CN114862023A - Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast - Google Patents
Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast Download PDFInfo
- Publication number
- CN114862023A CN114862023A CN202210509744.0A CN202210509744A CN114862023A CN 114862023 A CN114862023 A CN 114862023A CN 202210509744 A CN202210509744 A CN 202210509744A CN 114862023 A CN114862023 A CN 114862023A
- Authority
- CN
- China
- Prior art keywords
- photovoltaic power
- prediction
- power station
- centralized
- distributed photovoltaic
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000006870 function Effects 0.000 claims description 16
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- 238000010606 normalization Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 230000005855 radiation Effects 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 9
- 241000039077 Copula Species 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 5
- 238000003064 k means clustering Methods 0.000 claims description 4
- 230000035772 mutation Effects 0.000 claims description 4
- 230000007787 long-term memory Effects 0.000 claims description 2
- 238000010248 power generation Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 208000006440 Open Bite Diseases 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000009434 installation Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000000053 physical method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- 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
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Mathematical Optimization (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
Abstract
The invention discloses a distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting, which comprises the following steps: acquiring original data of a centralized photovoltaic power station in a certain time domain at each moment, and constructing a characteristic data set; inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment; and according to the prediction result, combining with a spatial correlation model to obtain the predicted output power of the distributed photovoltaic power station at the next moment. According to the method, the distribution characteristics of the photovoltaic power in time and space are reflected by calculating the non-coverage coefficient, a spatial correlation model between the centralized photovoltaic power station and the distributed photovoltaic power station is constructed, the centralized photovoltaic power prediction model is combined, the accurate prediction of the distributed photovoltaic power is realized, and the reliability of the prediction is improved.
Description
Technical Field
The invention belongs to the technical field of power control, and particularly relates to a distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecasting.
Background
With the progress of science and technology, photovoltaic power generation has been vigorously developed as a main new energy utilization mode. The photovoltaic power generation grid connection mainly comprises a centralized type and a distributed type, and with the centralized type large amount of grid connection, the defects are continuously shown, for example, the large-scale centralized type photovoltaic power generation cannot be locally absorbed, a large amount of light abandoning phenomena are generated, and resource waste is caused. The distributed photovoltaic system has the advantages of flexible installation, good environmental protection benefit, capability of relieving the shortage of power consumption in partial areas to a certain extent and the like, and is regarded as an optimal development mode of photovoltaic power generation. Therefore, in the face of the fact that the distributed photovoltaic with a high proportion is connected into a power grid, accurate prediction of the power generation power of the distributed photovoltaic has great significance for operation scheduling and safe and stable operation of the power grid.
In the prior art, most of prediction technologies for photovoltaic power aim at centralized photovoltaics with complete data, and mainly include a physical method, a statistical method, a combination method and the like. Compared with a centralized photovoltaic station, the distributed photovoltaic power station has the advantages that the installation places of the distributed photovoltaic power station are dispersed, the installation scale is unequal, the randomness and the intermittence of solar radiation, the diversity of geographic positions and the diversity of meteorological information make the prediction of the power generation power of the distributed photovoltaic power station more difficult, and the prediction accuracy is low.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a distributed photovoltaic power prediction method and a distributed photovoltaic power prediction system based on four-dimensional point-by-point meteorological forecasting.
To achieve the above object, one or more embodiments of the present invention provide the following technical solutions:
the invention provides a distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting.
A distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting comprises the following steps:
acquiring original data of a centralized photovoltaic power station in a certain time domain at each moment, and constructing a characteristic data set;
inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and according to the prediction result, combining with a spatial correlation model to obtain the predicted output power of the distributed photovoltaic power station at the next moment.
According to a further technical scheme, the characteristic data set comprises solar radiation, diffused radiation, humidity, temperature and active power.
In the further technical scheme, in the training process of the centralized photovoltaic power prediction model, a Bayesian optimization-based long-time memory neural network and a Bayesian optimization-based convolutional neural network are utilized to construct the centralized photovoltaic power prediction model.
