CN116502074A - Model fusion-based photovoltaic power generation power prediction method and system - Google Patents
Model fusion-based photovoltaic power generation power prediction method and system Download PDFInfo
- Publication number
- CN116502074A CN116502074A CN202310769184.7A CN202310769184A CN116502074A CN 116502074 A CN116502074 A CN 116502074A CN 202310769184 A CN202310769184 A CN 202310769184A CN 116502074 A CN116502074 A CN 116502074A
- Authority
- CN
- China
- Prior art keywords
- power generation
- photovoltaic power
- time sequence
- data
- features
- 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 114
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000004927 fusion Effects 0.000 title claims abstract description 32
- 238000000605 extraction Methods 0.000 claims abstract description 21
- 239000000203 mixture Substances 0.000 claims abstract description 15
- 230000001932 seasonal effect Effects 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000007774 longterm Effects 0.000 claims description 7
- 230000015654 memory Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000005096 rolling process Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000004590 computer program Methods 0.000 claims description 2
- 238000005520 cutting process Methods 0.000 claims description 2
- 238000013527 convolutional neural network Methods 0.000 description 12
- 239000004973 liquid crystal related substance Substances 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000009466 transformation Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013501 data transformation Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000006386 memory function Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—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
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- 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
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/04—Power grid distribution networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/06—Power analysis or power optimisation
-
- 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
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- 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
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
- H02J2300/22—The renewable source being solar energy
- H02J2300/24—The renewable source being solar energy of photovoltaic origin
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Power Engineering (AREA)
- Economics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Water Supply & Treatment (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Human Resources & Organizations (AREA)
- Public Health (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides a photovoltaic power generation power prediction method and system based on model fusion, and relates to the technical field of photovoltaic power generation. Firstly, historical data of a region to be predicted is obtained to conduct feature extraction, and the extracted features are clustered by utilizing a Gaussian mixture model to obtain photovoltaic power generation time sequence data. And then, respectively utilizing a GAF-CNN network model and an LSTM-Attention network model to extract long time sequence characteristics and short time sequence characteristics of the photovoltaic power generation time sequence data, adopting a multi-model fusion mode to fuse the short time sequence characteristics and the long time sequence characteristics, and outputting a photovoltaic power generation power prediction result, thereby improving the prediction precision of the distributed photovoltaic power generation power.
Description
Technical Field
The invention relates to the technical field of photovoltaic power generation, in particular to a photovoltaic power generation power prediction method and system based on model fusion.
Background
The rapid decline in reserves of traditional fossil energy has led to a growing market for renewable energy, becoming an important part of economic growth. Solar energy becomes the most ideal alternative energy source by virtue of abundant reserves, cleanness, high efficiency and the like. Photovoltaic power generation has been rapidly developed in recent years as a main form of solar energy utilization. However, because the photovoltaic power generation has intermittence and uncertainty due to the continuous change of solar energy along with day and night, seasons and meteorological factors, the photovoltaic power generation brings great challenges to the management and the dispatching of the power grid after being integrated into the power grid on a large scale, and the accuracy of the photovoltaic power generation prediction has great significance for the development of the photovoltaic.
Most of the existing prediction methods are a time sequence method, a regression method, an artificial neural network prediction method, an expert system method and the like. Patent CN113919545a proposes to use multiple regression models such as support vector machine regression, multiple linear regression, bayesian ridge regression model, etc. to fuse, and use Q-learning algorithm to fuse the prediction results of each sub-model, so as to improve the prediction accuracy through model fusion, but neglect the influence of the most original data features. Patent CN115222024B proposes to improve prediction accuracy by building a photovoltaic prediction neural network based on a gated cyclic neural network, and building a deep reinforcement learning feature selection network for feature selection at an input layer of the network to perform feature selection. The method has a good prediction effect on single and small-scale photovoltaic power, but cannot solve the problems of mutation and multiple dependence of photovoltaic power generation power on meteorological factors. Therefore, how to rapidly and accurately predict the photovoltaic power generation is a problem to be solved.
