and meteorological data of a moment to be predicted; classifying the historical data set and meteorological data at a moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages; processing the data sets of each power generation stage and the meteorological data of the moment to be predicted by adopting a similar day searching method to obtain a similar day sample set; training the LSTM neural network according to the similar day sample set to obtain an initial prediction model; and optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, wherein the prediction model is used for predicting the photovoltaic power generation power. The method can improve the prediction precision of the photovoltaic power generation power.
Description
Technical Field
The invention relates to the technical field of power generation prediction of power systems, in particular to a photovoltaic power generation power prediction method, a photovoltaic power generation power prediction system, electronic equipment and a medium.
Background
With global energy shortage and environmental issues becoming prominent, the use of renewable energy has received much attention. Photovoltaic power generation is taken as an important form of renewable energy, is one of the most large-scale power generation modes with development prospects for commercialization in the current renewable energy, and is more and more concerned by people.
Disclosure of Invention
The invention aims to provide a photovoltaic power generation power prediction method, a photovoltaic power generation power prediction system, an electronic device and a medium, which can improve the prediction accuracy of photovoltaic power generation power.
In order to achieve the purpose, the invention provides the following scheme:
a photovoltaic power generation power prediction method comprises the following steps:
acquiring a historical data set of a photovoltaic power station in a target area and meteorological data at a moment to be predicted; the historical data set comprises historical data at the same time within set days; the historical data comprises meteorological data and photovoltaic power generation power;
classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages;
processing the data sets of each power generation stage and the meteorological data of the moments to be predicted by adopting a similar day searching method to obtain a similar day sample set;
training the LSTM neural network according to the similar day sample set to obtain an initial prediction model;
and optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, wherein the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
Optionally, before the step of classifying the historical data set and the meteorological data at the time to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages, the method further includes:
and carrying out normalization processing on the historical data set to obtain a normalized historical data set.
Optionally, the classifying the historical data set and the meteorological data at the time to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of multiple power generation stages specifically includes:
calculating a membership matrix under the current iteration number according to Euclidean distances between each clustering center and each historical data in the historical data set under the current iteration number;
calculating each clustering center under the next iteration number according to the membership matrix under the current iteration number;
calculating a membership matrix under the next iteration number according to the Euclidean distance between each clustering center and each historical data in the historical data set under the next iteration number by taking the minimum target function as a target;
judging whether the norm of the difference value between each clustering center under the next iteration number and each clustering center under the current iteration number is smaller than an ending threshold value or not to obtain a first judgment result;
if the first judgment result is yes, clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages;
and if the first judgment result is negative, updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time.
Optionally, the processing the data sets of each power generation stage and the meteorological data of the time to be predicted by using a similar day search method to obtain a similar day sample set specifically includes:
calculating Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted;
and determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data at the time to be predicted.
A photovoltaic power generation power prediction system comprising:
the acquisition module is used for acquiring a historical data set of the photovoltaic power station in the target area and meteorological data of a moment to be predicted; the historical data set comprises historical data at the same time within set days; the historical data comprises meteorological data and photovoltaic power generation power;
the clustering module is used for classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages;
the searching module is used for processing the data sets of all power generation stages and the meteorological data of the moment to be predicted by adopting a similar day searching method to obtain a similar day sample set;
the training module is used for training the LSTM neural network according to the similar day sample set to obtain an initial prediction model;
and the optimization module is used for optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, and the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
Optionally, the photovoltaic power generation power prediction system further includes:
and the normalization module is used for carrying out normalization processing on the historical data set to obtain a normalized historical data set.