The further technical scheme is that the construction of the spatial correlation model of the centralized photovoltaic power station and the distributed photovoltaic power station comprises the following steps:
acquiring historical data of a centralized photovoltaic power station and a distributed photovoltaic power station;
according to historical data, adopting non-coverage coefficient normalization to perform data processing;
clustering the data processing results to finish the classification of the weather types;
and establishing a spatial correlation model by using a Copula function according to the clustering result.
According to the further technical scheme, historical data of the centralized photovoltaic power station and the distributed photovoltaic power station in the same area and the same time period are obtained, and the historical data comprise the output power of the photovoltaic power station at each moment in the time period.
In a further technical scheme, the data processing is performed by adopting non-coverage coefficient normalization, and a specific formula is as follows:
in the formula u ij The uncovered coefficient is the j time of the ith day in the test period; p ij Is the output power at the jth time of day i; w is a j The output power at the j time under clear sky; and m is the number of samples.
According to the further technical scheme, the clear coefficient and the mutation coefficient of each day are obtained based on the uncovered coefficient, the numerical values are normalized and then serve as clustering indexes, and the weather types are classified by using a K-means clustering method.
The invention provides a distributed photovoltaic power prediction system based on four-dimensional point-by-point meteorological forecasting.
A distributed photovoltaic power prediction system based on four-dimensional point-by-point meteorological forecasting comprises:
the data acquisition module is used for acquiring original data of the centralized photovoltaic power station in a certain time domain at each moment and constructing a characteristic data set;
the centralized photovoltaic power prediction module is used for inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and the distributed photovoltaic power prediction module is used for obtaining the predicted output power of the distributed photovoltaic power station at the next moment according to the prediction result by combining the spatial correlation model.
A third aspect of the invention provides an electronic device.
An electronic device comprising a memory and a processor and computer instructions stored on the memory and executed on the processor, the computer instructions, when executed by the processor, performing the steps of a four-dimensional point-by-point weather forecast based distributed photovoltaic power prediction method as described above.
A fourth aspect of the invention provides a computer-readable storage medium.
A computer readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to execute a method of distributed photovoltaic power prediction based on four-dimensional point-by-point meteorological forecasting as described above.
The above one or more technical solutions have the following beneficial effects:
(1) according to the distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecasting, the distribution characteristics of photovoltaic power in time and space are reflected by calculating the non-coverage coefficient, a spatial correlation model between a centralized photovoltaic power station and a distributed photovoltaic power station is constructed, the centralized photovoltaic power prediction model is combined, accurate prediction of the distributed photovoltaic power is achieved, and the reliability of prediction is improved.
(2) In order to describe the deviation of power output in a time domain and a space domain, the power data analysis is carried out by adopting an uncovered coefficient, and the connection characteristics of a centralized power station and a distributed power station are represented by establishing a space correlation model.
(3) According to the distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecasting, a four-dimensional space-time mixing model is constructed through weather classification, time and space correlation between a centralized photovoltaic power station and a distributed photovoltaic power station and meteorological information, and accurate prediction of distributed photovoltaic power is achieved. The distributed photovoltaic power generation prediction model provided by the disclosure has a better prediction effect in sunny weather, and is superior to the existing prediction method.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
Fig. 1 is a flowchart of a distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting according to an embodiment of the present invention;
FIG. 2 is a structural framework of a distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting according to an embodiment of the present invention;
FIG. 3 is a comparison graph of a predicted distributed photovoltaic power and an actual photovoltaic power in a case where a model is constructed by using different neural networks according to the prediction method of the first embodiment of the present invention;
fig. 4 is a comparison graph of the prediction distributed photovoltaic power and the actual photovoltaic power in different weather according to the prediction method of the embodiment of the present invention.
Detailed Description
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Example one
Aiming at the problems of low prediction accuracy and reliability of distributed photovoltaic power in the prior art, the invention provides a distributed photovoltaic power prediction method and system based on machine learning.