Disclosure of Invention
The invention aims to provide a photovoltaic power generation power prediction method and a system based on model fusion, which are characterized in that long and short time sequence features in photovoltaic time sequence data are respectively extracted in a multi-model fusion mode, and are subjected to feature fusion and prediction output to obtain a final photovoltaic power generation power prediction result so as to improve the prediction precision of distributed photovoltaic power generation power.
In order to solve the technical problems, the invention adopts the following technical scheme:
in a first aspect, the present application provides a photovoltaic power generation power prediction method based on model fusion, which includes:
acquiring historical data of a region to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data;
converting and extracting the photovoltaic power generation time sequence data by using a network model based on GAF-CNN to obtain corresponding short time sequence characteristics;
combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long time sequence features;
and fusing the short time sequence characteristic and the long time sequence characteristic through a power prediction mixed model based on GAF-CNN-LSTM-Attention, and outputting a photovoltaic power generation power prediction result.
Further, the step of obtaining the historical data of the area to be predicted to perform feature extraction, and clustering the extracted features by using a gaussian mixture model to obtain the photovoltaic power generation time sequence data includes:
acquiring historical meteorological data of a region to be predicted, preprocessing the historical meteorological data, and dividing the historical meteorological data into seasonal data according to seasons;
carrying out descriptive statistical feature extraction on preset meteorological indexes based on seasonal data, and clustering the extracted features by utilizing a Gaussian mixture model to obtain a corresponding weather type clustering result;
and acquiring historical photovoltaic power generation power data of the area to be predicted, and respectively extracting photovoltaic power generation power data under different weather types according to the weather type clustering result to form photovoltaic power generation time sequence data.
Further, the step of obtaining the historical meteorological data of the area to be predicted for preprocessing includes:
and acquiring historical meteorological data of the area to be predicted, removing abnormal values of the historical meteorological data by using a Box-Cox conversion method, and filling the missing values by using a linear interpolation method to obtain complete historical meteorological data.
Further, the step of converting and extracting the photovoltaic power generation time sequence data by using the network model based on the GAF-CNN to obtain the corresponding short time sequence feature comprises the following steps:
converting one-dimensional photovoltaic power generation time sequence data into a two-dimensional matrix image by using a GAF method;
and extracting the spatial features and the hidden features of the image by using a preset CNN network to obtain corresponding short-time sequence features.
Further, the steps of combining the photovoltaic power generation time sequence data with weather data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by using a network model based on LSTM-Attention to obtain corresponding long time sequence features include:
combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data;
normalizing the multi-variable photovoltaic power generation time sequence data, and rolling and cutting to obtain a plurality of groups of data to be detected;
and respectively inputting each group of data to be detected into a preset network model based on LSTM-attribute for feature extraction, and obtaining corresponding long time sequence features.
Further, the method further comprises the following steps: the effect evaluation is carried out on the power prediction mixed model based on the GAF-CNN-LSTM-Attention by the MAE index, and the calculation formula is as follows:
;
wherein n is the number of samples involved in the evaluation,is the predicted value of the photovoltaic power generation power, +.>Is a true value of photovoltaic power generation power.
In a second aspect, the present application provides a model fusion-based photovoltaic power generation power prediction system, comprising:
the data clustering module is used for acquiring historical data of the area to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data;
the short-time feature extraction module is used for converting and extracting the photovoltaic power generation time sequence data by utilizing a network model based on GAF-CNN to obtain corresponding short-time sequence features;
the long-term feature extraction module is used for combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long-term time sequence features;
the characteristic fusion prediction module is used for fusing the short-time sequence characteristic and the long-time sequence characteristic through a power prediction mixed model based on GAF-CNN-LSTM-Attention and outputting a photovoltaic power generation power prediction result.
In a third aspect, the present application provides an electronic device comprising a memory for storing one or more programs; a processor. The method as described in any one of the first aspects is implemented when the one or more programs are executed by the processor.
In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method as described in any of the first aspects above.