Optionally, the clustering module specifically includes:
the membership matrix calculating unit is used for calculating a membership matrix under the current iteration times according to Euclidean distances between each cluster center and each historical data in the historical data set under the current iteration times;
the cluster center calculating unit is used for calculating each cluster center under the next iteration number according to the membership matrix under the current iteration number;
a membership matrix updating unit, configured to calculate a membership matrix for the next iteration number according to euclidean distances between the respective clustering centers and the respective historical data in the historical data set for the next iteration number, with the minimum objective function as a target;
the judging unit is used for judging whether the norm of the difference value between each clustering center under the next iteration frequency and each clustering center under the current iteration frequency is smaller than an ending threshold value or not to obtain a first judging result;
the clustering unit is used for clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages if the first judgment result is yes;
and the updating unit is used for updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time if the first judgment result is negative.
Optionally, the search module specifically includes:
the Chebyshev distance calculating unit is used for calculating the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted;
and the similar day sample set determining unit is used for determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted.
An electronic device, comprising:
a memory for storing a computer program and a processor for executing the computer program to cause the electronic device to execute the photovoltaic power generation power prediction method according to the above.
A computer-readable storage medium storing a computer program which, when executed by a processor, implements a photovoltaic power generation power prediction method as described above.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects: classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages; processing the data sets of each power generation stage and the meteorological data of the moment to be predicted by adopting a similar day searching method to obtain a similar day sample set; training the LSTM neural network according to the similar day sample set to obtain an initial prediction model; and optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, dividing data into a plurality of power generation stages on the basis of fuzzy C-means clustering, introducing similar day search, adding a time variable to perform segmented prediction according to power generation characteristics of different time periods, reducing the dependence on single weather characteristics and improving the prediction precision.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
Fig. 1 is a flowchart of a photovoltaic power generation power prediction method according to an embodiment of the present invention;
fig. 2 is an overall framework diagram of a photovoltaic power generation power prediction method provided by an embodiment of the invention;
FIG. 3 is a flow chart for optimizing an LSTM neural network using a genetic algorithm;
FIG. 4 is a comparison graph of the actual value of the photovoltaic power generation power and the LSTM prediction result;
FIG. 5 is a graph comparing the actual value of the photovoltaic power generation power and the prediction result of GA-LSTM.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
As shown in fig. 1 and fig. 2, an embodiment of the present invention provides a photovoltaic power generation power prediction method, including:
step 101: acquiring a historical data set of a photovoltaic power station in a target area and meteorological data at a moment to be predicted; the historical data set comprises historical data at the same time within set days; the historical data includes meteorological data and photovoltaic power generation power.
Step 102: and classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages, wherein the data sets comprise a plurality of historical data. The power generation stage comprises a power generation peak stage, a power generation gradually increasing stage, a power generation gradually decreasing stage and a low power generation amount stage.
Step 103: and processing the data sets of all the power generation stages and the meteorological data of the moments to be predicted by adopting a similar day searching method to obtain a similar day sample set. The similar day sample set includes a plurality of historical data.
Step 104: and training the LSTM neural network according to the similar day sample set to obtain an initial prediction model.
Step 105: and optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, namely GA-LSTM, wherein the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
In practical application, before the step of classifying the historical data set and the meteorological data at the time to be predicted by using the fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages, the method further comprises the following steps:
and carrying out normalization processing on the historical data set to obtain a normalized historical data set.
In practical application, the classifying the historical data set and the meteorological data at the time to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages specifically comprises the following steps:
and calculating a membership matrix under the current iteration number according to the Euclidean distance between each clustering center and each historical data in the historical data set under the current iteration number.
And calculating each clustering center under the next iteration number according to the membership matrix under the current iteration number.
And calculating a membership matrix under the next iteration number according to the Euclidean distance between each clustering center and each historical data in the historical data set under the next iteration number by taking the minimum target function as a target.
And judging whether the norm of the difference value between each clustering center under the next iteration number and each clustering center under the current iteration number is smaller than an ending threshold value or not, and obtaining a first judgment result.
And if the first judgment result is yes, clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages.
And if the first judgment result is negative, updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time.
In practical application, the processing of the data sets of each power generation stage and the meteorological data of the time to be predicted by using a similar day searching method to obtain a similar day sample set specifically includes:
and calculating the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data at the time to be predicted.
And determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data at the time to be predicted.
The invention adopts meteorological data and historical power generation power data of a certain photovoltaic power station to introduce the photovoltaic power generation power prediction method in detail:
step 1: acquiring meteorological data and historical power generation power of a photovoltaic power station in a certain area at the same moment, identifying bad data and complementing the bad data to obtain original data.
And 2, step: and (3) performing linear change on the original data by adopting min-max standardization, and mapping the size of the data in a 0,1 interval to obtain a training sample.
The standardized processing formula is as follows:
And 3, step 3: and classifying the training samples and meteorological data of a prediction day by using fuzzy C-means clustering, and dividing the data into high, medium and low periods.
In practical application, the specific process of step 3 is as follows: and selecting an ending threshold epsilon and setting the maximum iteration number T.
According to the formula
Calculating the membership p of the ith clustering center and the jth data point in the training sample ij In the formula: z is the number of clusters, z belongs to (1, ∞); d ij Is the Euclidean distance between the ith data cluster center and the jth data point in the training sample; d is a radical of kj Is the Euclidean distance between the kth data clustering center and the jth data point in the training sample, and b is the total number of the clustering centers;
according to the formula
Calculating the ith cluster center B i (i =1,2,.., B) wherein B is i Is a n-dimensional cluster center, x j Is the jth data point in the training sample, i.e. the jth set of meteorological data and historical generated power in the training sample.
Setting a minimum objective function L:
taking the minimum of the objective function as the target according to the formula
Updating the membership degree:
if it is
Ending, otherwise, returning to the step of calculating the clustering center to enter the next iteration until the iteration number reaches T, wherein, the judgment result is based on the judgment result of the judgment result, and the judgment result is based on the judgment result>
Represents the ith cluster center at the number of iterations (k + 1) <' > in the cluster >>
The ith cluster center at the kth iteration number is represented.
And 4, step 4: similar day search is introduced, and the three steps are dividedAnd respectively screening all data matched with the weather type of the predicted day from the data in the three time periods of high, medium and low, and generating a similar day sample set C. The main weather characteristic parameters Q (k), Y (k) (solar radiation intensity and temperature) of the predicted day are compared with the Q of the ith day in the set C i (k),Y i (k) (solar radiation intensity and temperature on day i in set C) and calculating the Chebyshev distance, and the distance value D i Is shown as
In the formula: q (k) and Y (k) are respectively the main weather characteristic parameters of the predicted day, namely the solar radiation intensity and the temperature; rho Q,i Is Q (k) and Q i (k) Correlation coefficients of the numerical sequence; ρ is a unit of a gradient Y,i Is Y (k) and Y i (k) The correlation coefficient of the series of values, cov () is a function of the calculation of the covariance; e (Q (k) is the expectation of a sequence of Q (k) values, E (Q) i (k) Is Q) i (k) Expectation of a sequence of values; e (Y (k) is the expectation of the numerical sequence of Y (k); E (Y) i (k) Is Y) i (k) Expectation of a sequence of values; s (Q (k) is the variance of the Q (k) numerical sequence, S (Q) i (k) Is Q) i (k) A variance of the sequence of values; s (Y (k) is the variance of the numerical sequence of Y (k); S (Y) i (k) Is Y) i (k) The variance of the sequence of values.
Solving the Chebyshev distance between the similar day sample set C and the prediction day for all the days in the similar day sample set C to obtain { D 1 ,D 2 ,…,D j And setting a ratio coefficient eta to satisfy:
will { D 1 ,D 2 ,…,D j The median value is less than or equal to (1 + eta) · min { D } 1 ,D 2 ,…,D j All similar diaries of this sample set C are made to subset B.