The embodiment discloses a distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting, as shown in fig. 1, the method includes:
acquiring original data of a centralized photovoltaic power station in a certain time domain at each moment, and constructing a characteristic data set;
inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and according to the prediction result, combining with a spatial correlation model to obtain the predicted output power of the distributed photovoltaic power station at the next moment.
Figure 2 represents a research framework for the protocol described in the present disclosure. The method comprises the steps of constructing a centralized photovoltaic power prediction model by utilizing a long-time memory neural network and a convolutional neural network based on Bayesian Optimization (BO).
In order to solve the problem of disappearance of the RNN gradient of the recurrent neural network and realize long-term learning, the long-term and short-term memory neural network LSTM adds a new hidden state with an enhanced nonlinear mechanism on the basis of the hidden state of the recurrent neural network RNN, and the hidden state is called a cell state. LSTM uses simple gate functions to control the modification, updating or resetting of states, including input gates, output gates and forgetting gates.
A typical Bidirectional LSTM (BilsTM) consists of a forward LSTM and a backward LSTM, which are connected together to produce an output. The backward LSTM is equivalent to allowing the model to learn information about the future, which helps the model to better learn the way the data is represented.
While CNN-LSTM is a fusion of CNN and LSTM. First, the CNN part processes data to extract temporal features of the data, and then inputs a learned feature map (feature map) to the LSTM part to extract spatial features of the data.
Compared with the CNN-LSTM model, the LSTM-CNN model firstly extracts the time characteristics of the data, then extracts the spatial characteristics of the data, utilizes the LSTM-CNN model to predict the photovoltaic power, and determines that the prediction effect of the LSTM-CNN model is superior to that of the CNN-LSTM model through comparison of prediction results. Therefore, in the present embodiment, the LSTM-CNN model is used to construct the centralized photovoltaic power prediction model.
On the basis of determining the LSTM-CNN model, a Bayesian optimization algorithm is introduced to optimize parameters of a neural network in a training model. The input to Bayesian Optimization (BO) is a range of each parameter, which makes it easier to find the optimal combination of parameters than a discrete value setting. Meanwhile, the candidate points are randomized, so that the model is prevented from spending too much time on bad parameters, and the optimization process is easier and more efficient.
After entering the range of each parameter to be tuned and optimized, initializing the parameter x by using a Bayesian algorithm, solving a target function f, then using a Gaussian process as a structure of a proxy model simulation function, and finally optimizing an acquisition function defined by the proxy model by using the algorithm, and selecting the next sample position in a parameter space. This process is iterated until a stop condition is reached or convergence is reached.
And optimizing the super-parameter configuration of the LSTM-CNN model by using BO, wherein the super-parameters comprise a learning rate, a batch size, an activation function, a loss function, an optimizer, an implicit layer number, a unit number of each layer and the like.
In this embodiment, in the training process of the centralized photovoltaic power prediction model, for example, the power value after 1 hour is predicted by using the feature set data of the previous 24 hours, the raw data of the centralized photovoltaic power station within the previous 24 hours and after 1 hour is collected, where the raw data includes 9 features, which are respectively solar radiation (SR, W/m) 2 ) Diffuse radiation (DR, W/m) 2 ) Humidity (H,%), air pressure (AP, KPa), wind speed (WS, m/s), temperature (T, C), active power (P, MW), hour (H), month (m). The original data samples are divided into training set, verification set and test set according to the proportion of 70%, 20% and 10%. Normalizing the raw data to [0, 1 ]]. To prevent data leakage, the three data sets are divided sequentially, and the model is trained using only the maximum and minimum values of the training set.
During the training process, the features of the raw data were divided into 15 different feature sets, as shown in Table 1 below, for the sake of accuracyThe best input feature set of the model was determined and 15 different feature sets were tested for performance on the 1 hour ahead prediction task. And (3) training the baseline model for 5 times by each feature set, calculating the value of the evaluation index, and finally taking an average value to determine the optimal input feature set. In this embodiment, the solar radiation (SR, W/m) is determined 2 ) Diffuse radiation (DR, W/m) 2 ) Humidity (H,%), temperature (T, C) and active power (P, MW) are optimal feature sets, and a centralized photovoltaic power prediction model is trained.