Compared with the prior art, the invention has at least the following advantages or beneficial effects:
the application provides a photovoltaic power generation power prediction method and system based on model fusion. And then, converting and extracting the photovoltaic power generation time sequence data by using a network model based on GAF-CNN to obtain corresponding short time sequence characteristics, and greatly retaining the integrity and time dependence of the time sequence. Meanwhile, the photovoltaic power generation time sequence data and meteorological data at the same time are combined to form multi-variable photovoltaic power generation time sequence data, and a network model based on LSTM-Attention is utilized for feature extraction to obtain corresponding long time sequence features, so that model prediction errors are reduced. And finally, fusing the short time sequence characteristic and the long time sequence characteristic through a power prediction hybrid model, and outputting a photovoltaic power generation power prediction result, thereby improving the prediction precision of the distributed photovoltaic power generation power.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram illustrating steps of an embodiment of a method for predicting photovoltaic power based on model fusion according to the present invention;
FIG. 2 is a schematic flow chart of an embodiment of a photovoltaic power generation power prediction method based on model fusion provided by the invention;
FIG. 3 is a schematic diagram illustrating steps for extracting historical data features of an embodiment of a photovoltaic power generation power prediction method based on model fusion according to the present invention;
FIG. 4 is a block diagram illustrating an embodiment of a model fusion-based photovoltaic power generation power prediction system according to the present invention;
fig. 5 is a block diagram of an electronic device according to an embodiment of the present invention.
Icon: 1. a memory; 2. a processor; 3. a communication interface; 11. a data clustering module; 12. a short-time feature extraction module; 13. a long-term feature extraction module; 14. and the characteristic fusion prediction module.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
Examples
Some embodiments of the present application are described in detail below with reference to the accompanying drawings. The various embodiments and features of the embodiments described below may be combined with one another without conflict.
Referring to fig. 1 and 2, an embodiment of the present application provides a photovoltaic power generation power prediction method based on model fusion, which includes the following steps:
step S1: and acquiring historical data of the area to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data.
In the above steps, firstly, the historical data of the area to be predicted is obtained, the meteorological data can be divided into several blocks of seasonal data according to the seasonal division of the area, and then, the characteristic extraction is carried out on each block of seasonal data by taking the day as a unit. And clustering by adopting a Gaussian mixture model based on the extracted features to respectively obtain sunny days, cloudy days, rainy days, weather days of other categories and the like in different seasons. And then respectively extracting photovoltaic power generation data of different weather types in different seasons according to the clustering result to form photovoltaic power generation time sequence data. Specifically, referring to fig. 3, the process mainly includes the following steps:
step S1-1: and acquiring historical meteorological data of the area to be predicted, preprocessing the historical meteorological data, and dividing the historical meteorological data into seasonal data according to seasons.
In the above steps, firstly, historical meteorological data of an area to be predicted for at least one year is obtained, and the historical meteorological data mainly comprises the following index data: intensity of light radiation, cloud cover, temperature, humidity, wind direction, wind speed, rainfall, and the like. The data is then pre-processed. By way of example, the Box-Cox transformation method can be utilized to reject abnormal values of the historical meteorological data, and the linear interpolation method is adopted to fill the missing values, so that the complete historical meteorological data is obtained. Then, the weather data for one year is divided into different seasonal weather data according to seasons. The Box-Cox is a data transformation mode in statistical modeling, is used for the condition that continuous variables do not meet normal distribution, and can reduce unobservable errors to a certain extent. The specific formula is as follows:
;
wherein, the liquid crystal display device comprises a liquid crystal display device,for the new variable after transformation, +.>For the original continuous variable, ++>Is a transformation parameter; at->Logarithmic transformation when=0, ++>Reciprocal transformation when = -1, +.>The square root transformation is performed when=0.5.
Step S1-2: descriptive statistical feature extraction is carried out on preset meteorological indexes based on seasonal data, and the extracted features are clustered by utilizing a Gaussian mixture model to obtain a corresponding weather type clustering result.
