A method, device, equipment, medium and product for predicting power of photovoltaic power station

Abstract

The present invention discloses a power prediction method, device, equipment, medium, and product for a photovoltaic power station. The method includes: obtaining power adjacency matrices corresponding to a photovoltaic power station to be tested and a reference photovoltaic power station, and obtaining reference weather data sets corresponding to at least two preset weather conditions, respectively; determining a power spatial feature vector of the photovoltaic power station to be tested based on a spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data of the reference photovoltaic power station; determining, for each preset weather condition, a standard power sequence of the photovoltaic power station to be tested under the preset weather condition based on the historical power data to be tested of the photovoltaic power station to be tested, the power spatial feature vector, the reference weather data set corresponding to the preset weather condition, and a weather power prediction model; and obtaining, for each future moment, a target power of the photovoltaic power station to be tested at the future moment from the standard power sequence corresponding to the weather condition to be tested at the future moment, thereby improving the stability of power prediction.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation technology, and in particular to a power prediction method, device, equipment, medium and product for a photovoltaic power station. Background Art

[0002] With the proposal of my country's "dual carbon" goals and the continuous advancement of its energy transformation strategy, photovoltaic power generation, as a power generation technology with advantages such as short construction period, no risk of depletion and high energy quality, has gradually occupied an important position in the energy strategy.

[0003] In order to reduce the impact of photovoltaic output instability on the safe and economic operation of the power grid, it is necessary to accurately predict the power generation of photovoltaic power stations. Complex meteorological conditions have a great impact on the power generation capacity of photovoltaic power stations. Therefore, the predicted meteorological data of photovoltaic power stations is used as key reference data for power generation prediction.

[0004] However, in actual applications, photovoltaic power stations will inevitably have some missing meteorological data, so traditional power prediction methods cannot guarantee the prediction accuracy of power generation. Summary of the Invention

[0005] Embodiments of the present invention provide a power prediction method, device, equipment, medium and product for a photovoltaic power station to solve the problem of poor power prediction accuracy of a photovoltaic power station due to missing meteorological data and improve the stability of the power prediction method.

[0006] According to an embodiment of the present invention, a method for predicting power of a photovoltaic power station is provided, the method comprising:

[0007] Obtaining a power adjacency matrix corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station; wherein the power adjacency matrix represents the power correlation between the photovoltaic power stations;

[0008] Obtaining reference weather data sets corresponding to at least two preset weather conditions; wherein the reference weather data sets include historical meteorological characteristics of the reference photovoltaic power station;

[0009] Determining a power spatial feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station;

[0010] For each preset weather condition, a pre-trained weather power prediction model corresponding to the preset weather condition is obtained, and a standard power sequence of the photovoltaic power station under test under the preset weather condition is determined based on the historical power data to be measured of the photovoltaic power station under test, the power spatial feature vector, a reference weather data set corresponding to the preset weather condition, and the weather power prediction model; wherein the standard power sequence includes at least one standard power corresponding to each future moment;

[0011] For each future moment, the weather to be measured of the photovoltaic power station to be measured at the future moment is obtained, and the target power of the photovoltaic power station to be measured at the future moment is obtained from the standard power sequence corresponding to the weather to be measured.

[0012] According to another embodiment of the present invention, a power prediction device for a photovoltaic power station is provided, the device comprising:

[0013] A power adjacent matrix acquisition module is used to obtain power adjacent matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station; wherein the power adjacent matrix represents the power correlation between photovoltaic power stations;

[0014] A reference weather data set determination module is configured to obtain reference weather data sets corresponding to at least two preset weather conditions; wherein the reference weather data sets include historical meteorological characteristics of the reference photovoltaic power station;

[0015] a power space feature vector determination module, configured to determine the power space feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station;

[0016] a standard power sequence determination module, configured to obtain, for each preset weather condition, a pre-trained weather power prediction model corresponding to the preset weather condition, and determine, based on the historical power data to be measured of the photovoltaic power station to be tested, the power spatial feature vector, a reference weather data set corresponding to the preset weather condition, and the weather power prediction model, a standard power sequence for the photovoltaic power station to be tested under the preset weather condition; wherein the standard power sequence includes at least one standard power corresponding to each future moment;

[0017] The target power determination module is used to obtain the weather to be tested of the photovoltaic power station to be tested at each future moment, and obtain the target power of the photovoltaic power station to be tested at the future moment from the standard power sequence corresponding to the weather to be tested.

[0018] According to another embodiment of the present invention, an electronic device is provided, the electronic device including:

[0019] at least one processor; and

[0020] a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the power prediction method for a photovoltaic power station according to any embodiment of the present invention.

[0022] According to another embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power prediction method for a photovoltaic power station according to any embodiment of the present invention when executed.

[0023] According to another embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the computer program implements the power prediction method for a photovoltaic power station according to any embodiment of the present invention.