Table 1 sets of different features
Feature Sets | Included Features |
FS1 | P,SR |
FS2 | P,SR,H |
FS3 | P,SR,H,DR |
FS4 | P,SR,H,RH |
FS5 | P,SR,H,DR,T |
FS6 | P,SR,H,DR,AP |
FS7 | P,SR,H,DR,RH |
FS8 | P,SR,H,T,WS |
FS9 | P,SR,H,T,AP |
FS10 | P,SR,H,DR,T,WS |
FS11 | P,SR,H,DR,T,AP |
FS12 | P,SR,H,DR,T,WS,AP |
FS13 | P,SR,H,DR,WS,AP,RH |
FS14 | P,SR,H,DR,T,WS,AP,RH |
FS15 | P,SR,H,DR,M,T,WS,AP,RH |
In training the above model, RMSE (Root Mean Square Error) is used as a loss function, and RMSLE (Root Mean Square Logarithmic Error) and MAPE (Mean Absolute Percentage Error) are calculated simultaneously for measurement, specifically:
By introducing RMSE, RMSLE and MAPE, the method is used for evaluating the prediction performance, evaluating the deviation between a predicted value and a true value, and circulating the training process to ensure that the trained model is more accurate.
In the training process, data normalization is one of the most common analysis methods for processing multiple data sets containing a large amount of information. For example, a normalized approach is employed to evaluate the relationship between power output data between individual photovoltaic power plants. In addition, normalization is also used for columns with different units, such as temperature, air pressure and solar radiation. The general formula is as follows:
wherein X is a data variable array, X t Is a random variable, Y t Is a normalized result.
The normalization method can adjust data to the same unit level, but cannot consider the natural characteristics of photovoltaic output changing in the time domain and cannot truly reflect the randomness of the photovoltaic output under different time step lengths.
In this embodiment, the building of the spatial correlation model of the centralized photovoltaic power station and the distributed photovoltaic power station specifically includes:
acquiring historical data of a centralized photovoltaic power station and a distributed photovoltaic power station;
according to historical data, adopting non-coverage coefficient normalization to perform data processing;
clustering the data processing results to finish the classification of the weather types;
and establishing a spatial correlation model by using a Copula function according to the clustering result.
Specifically, firstly, historical data of the centralized photovoltaic power station and the distributed photovoltaic power station in the same area and the same time period, including output power of the photovoltaic power station at each moment, are obtained. The power output of the same photovoltaic power station at different moments in clear sky is considered to be similar. Therefore, the present embodiment employs non-coverage normalization to describe normalization between different weather types at the same time point, and the specific formula is as follows:
in the formula u ij In this embodiment, four quarters of a year are selected for the uncovered coefficients at the jth time on the ith day of the test period; p ij The output power at the j time of the ith day; m is the number of samples, representing the number of days in a quarter, i is greater than or equal to 1 and less than or equal to m, and in the embodiment, m is 90 days; w is a j The output power at the j-th time under clear sky, i.e., the maximum output power at the j-th time in the test period (one quarter).
And processing the weather classification of the centralized photovoltaic power station and the distributed photovoltaic power station by adopting a k-means clustering algorithm according to the result. Specifically, the maximum output power difference between the centralized power station and the distributed power station in each quarter without coverage is large, the main difference is the effective output duration, and the uncovered coefficient u obtained based on the calculation is ij And obtaining the clear coefficient Z and the mutation coefficient C of each day, and normalizing the numerical values to be used as clustering indexes. According to the different indexes of the normalized clear coefficient Z and the normalized mutation coefficient C, classifying the weather types by using a K-means clustering method, wherein the concrete formula is as follows:
in the formula u j The absolute value of the difference between the non-occlusion coefficient at the time j in a day, Z is the sum of the non-occlusion coefficients at the time j in the day, and C is the sum of the absolute values of the non-occlusion coefficients at the time points of the day and the previous time point, where N is the total number of the time points of the day, i.e., the sample size of the day, and in this embodiment, N is 288 time points.