And 5: constructing an LSTM model, importing the processed data, namely the similar day sample set, into the LSTM model for training to obtain an initial prediction model:
1) The LSTM memory unit comprises 4 elements of an input gate, an output gate, a forgetting gate and a memory cell which is circularly self-connected;
2) Let the output of the LSTM network at time t be n t ，t＝1,2,3...，n t The calculation is iterated through the following formula:
s t ＝sig(J xs x t +J hs h t-1 +J os o t-1 +b i )
w t ＝sig(J xw x t +J hw h t-1 +J ow o t-1 +b f )
y t ＝sig(J xy x t +J hy h t-1 +J oy o t-1 +b o )
in the formula, s t 、w t 、y t 、o t The outputs of the input gate, the forgetting gate, the memory cell and the output gate at the moment t are respectively; x is the number of t Is input; j. the design is a square xs 、J hs 、J os The weight matrix of the input information, the last moment output and the memory cell to the input gate respectively; j. the design is a square xw 、J hw 、J ow Respectively, the weight moments of the input information, the last moment output and the memory cell to the forgetting gateArraying; b i 、b f 、b o 、b c The offset of the input gate, the output gate, the forgetting gate and the memory cell are respectively; j. the design is a square hy 、J xy 、J oy The weight matrixes from the output and input information at the last moment and the memory cells to the input nodes respectively; j. the design is a square xo 、J ho The weight matrixes are respectively input information and weight matrixes output to an output gate at the last moment;
represents the change of the tanh function, o t-1 Output of the gate for time t-1, h t-1 Represents the cell state at time t-1, n t The output of the LSTM network at time t is shown, and sig represents the sigmoid activation function.
3) And importing the preprocessed data with the length of n into the LSTM, and training the LSTM model by taking meteorological data as input and photovoltaic power generation power as output to obtain an initial prediction model.
Step 6: as shown in FIG. 3, the initial prediction model is optimized by genetic algorithm to obtain the LSTM prediction model, namely GA-LSTM.
1) The population is initialized and encoded.
2) The LSTM network is initialized.
3) The fitness of the individual is determined. The fitness function targets the mean square error of the LSTM network,
the fitness function determination process is as follows:
in the formula: e min Is the minimum value in the range of the u (x) function, E max Is the maximum value within the range of the u (x) function, which is the objective function.
4) And (3) judging whether the target value of the fitness function reaches the optimal value, if so, continuing the next step, otherwise, carrying out evolution operation on the individual of the solution and updating the population, and returning to the step 3).
5) And acquiring a fitness target value and an optimal parameter.
6) And outputting the optimal parameters.
7) And obtaining the LSTM network according to the optimal parameters.
8) And predicting by using an LSTM network and outputting a prediction result.
And 7: inputting meteorological data in a similar day sample set into an LSTM prediction model to output predicted photovoltaic power generation power, and comparing the predicted photovoltaic power generation power with actual photovoltaic power generation power, and if the accuracy requirement is met, keeping the LSTM prediction model; otherwise, iterative training will continue until the accuracy requirement is met.
In practical application, step 7 specifically includes: selecting the average absolute percentage error M APE And root mean square error R MSE As an evaluation criterion, the prediction accuracy of the model in the present invention is evaluated when M is APE Not more than 3.5 and R MSE When the error is less than or equal to 1.5, the neural network prediction model is not modified, if one of the error is not satisfied, iteration is continued until the accuracy requirement is satisfied, and a formula used for calculating the error is as follows:
In the formula: n represents the total number of samples in the training sample; a. The i And F i The actual value and the predicted value of the ith sample are predicted respectively.
As shown in FIG. 4 and FIG. 5, M obtained by processing 25 samples of the conventional LSTM algorithm model and the prediction model (optimized LSTM model, namely GA-LSTM) obtained by the present invention APE (when R is MSE Not more than 1.5), and the comparison result is shown in table 1, so that MAPE after optimization is obviously reduced, and the prediction precision is improved.