[0024] The technical solution of the embodiment of the present invention utilizes the characteristic that the output characteristics between photovoltaic power stations are correlated to obtain the power adjacent matrices corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station, determines the power spatial feature vector of the photovoltaic power station to be tested based on the pre-trained spatial feature extraction model, the power adjacent matrix and the reference historical meteorological data corresponding to at least one reference photovoltaic power station, and adopts the weather dimension instead of the traditional meteorological dimension to obtain reference weather data sets corresponding to at least two preset weathers, and uses the weather power prediction model corresponding to each preset weather to predict the standard power sequence of the photovoltaic power station to be tested under the preset weather, and searches for the target power of the photovoltaic power station to be tested from the standard power sequence corresponding to the tested weather of the photovoltaic power station to be tested, thereby solving the problem of poor power prediction accuracy of the photovoltaic power station due to the lack of meteorological data, and improving the stability of the power prediction of the photovoltaic power station while ensuring the accuracy of the power prediction of the photovoltaic power station.

[0025] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 A flowchart of a method for predicting power of a photovoltaic power station provided by one embodiment of the present invention;

[0028] Figure 2 A flowchart of another photovoltaic power station power prediction method provided by one embodiment of the present invention;

[0029] Figure 3 A schematic diagram of a specific example of a power prediction method for a photovoltaic power station provided by one embodiment of the present invention;

[0030] Figure 4 A schematic structural diagram of a power prediction device for a photovoltaic power station provided by one embodiment of the present invention;

[0031] Figure 5 The present invention provides a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "to be measured", "preset", "target", "reference", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0034] Figure 1This is a flow chart of a power prediction method for a photovoltaic power station provided by an embodiment of the present invention. This embodiment is applicable to the case of predicting the power generated by a photovoltaic power station at a future time, and is particularly applicable to a distributed photovoltaic power station system. The method can be executed by a power prediction device of a photovoltaic power station, which can be implemented in the form of hardware and / or software, and can be configured in a terminal device. Figure 1 As shown, the method includes:

[0035] S110: Obtain power adjacency matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station.

[0036] In this embodiment, the power adjacency matrix represents the power correlation between photovoltaic power stations. Exemplarily, the power adjacency matrix is ​​an n-order block matrix, where n is a positive integer greater than or equal to 2, representing the total number of photovoltaic power stations corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station.

[0037] In an optional embodiment, the method further includes: obtaining preset power station locations corresponding to at least two preset photovoltaic power stations; for each preset photovoltaic power station, determining a power station distance based on the preset power station location corresponding to the preset photovoltaic power station and the power station location to be tested of the photovoltaic power station to be tested; and screening the at least two preset photovoltaic power stations based on the at least two power station distances to obtain at least one reference photovoltaic power station.

[0038] In an optional embodiment, at least two preset photovoltaic power stations are screened according to at least two power station distances to obtain at least one reference photovoltaic power station, including: using a preset photovoltaic power station whose power station distance is greater than a preset distance threshold as a reference photovoltaic power station.

[0039] Exemplarily, the preset distance threshold may be 10 km. The preset distance threshold is not limited here and may be customized according to actual needs.

[0040] In another optional embodiment, at least two preset photovoltaic power stations are screened according to the distances between at least two power stations to obtain at least one reference photovoltaic power station, including: sorting the at least one preset photovoltaic power station according to the distances between the at least two power stations, and determining at least one reference photovoltaic power station from the obtained sorting results according to a preset screening number.

[0041] Exemplarily, the preset number of filters can be 50. There is no limit on the preset number of filters here, and the specific settings can be customized according to actual needs.

[0042] Specifically, when the sorting method is ascending, a preset number of reference photovoltaic power stations are selected from the sorting results from left to right; when the sorting method is descending, a preset number of reference photovoltaic power stations are selected from the sorting results from right to left.

[0043] The advantage of this setting is that it takes advantage of the strong correlation between the output characteristics of adjacent photovoltaic power stations to perform distance screening on the preset photovoltaic power stations. On the one hand, it improves the accuracy of the power adjacency matrix, which in turn helps to improve the accuracy of the power prediction of the photovoltaic power station. On the other hand, it reduces the number of features contained in the power adjacency matrix, thereby improving the efficiency of the power prediction of the photovoltaic power station.

[0044] Specifically, the matrix elements in the power adjacency matrix represent the power correlation coefficients between the two PV plants. The power correlation coefficient represents the degree of correlation between the historical power data of the two PV plants. The power correlation coefficient typically ranges from -1 to 1. Values ​​closer to -1 indicate a negative correlation between the historical power data of the two PV plants, values ​​closer to 1 indicate a positive correlation, and values ​​closer to 0 indicate no correlation between the historical power data of the two PV plants.

[0045] In an optional embodiment, obtaining power adjacency matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station includes: inputting the historical power data to be tested of the photovoltaic power station to be tested and the reference historical power data corresponding to at least one reference photovoltaic power station into a pre-trained power-related prediction model to obtain an output power adjacency matrix.

[0046] The historical power data includes power data corresponding to at least one historical time point. For example, the historical time range corresponding to the historical power data can be 30 days or 60 days. The historical time range corresponding to the historical power data is not limited here and can be customized according to actual needs.