According to the fact that Z values and C values of the centralized photovoltaic power stations and the distributed photovoltaic power stations are different in the same time period, a Copula function is used for building a spatial correlation model of the centralized photovoltaic power stations and the distributed photovoltaic power stations, and connection characteristics of the centralized photovoltaic power stations and the distributed photovoltaic power stations are represented through the model.
The Copula function can model the correlation of a random variable distributed by a plurality of known edges. In general, if H (x, y) is a binary joint distribution function of f (x) and g (y) with continuous edge distribution, there is a unique Copula function, such that H (x, y) ═ C (f (x) × g (y)), that is, there is a unique Copula function, which can be expressed as:
C(x,y)=H(F -1 (x),G -1 (y))
in the above formula, F -1 (x) The inverse function of F (x) is shown.
According to the Z value and the C value of the centralized photovoltaic power station and the distributed photovoltaic power station in the same time period, a Copula function is used for establishing a spatial correlation model, and the spatial correlation model is used in combination with a centralized photovoltaic power prediction model to obtain the predicted power of the distributed photovoltaic power station at the next moment.
Namely, on the basis of obtaining the predicted power at the next moment output by the centralized photovoltaic power prediction model, the correlation characteristic between the output powers of the centralized photovoltaic power station and the distributed photovoltaic power station is determined through the spatial correlation model, so that the predicted power at the next moment of the distributed power station is obtained, and the predicted power is high in accuracy and reliable in data.
The scheme provided by the embodiment has a good prediction effect in sunny weather, and the prediction method provided by the disclosure is better than the existing prediction method through experimental result analysis.
Fig. 3 shows a comparison result between predicted distributed photovoltaic power and actual photovoltaic power in the case where a centralized photovoltaic power prediction model is constructed using different neural networks.
Fig. 4 shows a comparison result between the predicted distributed photovoltaic power and the actual photovoltaic power in different weather, and it can be seen that the difference between the predicted distributed photovoltaic power and the actual photovoltaic power is not large by the prediction method of the present disclosure, and the prediction result is more accurate.
Example two
The embodiment discloses a distributed photovoltaic power prediction system based on four-dimensional point-by-point meteorological forecasting, which comprises:
the data acquisition module is used for acquiring original data of the centralized photovoltaic power station in a certain time domain at each moment and constructing a characteristic data set;
the centralized photovoltaic power prediction module is used for inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and the distributed photovoltaic power prediction module is used for obtaining the predicted output power of the distributed photovoltaic power station at the next moment according to the prediction result by combining the spatial correlation model.
EXAMPLE III
The present embodiment provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps in the distributed photovoltaic power prediction method based on four-dimensional point-by-point weather forecast as described above are completed.
Example four
The present embodiment also provides a computer readable storage medium for storing computer instructions, which when executed by a processor, perform the steps of the method for predicting distributed photovoltaic power based on four-dimensional point-by-point meteorological forecasting.
The steps involved in the second to fourth embodiments correspond to the first embodiment of the method, and the detailed description thereof can be found in the relevant description of the first embodiment. The term "computer-readable storage medium" should be taken to include a single medium or multiple media containing one or more sets of instructions; it should also be understood to include any medium that is capable of storing, encoding or carrying a set of instructions for execution by a processor and that cause the processor to perform any of the methods of the present invention.
Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using general purpose computer means, or alternatively, they can be implemented using program code that is executable by computing means, such that they are stored in memory means for execution by the computing means, or they are separately fabricated into individual integrated circuit modules, or multiple modules or steps of them are fabricated into a single integrated circuit module. The present invention is not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (10)
1. A distributed photovoltaic power prediction method based on four-dimensional point-by-point meteorological forecasting is characterized by comprising the following steps:
acquiring original data of a centralized photovoltaic power station in a certain time domain at each moment, and constructing a characteristic data set;
inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and according to the prediction result, combining a spatial correlation model to obtain the predicted output power of the distributed photovoltaic power station at the next moment.