TABLE 1 different LSTM prediction algorithms M APE (%) comparative results table
In view of the above method, an embodiment of the present invention further provides a photovoltaic power generation power prediction system, including:
the acquisition module is used for acquiring a historical data set of the photovoltaic power station in the target area and meteorological data of a moment to be predicted; the historical data set comprises historical data at the same time within set days; the historical data comprises meteorological data and photovoltaic power generation power;
the clustering module is used for classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages;
the searching module is used for processing the data sets of all power generation stages and the meteorological data of the moment to be predicted by adopting a similar day searching method to obtain a similar day sample set;
the training module is used for training the LSTM neural network according to the similar day sample set to obtain an initial prediction model;
and the optimizing module is used for optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, and the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
In practical application, the photovoltaic power generation power prediction system further includes:
and the normalization module is used for carrying out normalization processing on the historical data set to obtain a normalized historical data set.
In practical application, the clustering module specifically includes:
the membership matrix calculating unit is used for calculating a membership matrix under the current iteration times according to Euclidean distances between each cluster center and each historical data in the historical data set under the current iteration times;
the cluster center calculating unit is used for calculating each cluster center under the next iteration number according to the membership matrix under the current iteration number;
a membership matrix updating unit, configured to calculate a membership matrix for the next iteration number according to an euclidean distance between each cluster center and each historical data in the historical data set for the next iteration number, with a minimum objective function as a target;
the judging unit is used for judging whether the norm of the difference value between each cluster center under the next iteration frequency and each cluster center under the current iteration frequency is smaller than an ending threshold value or not to obtain a first judging result;
the clustering unit is used for clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages if the first judgment result is yes;
and the updating unit is used for updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time if the first judgment result is negative.
In practical application, the search module specifically includes:
the Chebyshev distance calculating unit is used for calculating the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted;
and the similar day sample set determining unit is used for determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted.
An embodiment of the present invention further provides an electronic device, including:
a memory for storing a computer program and a processor for executing the computer program to cause the electronic device to execute the photovoltaic power generation power prediction method according to the above embodiments.
The embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the method for predicting photovoltaic power generation power according to the embodiment is implemented.
According to the method, various algorithms are combined, the objective function is optimized, data are divided into three types of time periods of high, medium and low on the basis of fuzzy C-means clustering, similar day searching is introduced, time variables are added, segmented prediction can be carried out according to power generation characteristics of different time periods, the dependency on single weather characteristics is reduced, a better prediction effect is obtained, and prediction accuracy is improved.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. A photovoltaic power generation power prediction method is characterized by comprising the following steps:
acquiring a historical data set of a photovoltaic power station in a target area and meteorological data at a moment to be predicted; the historical data set comprises historical data at the same time within set days; the historical data comprises meteorological data and photovoltaic power generation power;
classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages;
processing the data sets of each power generation stage and the meteorological data of the moment to be predicted by adopting a similar day searching method to obtain a similar day sample set;
training the LSTM neural network according to the similar day sample set to obtain an initial prediction model;
and optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, wherein the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
2. The method for predicting photovoltaic power generation power according to claim 1, wherein before the step of classifying the historical data set and the meteorological data at the time to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages, the method further comprises:
and carrying out normalization processing on the historical data set to obtain a normalized historical data set.
3. The photovoltaic power generation power prediction method according to claim 1, wherein the step of classifying the historical data set and the meteorological data at the time to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages specifically comprises:
calculating a membership matrix under the current iteration number according to Euclidean distances between each clustering center and each historical data in the historical data set under the current iteration number;
calculating each clustering center under the next iteration number according to the membership matrix under the current iteration number;
calculating a membership matrix under the next iteration number according to the Euclidean distance between each clustering center and each historical data in the historical data set under the next iteration number by taking the minimum target function as a target;
judging whether the norm of the difference value between each cluster center under the next iteration number and each cluster center under the current iteration number is smaller than an end threshold value or not, and obtaining a first judgment result;
if the first judgment result is yes, clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages;
and if the first judgment result is negative, updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time.
4. The photovoltaic power generation power prediction method according to claim 1, wherein the similar day search method is adopted to process the data sets of each power generation stage and the meteorological data of the time to be predicted to obtain a similar day sample set, and specifically includes:
calculating Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted;
and determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data at the time to be predicted.