[0047] In another optional embodiment, the power adjacency matrix is ​​calculated using a preset correlation coefficient algorithm. Exemplarily, the preset correlation coefficient algorithm includes but is not limited to the Pearson correlation coefficient, the Salman correlation coefficient, the Chebyshev correlation coefficient, the Gini coefficient, and the Kendall rank correlation coefficient, etc. The preset correlation coefficient algorithm is not limited here and can be customized according to actual needs.

[0048] In an optional embodiment, obtaining a power adjacency matrix corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station includes: determining the power covariance corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station based on the historical power data to be tested of the photovoltaic power station to be tested and the reference historical power data of the reference photovoltaic power station; determining the standard deviation of the power to be tested corresponding to the photovoltaic power station to be tested based on the historical power data to be tested, and determining the reference power standard deviation corresponding to the reference photovoltaic power station based on the reference historical power data; determining the power correlation coefficient corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station based on the power covariance, the standard deviation of the power to be tested, and the reference power standard deviation; and using the power correlation coefficient as the matrix element corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station in the power adjacency matrix.

[0049] Wherein, illustratively, the power covariance cov(X, Y) satisfies the formula:

[0050]

[0051] Among them, X represents the historical power data to be measured, Y represents the reference historical power data, and X i Indicates the i-th power data in the historical power data to be measured, Indicates the expected value corresponding to the historical power data to be measured, Y i Indicates the i-th power data in the reference historical power data, represents the expected value corresponding to the reference historical power data, and m represents the data volume of the historical power data to be measured or the power data in the reference historical power data.

[0052] Wherein, for example, the standard deviation of the power to be measured σ X Satisfies the formula:

[0053]

[0054] Where, for example, the reference power standard deviation σ Y Satisfy the formula

[0055]

[0056] Among them, exemplary, the power correlation coefficient r XY Satisfies the formula:

[0057]

[0058] It can be understood that this embodiment only provides an illustrative example of the method for calculating the power correlation coefficient between the photovoltaic power station to be tested and the reference photovoltaic power station. The method for calculating the power correlation coefficient between the two reference photovoltaic power stations is the same or similar to the method for calculating the power correlation coefficient in the above embodiment, and will not be repeated in this embodiment.

[0059] S120: Obtain reference weather data sets corresponding to at least two preset weather conditions.

[0060] In this embodiment, the reference weather dataset includes historical meteorological characteristics of the reference photovoltaic power station. These historical meteorological characteristics include at least one meteorological characteristic of the reference photovoltaic power station at a historical time point within a meteorological timeframe. Exemplary historical meteorological characteristics include, but are not limited to, temperature, air pressure, humidity, wind speed, wind direction, precipitation, visibility, light intensity, and UV index. These historical meteorological characteristics are not limited here and can be customized based on actual needs.

[0061] In an optional embodiment, the meteorological time limit corresponding to the historical meteorological characteristics is one day. For example, the reference weather dataset includes the historical meteorological characteristics of photovoltaic power station A on December 1 and the historical meteorological characteristics of photovoltaic power station B on December 2.

[0062] In an optional embodiment, the preset weather is determined using numerical weather prediction technology (NWP), and accordingly, at least two reference weather data sets corresponding to the preset weather are obtained, including: for each preset weather, adding the historical meteorological characteristics of the reference photovoltaic power station that matches the preset weather to the reference weather data set.

[0063] In this embodiment, historical meteorological characteristics are obtained using numerical weather forecasting technology. Numerical weather forecasting refers to a method for predicting atmospheric motion and weather phenomena for a specific period in the future by using large computers to perform numerical calculations based on atmospheric conditions, under certain initial and boundary conditions, to solve the fluid dynamics and thermodynamics equations that describe weather evolution.

[0064] For example, the preset weather includes but is not limited to sunny, rainy, cloudy, foggy, snowy, etc. There is no limitation on the preset weather here, and the settings can be customized according to actual needs.

[0065] S130 , determining a power spatial feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, a power adjacent matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station.

[0066] Specifically, the spatial feature extraction model is used to extract the power spatial feature vector corresponding to the photovoltaic power station under test. Exemplary, the model architecture of the spatial feature extraction model includes but is not limited to residual networks (ResNet), Transformer networks, CNN networks (Convolutional Neural Networks), FCN networks (Fully Convolutional Networks), DNN networks (Deep Neural Networks), or RNN networks (Recurrent Neural Networks), etc. The model architecture of the spatial feature extraction model is not limited here and can be customized according to actual needs.

[0067] Specifically, the reference historical meteorological data includes historical meteorological characteristics corresponding to at least one historical time point. For example, the historical time range corresponding to the reference historical meteorological data may be 30 days or 60 days. The historical time range corresponding to the reference historical meteorological data is not limited here and can be customized according to actual needs.

[0068] In an optional embodiment, the historical meteorological features in the reference historical meteorological data are collected in real time by meteorological sensors and / or acquired by numerical weather prediction (NWP) technology.

[0069] In an optional embodiment, the power spatial feature vector of the photovoltaic power station to be tested is determined based on a pre-trained spatial feature extraction model, a power adjacent matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station, including: inputting the power adjacent matrix and the reference historical meteorological data corresponding to at least one reference photovoltaic power station into the spatial feature extraction model to obtain the output power spatial feature vector of the photovoltaic power station to be tested.