2. The method of claim 1, wherein the characteristic data set comprises solar radiation, diffuse radiation, humidity, temperature and active power.
3. The method as claimed in claim 1, wherein in the training process of the centralized photovoltaic power prediction model, a Bayesian optimization-based long-term memory neural network and a Bayesian optimization-based convolutional neural network are used to construct the centralized photovoltaic power prediction model.
4. The method of claim 1, wherein the building of the spatial correlation model of the centralized photovoltaic power station and the distributed photovoltaic power station comprises:
acquiring historical data of a centralized photovoltaic power station and a distributed photovoltaic power station;
according to historical data, adopting non-coverage coefficient normalization to perform data processing;
clustering the data processing results to finish the classification of the weather types;
and establishing a spatial correlation model by using a Copula function according to the clustering result.
5. The method as claimed in claim 4, wherein historical data of the centralized photovoltaic power station and the distributed photovoltaic power station in the same area and in the same time period, including the output power of the photovoltaic power station at each moment in the time period, are obtained.
6. The method as claimed in claim 4, wherein the data processing is performed by non-coverage coefficient normalization, and the specific formula is as follows:
in the formula u ij The uncovered coefficient is the j time of the ith day in the test period; p ij The output power at the j time of the ith day; w is a j The output power at the j time under clear sky; and m is the number of samples.
7. The method as claimed in claim 4, wherein the distributed photovoltaic power prediction method based on four-dimensional point-by-point weather forecast is characterized in that a clear coefficient and a mutation coefficient of each day are obtained based on an uncovered coefficient, numerical values are normalized to serve as clustering indexes, and a K-means clustering method is used for classifying weather types.
8. A distributed photovoltaic power prediction system based on four-dimensional point-by-point meteorological forecasting is characterized by comprising the following components:
the data acquisition module is used for acquiring original data of the centralized photovoltaic power station in a certain time domain at each moment and constructing a characteristic data set;
the centralized photovoltaic power prediction module is used for inputting the characteristic data set into a trained centralized photovoltaic power prediction model to obtain a prediction result of the output power of the centralized photovoltaic power station at the next moment;
and the distributed photovoltaic power prediction module is used for obtaining the predicted output power of the distributed photovoltaic power station at the next moment according to the prediction result by combining the spatial correlation model.
9. An electronic device, characterized by: comprising a memory and a processor and computer instructions stored on the memory and executed on the processor, which when executed by the processor, perform the steps of a method for four-dimensional point-by-point weather forecast based distributed photovoltaic power prediction according to any of claims 1-7.