5. A photovoltaic power generation power prediction system, comprising:
the acquisition module is used for acquiring a historical data set of the photovoltaic power station in the target area and meteorological data of a moment to be predicted; the historical data set comprises historical data at the same moment in a set number of days; the historical data comprises meteorological data and photovoltaic power generation power;
the clustering module is used for classifying the historical data set and the meteorological data at the moment to be predicted by using a fuzzy C-means clustering algorithm to obtain data sets of a plurality of power generation stages;
the searching module is used for processing the data sets of all power generation stages and the meteorological data at the time to be predicted by adopting a similar day searching method to obtain a similar day sample set;
the training module is used for training the LSTM neural network according to the similar day sample set to obtain an initial prediction model;
and the optimizing module is used for optimizing the initial prediction model by adopting a genetic algorithm to obtain a prediction model, and the prediction model is used for predicting the photovoltaic power generation power at the moment to be predicted.
6. The photovoltaic generated power prediction system of claim 5, further comprising:
and the normalization module is used for performing normalization processing on the historical data set to obtain a normalized historical data set.
7. The photovoltaic power generation power prediction system according to claim 5, wherein the clustering module specifically includes:
the membership matrix calculating unit is used for calculating a membership matrix under the current iteration times according to Euclidean distances between each clustering center and each historical data in the historical data set under the current iteration times;
the cluster center calculating unit is used for calculating each cluster center under the next iteration number according to the membership matrix under the current iteration number;
a membership matrix updating unit, configured to calculate a membership matrix for the next iteration number according to an euclidean distance between each cluster center and each historical data in the historical data set for the next iteration number, with a minimum objective function as a target;
the judging unit is used for judging whether the norm of the difference value between each clustering center under the next iteration frequency and each clustering center under the current iteration frequency is smaller than an ending threshold value or not to obtain a first judging result;
the clustering unit is used for clustering the historical data set and the meteorological data at the moment to be predicted according to each clustering center under the next iteration number to obtain data sets of a plurality of power generation stages if the first judgment result is yes;
and the updating unit is used for updating the iteration times and returning to the step of calculating each clustering center under the next iteration time according to the membership matrix under the current iteration time if the first judgment result is negative.
8. The photovoltaic power generation power prediction system according to claim 5, wherein the search module specifically includes:
the Chebyshev distance calculating unit is used for calculating the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted;
and the similar day sample set determining unit is used for determining a similar day sample set according to the Chebyshev distance between each meteorological data in the similar day sample set and the meteorological data of the time to be predicted.
9. An electronic device, comprising:
a memory for storing a computer program and a processor that executes the computer program to cause the electronic device to perform the photovoltaic power generation power prediction method of any one of claims 1 to 4.