[0070] S140. For each preset weather, obtain a pre-trained weather power prediction model corresponding to the preset weather, and determine the standard power sequence of the photovoltaic power station to be tested under the preset weather based on the historical power data to be tested, the power space feature vector, the reference weather data set corresponding to the preset weather, and the weather power prediction model.

[0071] Exemplarily, the model architecture of the weather power prediction model includes but is not limited to the temporal convolutional network (TCN), the support vector machine model and the long short-term memory network (LSTM), etc. The model architecture of the weather power prediction model is not limited here, and can be customized according to actual needs.

[0072] In an optional embodiment, the standard power sequence of the photovoltaic power station to be tested under the preset weather is determined based on the historical power data to be tested, the power space characteristic vector, the reference weather data set corresponding to the preset weather, and the weather power prediction model of the photovoltaic power station to be tested, including: inputting the historical power data to be tested, the power space characteristic vector, and the reference weather data set corresponding to the preset weather into the weather power prediction model to obtain the output standard power sequence of the photovoltaic power station to be tested under the preset weather.

[0073] In this embodiment, the standard power sequence includes at least one standard power corresponding to a future time. In an optional embodiment, the future time is a time within a day of a future date. For example, the standard power sequence includes standard powers corresponding to 1:00, 2:00, and 3:00 on the future date.

[0074] S150 , for each future moment, obtaining the weather conditions to be measured for the photovoltaic power station to be measured at the future moment, and obtaining the target power of the photovoltaic power station to be measured at the future moment from a standard power sequence corresponding to the weather conditions to be measured.

[0075] In an optional embodiment, obtaining the weather to be measured at the photovoltaic power station to be measured at a future time includes: using numerical weather forecasting technology to obtain the weather to be measured at the photovoltaic power station to be measured at a future time.

[0076] In another optional embodiment, obtaining the weather to be measured at the photovoltaic power station to be tested at a future time includes: classifying the future meteorological data to be measured at the photovoltaic power station to be tested at a future time to obtain the weather to be measured at the photovoltaic power station to be tested at a future time.

[0077] In another optional embodiment, obtaining the weather to be measured at the photovoltaic power station to be measured at a future moment includes: clustering the future meteorological data to be measured corresponding to the photovoltaic power station to be measured and at least two future moments, to obtain the weather to be measured at each future moment of the photovoltaic power station to be measured.

[0078] There is no limitation on the classification or clustering algorithm used here, and you can customize the settings according to actual needs.

[0079] For example, suppose the standard power sequence corresponding to preset weather A contains the standard power corresponding to 1:00 and 2:00 in the future date, which is 100W / m 2 and 220W / m 2 The standard power sequence corresponding to the preset weather B includes the standard power corresponding to 1:00 and 2:00 in the future date, which is 300W / m 2 and 500W / m 2 If the weather conditions for the PV power station to be tested at 1:00 and 2:00 in the future are preset weather A and preset weather B respectively, then the target power for the PV power station to be tested at 1:00 and 2:00 in the future is 100W / m 2 and 500W / m 2 .

[0080] The technical solution of this embodiment utilizes the characteristic that the output characteristics between photovoltaic power stations are correlated to obtain the power adjacent matrices corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station, determines the power spatial feature vector of the photovoltaic power station to be tested based on the pre-trained spatial feature extraction model, the power adjacent matrix and the reference historical meteorological data corresponding to at least one reference photovoltaic power station, and adopts the weather dimension instead of the traditional meteorological dimension to obtain reference weather data sets corresponding to at least two preset weathers, and uses the weather power prediction model corresponding to each preset weather to predict the standard power sequence of the photovoltaic power station to be tested under the preset weather, and searches for the target power of the photovoltaic power station to be tested from the standard power sequence corresponding to the test weather of the photovoltaic power station to be tested, thereby solving the problem of poor power prediction accuracy of the photovoltaic power station due to the lack of meteorological data, and improving the stability of the power prediction of the photovoltaic power station while ensuring the accuracy of the power prediction of the photovoltaic power station.

[0081] Figure 2 This is a flow chart of another method for predicting power of a photovoltaic power station provided by an embodiment of the present invention. This embodiment further refines the above embodiment's "determining the standard power sequence of the photovoltaic power station under test under preset weather conditions based on the historical power data of the photovoltaic power station under test, the power space feature vector, the reference weather data set corresponding to the preset weather conditions, and the weather power prediction model." Figure 2 As shown, the method includes:

[0082] S210: Obtain power adjacency matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station.

[0083] S210 in this embodiment is the same as that in the above embodiment. Figure 1 The steps in S110 shown are the same or similar and will not be described in detail in this embodiment.

[0084] S220: Obtain reference weather data sets corresponding to at least two preset weather conditions.

[0085] In an optional embodiment, S220 is the same as in the above embodiment. Figure 1 The S120 shown corresponds to the same or similar steps, and will not be described in detail in this embodiment.

[0086] In another optional embodiment, obtaining reference weather data sets corresponding to at least two preset weather conditions includes: classifying reference historical meteorological data corresponding to at least one reference photovoltaic power station to obtain reference weather data sets corresponding to at least two preset weather conditions.