In the above step, for each seasonal meteorological data, a descriptive statistical feature extraction is performed on a preset meteorological index by taking a day as a unit, and the method mainly comprises the following features: daily maximum, daily minimum, daily average, daily variance, etc. And clustering the daily features by adopting a Gaussian mixture model according to the extracted daily descriptive statistical features to respectively obtain weather daily clustering results of sunny days, cloudy days, rainy and snowy days and other types in each season. The formula of the Gaussian mixture model is as follows:
;
k is the number of models, namely the number of clusters;to belong to the probability of the kth gaussian cluster, it needs to satisfy a probability greater than zero, and for one x, +.>The sum is equal to 1; />Probability density for kth Gaussian cluster with mean vector of +.>;/>Is a covariance matrix.
Step S1-3: and acquiring historical photovoltaic power generation power data of the area to be predicted, and respectively extracting photovoltaic power generation power data under different weather types according to the weather type clustering result to form photovoltaic power generation time sequence data.
According to the weather type clustering result, the photovoltaic power generation power data corresponding to different weather types in different seasons are respectively extracted, and the photovoltaic power generation time sequence data is formed and used for subsequent further feature extraction.
Step S2: and converting and extracting the photovoltaic power generation time sequence data by using a network model based on GAF-CNN to obtain corresponding short time sequence characteristics.
The method mainly comprises the following steps:
step S2-1: and converting the one-dimensional photovoltaic power generation time sequence data into a two-dimensional matrix image by using a GAF method.
In the above steps, the GAF (Gramian Angular Field, glamer angle field) method is to convert a one-dimensional time sequence in a cartesian coordinate system into a polar coordinate system representation, and then generate a GAF matrix using a trigonometric function. The calculation formula is as follows:
;
wherein, the liquid crystal display device comprises a liquid crystal display device,for the time stamp, N is dividing the unit length of the polar coordinate into N equal parts, +.>For the time series values after normalization, +.>Is the time series after normalization. The GAF matrix can be obtained by the above formula as follows:
;
wherein, the liquid crystal display device comprises a liquid crystal display device,is the angle of the ith time point on the polar coordinates.
The size of the GAF matrix is smaller than that of the time series, and the longer time series length leads to the increase of the dimension of the GAF matrix, which is less computationally efficient. Based on this, the data processing method in step S1 divides the longer time series data into time series data in different seasons and different weather, shortens the sequence length while retaining the sequence trend, effectively reduces the dimension of the generated GAF matrix, and is beneficial to improving the calculation efficiency. By using the GAF method, a one-dimensional time series can be converted into a two-dimensional image, which retains the integrity and time dependency of the time series to a great extent.
Step S2-2: and extracting the spatial features and the hidden features of the image by using a preset CNN network to obtain corresponding short-time sequence features.
In the above steps, the constructed CNN network model mainly includes an input layer, a two-dimensional convolution layer, a pooling layer, a full connection layer, an output layer, and the like. Due to the characteristics of sparse connection, weight sharing, pooling operation and the like, the CNN still has higher expression capability under the condition of having fewer network layers. In addition, the CNN introduces a convolution kernel, so that the risks of parameter connection and overfitting are reduced, and parameter sharing can be realized. The pooling operation plays a role in secondary extraction of the features, and the calculated amount is greatly reduced. Illustratively, an Adam optimization algorithm is used in the model to minimize the loss function of the network and an MSE index is used as the loss function. The acquired GAF two-dimensional image is input into a built CNN network model, and the time sequence characteristics and the short time sequence characteristics of the photovoltaic time sequence hidden in the two-dimensional image can be acquired.
Step S3: and combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long time sequence features.
In the above steps, the photovoltaic power generation time series data and the meteorological data at the same time are combined to form the multivariable photovoltaic power generation time series data. The meteorological data comprise light radiation intensity, cloud cover, temperature, humidity, wind direction, wind speed, rainfall and the like. And then carrying out normalization processing on the multi-variable photovoltaic power generation time sequence data, and carrying out rolling segmentation to obtain a plurality of groups of data to be detected. And then, respectively inputting each group of data to be detected into a preset network model based on LSTM-attribute for feature extraction, and obtaining corresponding long time sequence features.