10. A computer-readable storage medium characterized by: a plurality of instructions stored therein, the instructions adapted to be loaded by a processor of a terminal device and to perform a method of four-dimensional point-by-point weather forecast based distributed photovoltaic power prediction according to any of claims 1-7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210509744.0A CN114862023A (en) | 2022-05-11 | 2022-05-11 | Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210509744.0A CN114862023A (en) | 2022-05-11 | 2022-05-11 | Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114862023A true CN114862023A (en) | 2022-08-05 |
Family
ID=82638278
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210509744.0A Pending CN114862023A (en) | 2022-05-11 | 2022-05-11 | Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114862023A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115564100A (en) * | 2022-09-22 | 2023-01-03 | 济南大学 | Photovoltaic power prediction method, system and equipment |
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 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN112508279A (en) * | 2020-12-10 | 2021-03-16 | 国网山东省电力公司电力科学研究院 | Regional distributed photovoltaic prediction method and system based on spatial correlation |
US20210194424A1 (en) * | 2019-04-25 | 2021-06-24 | Shandong University | Method and system for power prediction of photovoltaic power station based on operating data of grid-connected inverters |
CN114186723A (en) * | 2021-11-26 | 2022-03-15 | 内蒙古工业大学 | Distributed photovoltaic power grid virtual prediction system based on space-time correlation |
-
2022
- 2022-05-11 CN CN202210509744.0A patent/CN114862023A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
US20210194424A1 (en) * | 2019-04-25 | 2021-06-24 | Shandong University | Method and system for power prediction of photovoltaic power station based on operating data of grid-connected inverters |
CN112508279A (en) * | 2020-12-10 | 2021-03-16 | 国网山东省电力公司电力科学研究院 | Regional distributed photovoltaic prediction method and system based on spatial correlation |
CN114186723A (en) * | 2021-11-26 | 2022-03-15 | 内蒙古工业大学 | Distributed photovoltaic power grid virtual prediction system based on space-time correlation |
Non-Patent Citations (2)
Title |
---|
嵇灵;牛东晓;汪鹏;: "基于相似日聚类和贝叶斯神经网络的光伏发电功率预测研究", 中国管理科学, no. 03, 15 March 2015 (2015-03-15), pages 118 - 122 * |
张家安;王琨?;陈建;郭凌旭;黄潇潇;范瑞卿;李志军;: "基于空间相关性的分布式光伏出力预测", 电力建设, no. 03, 1 March 2020 (2020-03-01), pages 47 - 53 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115564100A (en) * | 2022-09-22 | 2023-01-03 | 济南大学 | Photovoltaic power prediction method, system and equipment |
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 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110070226B (en) | Photovoltaic power prediction method and system based on convolutional neural network and meta-learning | |
US11874429B2 (en) | High-temperature disaster forecast method based on directed graph neural network | |
CN111310968A (en) | LSTM neural network circulation hydrological forecasting method based on mutual information | |
CN108985965A (en) | A kind of photovoltaic power interval prediction method of combination neural network and parameter Estimation | |
Fatemi et al. | Parametric methods for probabilistic forecasting of solar irradiance | |
CN114862023A (en) | Distributed photovoltaic power prediction method and system based on four-dimensional point-by-point meteorological forecast | |
CN111275571B (en) | Resident load probability prediction deep learning method considering microclimate and user mode | |
CN110942205A (en) | Short-term photovoltaic power generation power prediction method based on HIMVO-SVM | |
CN115115125A (en) | Photovoltaic power interval probability prediction method based on deep learning fusion model | |
CN115237896A (en) | Data preprocessing method and system for forecasting air quality based on deep learning | |
CN116796168A (en) | CNN-BiLSTM high-altitude multi-factor power transmission line audible noise prediction method based on multi-head attention mechanism | |
CN115759458A (en) | Load prediction method based on comprehensive energy data processing and multi-task deep learning | |
CN114004152B (en) | Multi-wind-field wind speed space-time prediction method based on graph convolution and recurrent neural network | |
Bianchini et al. | Estimation of photovoltaic generation forecasting models using limited information | |
Chen et al. | Probabilistic Prediction of Photovoltaic Power Using Bayesian Neural Network-LSTM Model | |
CN117613883A (en) | Method and device for predicting generated power, computer equipment and storage medium | |
CN115587664B (en) | Solar radiation prediction method and system based on graph convolution long-short-term memory network | |
CN112734073A (en) | Photovoltaic power generation short-term prediction method based on long and short-term memory network | |
CN114462684B (en) | Wind speed multipoint synchronous prediction method coupling numerical weather forecast and measured data | |
CN112446550B (en) | Short-term building load probability density prediction method | |
CN115860165A (en) | Neural network basin rainfall runoff forecasting method and system considering initial loss | |
CN115481788A (en) | Load prediction method and system for phase change energy storage system | |
CN115563848A (en) | Distributed photovoltaic total radiation prediction method and system based on deep learning | |
CN113449642A (en) | PM based on multi-scale feature extraction and deep learning2.5Concentration prediction method and system | |
Xiao et al. | Multi-station low-visibility prediction based on spatio-temporal fusion 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 |