[0087] Among them, exemplary classification algorithms include but are not limited to K-nearest neighbor algorithm, decision tree algorithm, naive Bayes algorithm, support vector machine algorithm, logistic regression algorithm, neural network algorithm, random forest algorithm, gradient boosting algorithm or AdaBoost algorithm and XGBoost algorithm. There is no limitation on the classification algorithm used here, and the specific settings can be customized according to actual needs.

[0088] In another optional embodiment, obtaining reference weather data sets corresponding to at least two preset weather conditions includes: using a hierarchical clustering algorithm to perform clustering operations on reference historical meteorological data corresponding to at least one reference photovoltaic power station to obtain reference weather data sets corresponding to at least two preset weather conditions.

[0089] Exemplarily, the algorithm type of the hierarchical clustering algorithm is an agglomerative hierarchical algorithm or a divisive hierarchical algorithm. Among them, the agglomerative hierarchical algorithm can be to assume that each data point is an independent cluster, and then merge the two most similar clusters into a new cluster by calculating the distance or similarity between all clusters. This process is repeated until all data points are merged into one cluster or the preset number of clusters is reached. The divisive hierarchical algorithm assumes that all data points are in one cluster, and then continuously splits this cluster into smaller clusters until each data point becomes a separate cluster or the preset number of clusters is reached. There is no limitation on the algorithm type of the hierarchical clustering algorithm here, and the specific settings can be customized according to actual needs.

[0090] In an optional embodiment, the hierarchical clustering algorithm is a balanced iterative reducing and clustering using hierarchies (BIRCH) algorithm.

[0091] The advantage of setting up clustering is that, since weather conditions in different regions are different, obtaining a reference weather dataset through clustering avoids the inaccuracy of pre-set weather conditions and improves the accuracy of the reference weather dataset.

[0092] Based on the above embodiment, optionally, the method further includes: preprocessing the historical power data to be measured corresponding to the photovoltaic power station to be measured and the reference historical power data corresponding to at least one reference photovoltaic power station to obtain the preprocessed historical power data to be measured and at least one reference historical power data.

[0093] Based on the above embodiment, optionally, the method further includes: preprocessing the future meteorological data to be measured corresponding to the photovoltaic power station to be measured and the reference historical meteorological data corresponding to at least one reference photovoltaic power station to obtain the preprocessed future meteorological data to be measured and at least one reference historical meteorological data.

[0094] Exemplarily, preprocessing includes, but is not limited to, data cleaning, data encoding, and normalization. Taking historical meteorological reference data as an example, data cleaning can include deleting missing values, filling out outliers, and removing duplicate values ​​within a meteorological timeframe. Data encoding can refer to the use of coded symbols to represent textual information within the historical meteorological reference data. Normalization is used to eliminate dimensional differences between different historical meteorological reference data, between different meteorological data, or between different meteorological characteristics.

[0095] The benefit of setting up preprocessing is to further improve the accuracy of power prediction of photovoltaic power plants.

[0096] S230 , determining a power spatial feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, a power adjacent matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station.

[0097] In an alternative embodiment, S230 is the same as in the above embodiment. Figure 1 The steps in S130 shown are the same or similar and will not be described again in detail in this embodiment.

[0098] In another optional embodiment, the model architecture of the spatial feature extraction model is a graph convolutional neural network. Accordingly, the power spatial feature vector of the photovoltaic power station to be tested is determined based on the pre-trained spatial feature extraction model, the power adjacent matrix and the reference historical meteorological data corresponding to at least one reference photovoltaic power station, including: constructing a meteorological feature matrix based on the reference historical meteorological data corresponding to at least one reference photovoltaic power station; inputting the power adjacent matrix and the meteorological feature matrix into the spatial feature extraction model to obtain the output power spatial feature vector of the photovoltaic power station to be tested.

[0099] Specifically, a graph convolutional network (GCN) is a neural network that uses convolution operations to process data with a generalized topological graph structure.

[0100] In this embodiment, the meteorological characteristic matrix includes a meteorological block matrix corresponding to each reference photovoltaic power station. Specifically, the matrix elements in the meteorological block matrix are meteorological characteristics of the reference photovoltaic power station.

[0101] The purpose of setting the meteorological feature matrix is ​​to make the input format of the reference historical meteorological data meet the format requirements of the graph convolutional neural network for input data.

[0102] S240: For each preset weather, obtain a pre-trained weather power prediction model corresponding to the preset weather.

[0103] In this embodiment, the model architecture of the weather power prediction model is a bidirectional long short-term memory network. The bidirectional long short-term memory network (BiLSTM) consists of two independent long short-term memory networks, one responsible for forward propagation and the other for backward propagation. At least two time series feature sequences are input into the two long short-term memory networks in forward and reverse order for feature extraction, respectively, to learn the context feature information before the current moment and the context feature information after the current moment. The standard power sequence of the photovoltaic power station under the preset weather conditions is determined based on the splicing results of the feature vectors extracted by the two long short-term memory networks.