Specifically, each group of data to be detected comprises a detection number and characteristic variables, each group of data is respectively input into the constructed multilayer LSTM network, and time sequence change information of nonlinear data such as photovoltaic power generation power, weather and the like is extracted by utilizing a memory function of the LSTM. The LSTM network introduces an input gate, a forget gate, an output gate, and also adds candidate states, cellular states, and hidden states. The problem of gradient disappearance can be alleviated by long-term memory of cell state storage, and short-term memory by hidden state storage. Information in a multilayer LSTM network is transferred layer by layer until the output of the last LSTM hidden layer enters the Attention (Attention mechanism) layer. The attention mechanism may enhance the effect of important time steps in LSTM, thereby further reducing model prediction errors. The LSTM hidden layer output vector serves as the input to the attention layer, essentially being a weighted average of the last layer LSTM output vector. Finally training is carried out through a full-connection layer, and the output of the full-connection layer is normalized by using a softmax function to obtain the distribution weight of each hidden layer vector. The weight size represents the importance degree of the hidden state of each time step to the prediction result, and the training process is as follows:
;
and then the trained weight is used for carrying out weighted average summation on the hidden layer output vector, and the calculation result is as follows:
;
wherein, the liquid crystal display device comprises a liquid crystal display device,score output for each hidden layer, +.>Is a weight coefficient>For the weighted sum result, +.>For the output of the last LSTM hidden layer, softmax is the activation function.
Step S4: and fusing the short time sequence characteristic and the long time sequence characteristic through a power prediction mixed model based on GAF-CNN-LSTM-Attention, and outputting a photovoltaic power generation power prediction result.
In the above step, the short time sequence features and the long time sequence features extracted in the steps S2 and S3 are fused through a power prediction mixed model formed by GAF-CNN-LSTM-Attention, and the fused time sequence features are subjected to feature transformation by a full-connection layer positioned at the tail end of the whole mixed network model, so that the photovoltaic power generation power prediction data at the next moment is finally directly output.
In some embodiments of the invention, further comprising:
the effect evaluation is carried out on the power prediction mixed model based on the GAF-CNN-LSTM-attribute through the MAE index so as to optimize the model, and the calculation formula is as follows:
;
wherein n is the number of samples involved in the evaluation,is the predicted value of the photovoltaic power generation power, +.>Is a true value of photovoltaic power generation power.
Based on the same inventive concept, the invention also provides a photovoltaic power generation power prediction system based on model fusion, referring to fig. 4, fig. 4 is a structural block diagram of the photovoltaic power generation power prediction system based on model fusion provided in the embodiment of the present application. The system comprises:
the data clustering module 11 is used for acquiring historical data of a region to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data;
the short-time feature extraction module 12 is configured to convert and extract the above-mentioned photovoltaic power generation time sequence data by using a network model based on GAF-CNN, so as to obtain corresponding short-time sequence features;
the long-term feature extraction module 13 is used for combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long-term time sequence features;
the feature fusion prediction module 14 is configured to fuse the short-time sequence feature and the long-time sequence feature through a power prediction hybrid model based on GAF-CNN-LSTM-Attention, and output a photovoltaic power generation power prediction result.
Referring to fig. 5, fig. 5 is a block diagram of an electronic device according to an embodiment of the present application. The electronic device comprises a memory 1, a processor 2 and a communication interface 3, wherein the memory 1, the processor 2 and the communication interface 3 are electrically connected with each other directly or indirectly so as to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 1 may be used for storing software programs and modules, such as program instructions/modules corresponding to a model fusion-based photovoltaic power generation power prediction system provided in the embodiments of the present application, and the processor 2 executes the software programs and modules stored in the memory 1, thereby executing various functional applications and data processing. The communication interface 3 may be used for communication of signaling or data with other node devices.
The foregoing description is only of the preferred embodiments of the present application and is not intended to limit the same, but rather, various modifications and variations may be made by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (9)
1. The photovoltaic power generation power prediction method based on model fusion is characterized by comprising the following steps of:
acquiring historical data of a region to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data;
converting and extracting the photovoltaic power generation time sequence data by using a network model based on GAF-CNN to obtain corresponding short time sequence characteristics;
combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long time sequence features;
and fusing the short time sequence characteristic and the long time sequence characteristic through a power prediction mixed model based on GAF-CNN-LSTM-Attention, and outputting a photovoltaic power generation power prediction result.