10. A computer-readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements a photovoltaic power generation power prediction method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211574793.9A CN115907195A (en) | 2022-12-08 | 2022-12-08 | Photovoltaic power generation power prediction method, system, electronic device and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211574793.9A CN115907195A (en) | 2022-12-08 | 2022-12-08 | Photovoltaic power generation power prediction method, system, electronic device and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115907195A true CN115907195A (en) | 2023-04-04 |
Family
ID=86470934
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211574793.9A Pending CN115907195A (en) | 2022-12-08 | 2022-12-08 | Photovoltaic power generation power prediction method, system, electronic device and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115907195A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116432874A (en) * | 2023-06-14 | 2023-07-14 | 青岛鼎信通讯科技有限公司 | Distributed photovoltaic power prediction method based on characteristic power |
| CN116706907A (en) * | 2023-08-09 | 2023-09-05 | 深圳航天科创泛在电气有限公司 | Photovoltaic power generation prediction method based on fuzzy reasoning and related equipment |
| CN117408394A (en) * | 2023-12-14 | 2024-01-16 | 国网天津市电力公司电力科学研究院 | Carbon emission factor prediction method and device for electric power system and electronic equipment |
| CN117973639A (en) * | 2024-03-29 | 2024-05-03 | 南京师范大学 | Photovoltaic power prediction method based on multiscale similarity days and improved integration algorithm |
-
2022
- 2022-12-08 CN CN202211574793.9A patent/CN115907195A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116432874A (en) * | 2023-06-14 | 2023-07-14 | 青岛鼎信通讯科技有限公司 | Distributed photovoltaic power prediction method based on characteristic power |
| CN116706907A (en) * | 2023-08-09 | 2023-09-05 | 深圳航天科创泛在电气有限公司 | Photovoltaic power generation prediction method based on fuzzy reasoning and related equipment |
| CN116706907B (en) * | 2023-08-09 | 2024-01-23 | 深圳航天科创泛在电气有限公司 | Photovoltaic power generation prediction method based on fuzzy reasoning and related equipment |
| CN117408394A (en) * | 2023-12-14 | 2024-01-16 | 国网天津市电力公司电力科学研究院 | Carbon emission factor prediction method and device for electric power system and electronic equipment |
| CN117408394B (en) * | 2023-12-14 | 2024-05-31 | 国网天津市电力公司电力科学研究院 | Carbon emission factor prediction method and device for electric power system and electronic equipment |
| CN117973639A (en) * | 2024-03-29 | 2024-05-03 | 南京师范大学 | Photovoltaic power prediction method based on multiscale similarity days and improved integration algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115907195A (en) | Photovoltaic power generation power prediction method, system, electronic device and medium | |
| CN108564192B (en) | Short-term photovoltaic power prediction method based on meteorological factor weight similarity day | |
| CN107766990B (en) | Method for predicting power generation power of photovoltaic power station | |
| CN113282122B (en) | Commercial building energy consumption prediction optimization method and system | |
| CN111027772B (en) | Multi-factor short-term load prediction method based on PCA-DBILSTM | |
| CN112564098B (en) | High-proportion photovoltaic power distribution network voltage prediction method based on time convolution neural network | |
| Zhang et al. | Wind speed prediction research considering wind speed ramp and residual distribution | |
| CN114757427B (en) | Autoregressive-corrected LSTM intelligent wind power plant ultra-short-term power prediction method | |
| CN114897129A (en) | Photovoltaic power station short-term power prediction method based on similar daily clustering and Kmeans-GRA-LSTM | |
| CN111915092A (en) | Ultra-short-term wind power prediction method based on long-time and short-time memory neural network | |
| CN117117859A (en) | Photovoltaic power generation power prediction method and system based on neural network | |
| CN114511132A (en) | Photovoltaic output short-term prediction method and prediction system | |
| CN111242355A (en) | Photovoltaic probability prediction method and system based on Bayesian neural network | |
| CN114298140A (en) | Wind power short-term power prediction correction method considering unit classification | |
| CN115115125A (en) | Photovoltaic power interval probability prediction method based on deep learning fusion model | |
| CN113344243B (en) | Wind speed prediction method and system for optimizing ELM (ELM) based on improved Harris eagle algorithm | |
| CN116646927B (en) | Wind power prediction method based on segmented filtering and longitudinal and transverse clustering | |
| CN116826737A (en) | Photovoltaic power prediction method, device, storage medium and equipment | |
| CN117151770A (en) | Attention mechanism-based LSTM carbon price prediction method and system | |
| CN116226689A (en) | Power distribution network typical operation scene generation method based on Gaussian mixture model | |
| CN114881429A (en) | Data-driven transformer area line loss quantification method and system | |
| CN109858667A (en) | It is a kind of based on thunder and lightning weather to the short term clustering method of loading effects | |
| CN117688362A (en) | Photovoltaic power interval prediction method and device based on multivariate data feature enhancement | |
| CN117349687A (en) | Daily load curve clustering method based on variable convolution self-encoder | |
| CN116865255A (en) | Short-term wind power prediction method based on improved entropy weight method and SECEEMD |
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 |