[0104] S250 , performing modal decomposition on a target feature sequence consisting of the historical power data to be measured, the power space feature vector, and the reference weather data set to obtain at least two time series feature sequences.

[0105] Exemplarily, modal decomposition algorithms include but are not limited to empirical mode decomposition, ensemble empirical mode decomposition, complementary ensemble empirical mode decomposition, empirical wavelet decomposition and variational mode decomposition, etc. There is no limitation on the modal decomposition algorithm used here, and the specific settings can be customized according to actual needs.

[0106] S260: Input at least two time series feature sequences into a weather power prediction model to obtain an output standard power sequence of the photovoltaic power station to be tested under preset weather conditions.

[0107] Based on the above embodiment, optionally, the method further includes: performing denormalization processing on the standard power sequences corresponding to at least two preset weather conditions, to obtain at least two processed standard power sequences.

[0108] The advantage of this setting is that the scale information and distribution information of the standard power in the standard power sequence is retained, thereby improving the accuracy of the standard power sequence.

[0109] S270 , for each future moment, obtaining the weather conditions to be measured for the photovoltaic power station to be measured at the future moment, and obtaining the target power of the photovoltaic power station to be measured at the future moment from a standard power sequence corresponding to the weather conditions to be measured.

[0110] S270 in this embodiment is the same as that in the above embodiment. Figure 1 The S150 shown corresponds to the same or similar steps, and will not be described in detail in this embodiment.

[0111] The technical solution of this embodiment is to set the model architecture of the weather power prediction model to a bidirectional long short-term memory network, perform modal decomposition on the target feature sequence composed of the historical power data to be tested, the power space feature vector and the reference weather data set, and obtain at least two time series feature sequences. The at least two time series feature sequences are input into the weather power prediction model to obtain the output standard power sequence of the photovoltaic power station to be tested under preset weather conditions, which solves the problem of less feature information directly represented by a single feature sequence, enables the input data to reflect more time series change characteristics of historical data, and further improves the accuracy of power prediction of photovoltaic power stations.

[0112] Figure 3 A schematic diagram of a specific example of a power prediction method for a photovoltaic power station provided by an embodiment of the present invention. Specifically, in the data processing process, distributed photovoltaic power station data is obtained, wherein the distributed photovoltaic power station data includes the future meteorological data to be measured, the historical power data to be measured, and the reference historical meteorological data and reference historical power data of the reference photovoltaic power station. The distributed photovoltaic power station data is preprocessed. Exemplarily, the preprocessing includes data cleaning, data encoding, and normalization to obtain meteorological data and power data. The meteorological data includes the future meteorological data to be measured and the reference historical meteorological data, and the power data includes the historical power data to be measured and the reference historical power data. The reference historical meteorological data in the meteorological data is clustered to obtain a cloudy data set, a sunny data set, a rainy data set, and the like.

[0113] In the power prediction process, a meteorological feature matrix is ​​constructed based on the reference historical meteorological data, and a power adjacency matrix is ​​constructed based on the power data. These two matrixes are then fed into a pre-trained graph convolutional neural network to generate the power spatial feature vectors for the PV power plant under test. Taking a cloudy day as an example, modal decomposition is performed on the target feature sequence consisting of the cloudy day dataset, the historical power data to be measured, and the power spatial feature vectors to generate at least two time series feature sequences. These at least two time series feature sequences are then fed into a pre-trained BiLSTM model for cloudy days to generate the standard power sequence corresponding to the cloudy day.

[0114] In the target power determination process, the standard power sequence corresponding to each weather is denormalized, and the target power sequence is determined from at least one standard power sequence based on the weather data to be measured corresponding to the photovoltaic power station to be measured, wherein the weather data to be measured includes the weather to be measured of the photovoltaic power station to be measured at each future moment, and the target power sequence includes the target power of the photovoltaic power station to be measured at each future moment.

[0115] The following is an embodiment of a power prediction device for a photovoltaic power station provided by an embodiment of the present invention. The device and the power prediction method for a photovoltaic power station of the above embodiment belong to the same inventive concept. For details not fully described in the embodiment of the power prediction device for a photovoltaic power station, please refer to the content of the power prediction method for a photovoltaic power station in the above embodiment.

[0116] Figure 4 This is a schematic diagram of the structure of a power prediction device for a photovoltaic power station provided by one embodiment of the present invention. Figure 4 As shown, the apparatus includes: a power adjacent matrix acquisition module 310 , a reference weather data set determination module 320 , a power space eigenvector determination module 330 , a standard power sequence determination module 340 and a target power determination module 350 .

[0117] The power adjacent matrix acquisition module 310 is configured to acquire power adjacent matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station; wherein the power adjacent matrix represents the power correlation between photovoltaic power stations;

[0118] The reference weather data set determination module 320 is configured to obtain reference weather data sets corresponding to at least two preset weather conditions; wherein the reference weather data sets include historical meteorological characteristics of the reference photovoltaic power station;

[0119] A power spatial feature vector determination module 330 is configured to determine a power spatial feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, a power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station.