2. The method for predicting the photovoltaic power generation power based on model fusion according to claim 1, wherein the steps of obtaining the historical data of the area to be predicted, extracting the features, and clustering the extracted features by using a gaussian mixture model to obtain the photovoltaic power generation time sequence data comprise the following steps:
acquiring historical meteorological data of a region to be predicted, preprocessing the historical meteorological data, and dividing the historical meteorological data into seasonal data according to seasons;
carrying out descriptive statistical feature extraction on preset meteorological indexes based on seasonal data, and clustering the extracted features by utilizing a Gaussian mixture model to obtain a corresponding weather type clustering result;
and acquiring historical photovoltaic power generation power data of the area to be predicted, and respectively extracting photovoltaic power generation power data under different weather types according to the weather type clustering result to form photovoltaic power generation time sequence data.
3. The method for predicting photovoltaic power generation power based on model fusion according to claim 2, wherein the step of obtaining historical meteorological data of the area to be predicted for preprocessing comprises:
and acquiring historical meteorological data of the area to be predicted, removing abnormal values of the historical meteorological data by using a Box-Cox conversion method, and filling the missing values by using a linear interpolation method to obtain complete historical meteorological data.
4. The method for predicting photovoltaic power generation power based on model fusion according to claim 1, wherein the step of converting and extracting features of the photovoltaic power generation time sequence data by using a network model based on GAF-CNN to obtain corresponding short time sequence features comprises:
converting one-dimensional photovoltaic power generation time sequence data into a two-dimensional matrix image by using a GAF method;
and extracting the spatial features and the hidden features of the image by using a preset CNN network to obtain corresponding short-time sequence features.
5. The method for predicting photovoltaic power generation power based on model fusion according to claim 1, wherein the steps of combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by using a network model based on LSTM-Attention to obtain corresponding long time sequence features comprise:
combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data;
normalizing the multi-variable photovoltaic power generation time sequence data, and rolling and cutting to obtain a plurality of groups of data to be detected;
and respectively inputting each group of data to be detected into a preset network model based on LSTM-attribute for feature extraction, and obtaining corresponding long time sequence features.
6. The model fusion-based photovoltaic power generation power prediction method according to claim 1, further comprising: the effect evaluation is carried out on the power prediction mixed model based on the GAF-CNN-LSTM-Attention by the MAE index, and the calculation formula is as follows:
;
wherein n is the number of samples involved in the evaluation,is the predicted value of the photovoltaic power generation power, +.>Is a true value of photovoltaic power generation power.
7. The utility model provides a photovoltaic power generation power prediction system based on model fuses which characterized in that includes:
the data clustering module is used for acquiring historical data of the area to be predicted, extracting features, and clustering the extracted features by using a Gaussian mixture model to obtain photovoltaic power generation time sequence data;
the short-time feature extraction module is used for converting and extracting the photovoltaic power generation time sequence data by utilizing a network model based on GAF-CNN to obtain corresponding short-time sequence features;
the long-term feature extraction module is used for combining the photovoltaic power generation time sequence data with meteorological data at the same time to form multivariable photovoltaic power generation time sequence data, and extracting features by utilizing a network model based on LSTM-attribute to obtain corresponding long-term time sequence features;
the characteristic fusion prediction module is used for fusing the short-time sequence characteristic and the long-time sequence characteristic through a power prediction mixed model based on GAF-CNN-LSTM-Attention and outputting a photovoltaic power generation power prediction result.
8. An electronic device, comprising:
a memory for storing one or more programs;
a processor;
a method of model fusion-based photovoltaic power generation power prediction as claimed in any one of claims 1-6, when said one or more programs are executed by said processor.
9. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a model fusion based photovoltaic power generation power prediction method according to any of claims 1-6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310769184.7A CN116502074A (en) | 2023-06-28 | 2023-06-28 | Model fusion-based photovoltaic power generation power prediction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310769184.7A CN116502074A (en) | 2023-06-28 | 2023-06-28 | Model fusion-based photovoltaic power generation power prediction method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116502074A true CN116502074A (en) | 2023-07-28 |
Family
ID=87330539
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310769184.7A Pending CN116502074A (en) | 2023-06-28 | 2023-06-28 | Model fusion-based photovoltaic power generation power prediction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116502074A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117353302A (en) * | 2023-12-04 | 2024-01-05 | 北京东润环能科技股份有限公司 | New energy power generation power prediction method, device, equipment and medium |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113128793A (en) * | 2021-05-19 | 2021-07-16 | 中国南方电网有限责任公司 | Photovoltaic power combination prediction method and system based on multi-source data fusion |
-
2023
- 2023-06-28 CN CN202310769184.7A patent/CN116502074A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113128793A (en) * | 2021-05-19 | 2021-07-16 | 中国南方电网有限责任公司 | Photovoltaic power combination prediction method and system based on multi-source data fusion |
Non-Patent Citations (2)
Title |
---|
张淑清等: ""基于格拉姆角场与改进的风电功率预测方法"", 《电网技术》, pages 1540 - 1548 * |
王艺霏等: ""基于长-短时序特征融合的资源负载预测模型"", 《计算机应用》, pages 1508 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117353302A (en) * | 2023-12-04 | 2024-01-05 | 北京东润环能科技股份有限公司 | New energy power generation power prediction method, device, equipment and medium |
CN117353302B (en) * | 2023-12-04 | 2024-02-13 | 北京东润环能科技股份有限公司 | New energy power generation power prediction method, device, equipment and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Li et al. | Photovoltaic power forecasting with a hybrid deep learning approach | |
Lin et al. | Short-term power prediction for photovoltaic power plants using a hybrid improved Kmeans-GRA-Elman model based on multivariate meteorological factors and historical power datasets | |
CN107194495B (en) | Photovoltaic power longitudinal prediction method based on historical data mining | |
Yang et al. | Day-ahead forecasting of photovoltaic output power with similar cloud space fusion based on incomplete historical data mining | |
CN114792156B (en) | Photovoltaic output power prediction method and system based on curve characteristic index clustering | |
Wang et al. | Comparative study of machine learning approaches for predicting short-term photovoltaic power output based on weather type classification | |
CN110929953A (en) | Photovoltaic power station ultra-short term output prediction method based on cluster analysis | |
CN111832812A (en) | Wind power short-term prediction method based on deep learning | |
CN111242355A (en) | Photovoltaic probability prediction method and system based on Bayesian neural network | |
CN114399081A (en) | Photovoltaic power generation power prediction method based on weather classification | |
CN115796004A (en) | Photovoltaic power station ultra-short term power intelligent prediction method based on SLSTM and MLSTNet models | |
CN115829126A (en) | Photovoltaic power generation power prediction method based on multi-view self-adaptive feature fusion | |
CN116502074A (en) | Model fusion-based photovoltaic power generation power prediction method and system | |
He et al. | A power forecasting approach for PV plant based on irradiance index and LSTM | |
CN115115125A (en) | Photovoltaic power interval probability prediction method based on deep learning fusion model | |
CN116402203A (en) | Method, system and medium for predicting short-time photovoltaic power generation capacity considering weather conditions | |
CN116451821A (en) | Deep learning-based large-area new energy generation power prediction method and system | |
Yu et al. | Prediction of solar irradiance one hour ahead based on quantum long short-term memory network | |
CN115456235A (en) | Photovoltaic power prediction system based on multi-mode fusion | |
Omar et al. | Seasonal clustering forecasting technique for intelligent hourly solar irradiance systems | |
CN111275256A (en) | Photovoltaic power generation power day-ahead prediction method based on image feature extraction | |
CN108694475B (en) | Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model | |
CN113837434A (en) | Solar photovoltaic power generation prediction method and device, electronic equipment and storage medium | |
Ziyabari et al. | Short-term solar irradiance forecasting based on multi-branch residual network | |
Xu et al. | One-day ahead forecast of PV output based on deep belief network and weather classification |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230728 |