[0120] The standard power sequence determination module 340 is configured to obtain, for each preset weather condition, a pre-trained weather power prediction model corresponding to the preset weather condition, and determine a standard power sequence for the PV power station under test under the preset weather condition based on the historical power data to be measured, the power spatial feature vector, the reference weather dataset corresponding to the preset weather condition, and the weather power prediction model. The standard power sequence includes at least one standard power corresponding to each future moment.

[0121] The target power determination module 350 is used to obtain the weather conditions of the photovoltaic power station at the future moment and obtain the target power of the photovoltaic power station at the future moment from the standard power sequence corresponding to the weather conditions.

[0122] The technical solution of this embodiment utilizes the characteristic that the output characteristics between photovoltaic power stations are correlated to obtain the power adjacent matrices corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station, determines the power spatial feature vector of the photovoltaic power station to be tested based on the pre-trained spatial feature extraction model, the power adjacent matrix and the reference historical meteorological data corresponding to at least one reference photovoltaic power station, and adopts the weather dimension instead of the traditional meteorological dimension to obtain reference weather data sets corresponding to at least two preset weathers, and uses the weather power prediction model corresponding to each preset weather to predict the standard power sequence of the photovoltaic power station to be tested under the preset weather, and searches for the target power of the photovoltaic power station to be tested from the standard power sequence corresponding to the test weather of the photovoltaic power station to be tested, thereby solving the problem of poor power prediction accuracy of the photovoltaic power station due to the lack of meteorological data, and improving the stability of the power prediction of the photovoltaic power station while ensuring the accuracy of the power prediction of the photovoltaic power station.

[0123] In an optional embodiment, the model architecture of the spatial feature extraction model is a graph convolutional neural network. Accordingly, the power space feature vector determination module 330 is specifically configured to:

[0124] Constructing a meteorological characteristic matrix based on reference historical meteorological data corresponding to at least one reference photovoltaic power station; wherein the meteorological characteristic matrix includes a meteorological block matrix corresponding to each reference photovoltaic power station;

[0125] The power adjacency matrix and the meteorological characteristic matrix are input into the spatial feature extraction model to obtain the output power spatial characteristic vector of the photovoltaic power station to be tested.

[0126] In an optional embodiment, the model architecture of the weather power prediction model is a bidirectional long short-term memory network. Accordingly, the standard power sequence determination module 340 is specifically configured to:

[0127] Performing modal decomposition on a target feature sequence consisting of the historical power data to be measured, the power space feature vector, and the reference weather data set to obtain at least two time series feature sequences;

[0128] At least two time series feature sequences are input into the weather power prediction model to obtain the output standard power sequence of the photovoltaic power station to be tested under preset weather conditions.

[0129] In an optional embodiment, the power adjacent matrix acquisition module 310 is specifically configured to:

[0130] Determining the power covariance corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station based on the historical power data to be tested of the photovoltaic power station to be tested and the reference historical power data of the reference photovoltaic power station;

[0131] Determine the standard deviation of the power to be measured corresponding to the photovoltaic power station to be measured based on the historical power data to be measured, and determine the standard deviation of the reference power corresponding to the reference photovoltaic power station based on the reference historical power data;

[0132] Determine the power correlation coefficients corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station based on the power covariance, the standard deviation of the power to be tested, and the standard deviation of the reference power;

[0133] The power correlation coefficient is used as the matrix element corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station in the power adjacency matrix.

[0134] In an optional embodiment, the reference weather data set determination module 320 is specifically configured to:

[0135] A hierarchical clustering algorithm is used to perform clustering operations on reference historical meteorological data corresponding to at least one reference photovoltaic power station, to obtain reference weather data sets corresponding to at least two preset weather conditions.

[0136] In an optional embodiment, the meteorological time limit corresponding to the historical meteorological feature is one day, and the future time is a time within one day of a future date.

[0137] The power prediction device for a photovoltaic power station provided by an embodiment of the present invention can execute the power prediction method for a photovoltaic power station provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0138] Figure 5 A schematic diagram of the structure of an electronic device provided for one embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0139] like Figure 5As shown, the electronic device 10 includes at least one processor 11 and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by the at least one processor 11, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12 and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0140] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information or data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0141] The processor 11 can be a variety of general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the power prediction method for a photovoltaic power station provided in the above embodiments.

[0142] In some embodiments, the photovoltaic power plant power prediction method provided in the above embodiments can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the photovoltaic power plant power prediction method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the photovoltaic power plant power prediction method through any other appropriate means (e.g., via firmware).

[0143] Various embodiments of the systems and techniques described herein can be implemented in the following systems or combinations thereof: digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard parts (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0144] Computer programs for implementing the photovoltaic power plant power prediction method of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0145] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable storage medium. Examples of machine-readable storage media can include an electrical connection based on at least one line, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0146] To provide interaction with a user, the systems and techniques described herein can be implemented on a terminal device having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball), through which the user can provide input to the terminal device. Other types of devices can also provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0147] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0148] A computing system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. The client-server relationship arises through computer programs running on the respective computers and establishing a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within a cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosts and virtual private server (VPS) services.

[0149] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0150] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for predicting power of a photovoltaic power station, characterized in that: include: Obtaining a power adjacency matrix corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station; wherein the power adjacency matrix represents the power correlation between the photovoltaic power stations; Obtaining reference weather data sets corresponding to at least two preset weather conditions; wherein the reference weather data sets include historical meteorological characteristics of the reference photovoltaic power station; Determining a power spatial feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station; For each preset weather condition, a pre-trained weather power prediction model corresponding to the preset weather condition is obtained, and a standard power sequence of the photovoltaic power station under test under the preset weather condition is determined based on the historical power data to be measured of the photovoltaic power station under test, the power spatial feature vector, a reference weather data set corresponding to the preset weather condition, and the weather power prediction model; wherein the standard power sequence includes at least one standard power corresponding to each future moment; For each future moment, the weather to be measured of the photovoltaic power station to be measured at the future moment is obtained, and the target power of the photovoltaic power station to be measured at the future moment is obtained from the standard power sequence corresponding to the weather to be measured.

2. The method according to claim 1, characterized in that The model architecture of the spatial feature extraction model is a graph convolutional neural network. Accordingly, based on the pre-trained spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station, the power spatial feature vector of the photovoltaic power station to be tested is determined, including: Constructing a meteorological characteristic matrix based on reference historical meteorological data corresponding to at least one reference photovoltaic power station; wherein the meteorological characteristic matrix includes a meteorological block matrix corresponding to each reference photovoltaic power station; The power adjacent matrix and the meteorological characteristic matrix are input into the spatial feature extraction model to obtain an output power spatial characteristic vector of the photovoltaic power station to be tested.

3. The method according to claim 1, characterized in that The model architecture of the weather power prediction model is a bidirectional long short-term memory network. Accordingly, the method of determining the standard power sequence of the photovoltaic power station under the preset weather conditions based on the historical power data of the photovoltaic power station under test, the power space feature vector, the reference weather data set corresponding to the preset weather conditions, and the weather power prediction model includes: Performing modal decomposition on a target feature sequence consisting of the historical power data to be measured, the power space feature vector, and the reference weather data set to obtain at least two time series feature sequences; The at least two time series feature sequences are input into the weather power prediction model to obtain an output of a standard power sequence of the photovoltaic power station to be tested under the preset weather conditions.

4. The method according to claim 1, wherein The obtaining of power adjacency matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station includes: Determining the power covariance corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station according to the historical power data to be tested of the photovoltaic power station to be tested and the reference historical power data of the reference photovoltaic power station; Determining a standard deviation of the power to be measured corresponding to the photovoltaic power station to be measured based on the historical power data to be measured, and determining a standard deviation of the reference power corresponding to the reference photovoltaic power station based on the reference historical power data; Determining power correlation coefficients corresponding to the photovoltaic power station to be measured and the reference photovoltaic power station according to the power covariance, the standard deviation of the power to be measured, and the standard deviation of the reference power; The power correlation coefficient is used as a matrix element in the power adjacency matrix corresponding to the photovoltaic power station to be tested and the reference photovoltaic power station.

5. The method according to claim 1, wherein The obtaining of reference weather data sets corresponding to at least two preset weather conditions includes: A hierarchical clustering algorithm is used to perform clustering operations on reference historical meteorological data corresponding to at least one reference photovoltaic power station, to obtain reference weather data sets corresponding to at least two preset weather conditions.

6. The method according to claim 1, wherein The meteorological time limit corresponding to the historical meteorological characteristics is one day, and the future time is a time within one day of a future date.

7. A power prediction device for a photovoltaic power station, characterized in that: include: A power adjacent matrix acquisition module is used to obtain power adjacent matrices corresponding to the photovoltaic power station to be tested and at least one reference photovoltaic power station; wherein the power adjacent matrix represents the power correlation between photovoltaic power stations; A reference weather data set determination module is configured to obtain reference weather data sets corresponding to at least two preset weather conditions; wherein the reference weather data sets include historical meteorological characteristics of the reference photovoltaic power station; a power space feature vector determination module, configured to determine the power space feature vector of the photovoltaic power station to be tested based on a pre-trained spatial feature extraction model, the power adjacency matrix, and reference historical meteorological data corresponding to at least one reference photovoltaic power station; a standard power sequence determination module, configured to obtain, for each preset weather condition, a pre-trained weather power prediction model corresponding to the preset weather condition, and determine, based on the historical power data to be measured of the photovoltaic power station to be tested, the power spatial feature vector, a reference weather data set corresponding to the preset weather condition, and the weather power prediction model, a standard power sequence for the photovoltaic power station to be tested under the preset weather condition; wherein the standard power sequence includes at least one standard power corresponding to each future moment; The target power determination module is used to obtain the weather to be tested of the photovoltaic power station to be tested at each future moment, and obtain the target power of the photovoltaic power station to be tested at the future moment from the standard power sequence corresponding to the weather to be tested.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor. The computer program is executed by the at least one processor so that the at least one processor can execute the power prediction method for a photovoltaic power station according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power prediction method for a photovoltaic power station according to any one of claims 1 to 6 when executed. 10 . A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the power prediction method for a photovoltaic power station according to any one of claims 1 to 6.

Images