Photovoltaic power station generating capacity prediction method and system combined with dynamic updating mechanism

Abstract

The invention discloses a photovoltaic power station generating capacity prediction method and system combined with a dynamic updating mechanism, and relates to the related field of photovoltaic power generation, and the method comprises the steps: collecting the power generation data of a photovoltaic module array in a target photovoltaic power station in a preset historical window; performing iterative data fluctuation analysis on the power generation data array sequence according to a time sequence; determining a power generation capacity prediction window according to the size of the data fluctuation factor, traversing and extracting power generation data of the photovoltaic module array in the power generation capacity prediction window, and obtaining a prediction window power generation data array sequence; performing centralized power generation data screening on the prediction window power generation data array sequence, and determining a target prediction window power generation data array; and calling a generating capacity prediction network layer to identify the target prediction window generating data array to obtain a generating capacity prediction result. The technical problem of low prediction precision of the existing photovoltaic power station generating capacity prediction is solved, and the technical effect of improving the photovoltaic power station generating capacity prediction precision is achieved.

Description

Technical Field

[0001] This application relates to the field of photovoltaic power generation, and particularly to a method and system for predicting the power generation of a photovoltaic power station combined with a dynamic update mechanism. Background Art

[0002] In the operation and management of a photovoltaic power station, accurately predicting the power generation is crucial for optimizing the operation of the power station and improving the energy utilization efficiency. At present, the power generation prediction of a photovoltaic power station mainly adopts methods based on statistical learning. These methods use historical power generation data and environmental factor data, and through statistical means such as regression analysis and time series analysis to predict the power generation. Their prediction performance depends on a large amount of historical data and stable environmental factors. Since the power generation process of a photovoltaic power station is affected by various complex factors, it is difficult for these methods to accurately capture the change laws of these factors when dealing with dynamically changing power generation data, resulting in low prediction accuracy.

[0003] In the related technologies at the present stage, there is a technical problem of low prediction accuracy in the power generation prediction of a photovoltaic power station. Summary of the Invention

[0004] This application provides a method and system for predicting the power generation of a photovoltaic power station combined with a dynamic update mechanism. By collecting the power generation data of the photovoltaic module array in a target photovoltaic power station within a preset historical window to obtain a power generation data array sequence, performing iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor, determining a power generation prediction window according to the size of the data fluctuation factor, traversing and extracting the power generation data of the photovoltaic module array within the power generation prediction window to obtain a prediction window power generation data array sequence, performing centralized power generation data screening on the prediction window power generation data array sequence to determine a target prediction window power generation data array, and calling a power generation prediction network layer to identify the target prediction window power generation data array to obtain a power generation prediction result and other technical means, the technical effect of improving the power generation prediction accuracy of a photovoltaic power station is achieved.

[0005] This application provides a method for predicting the power generation of a photovoltaic power station combined with a dynamic update mechanism, including: collecting the power generation data of the photovoltaic module array in a target photovoltaic power station within a preset historical window to obtain a power generation data array sequence; performing iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor; determining a power generation prediction window according to the size of the data fluctuation factor, traversing and extracting the power generation data of the photovoltaic module array within the power generation prediction window to obtain a prediction window power generation data array sequence; performing centralized power generation data screening on the prediction window power generation data array sequence to determine a target prediction window power generation data array; and calling a power generation prediction network layer to identify the target prediction window power generation data array to obtain a power generation prediction result.

[0006] In a possible implementation, iterative data fluctuation analysis is performed on the sequence of power generation data arrays in chronological order to obtain a data fluctuation factor, and the following processing is executed: extract a first power generation data array and a second power generation data array from the sequence of power generation data arrays in chronological order; perform iterative correlation on the first power generation data array and the second power generation data array to obtain a first interaction correlation matrix and a first interaction correlation power generation data array; perform iterative correlation on the first interaction correlation power generation data array and a third power generation data array extracted from the sequence of power generation data arrays in chronological order to obtain a second interaction correlation matrix and a second interaction correlation power generation data array; perform iterative correlation on the (N - 2)th interaction correlation power generation data array and the Nth power generation data array extracted from the sequence of power generation data arrays in chronological order to obtain the (N - 1)th interaction correlation matrix, where N is the number of power generation data arrays in the sequence of power generation data arrays; call an iterative data fluctuation analysis network layer to perform fluctuation analysis on the first interaction correlation matrix, the second interaction correlation matrix, the (N - 2)th interaction correlation matrix, and the (N - 1)th interaction correlation matrix to obtain the data fluctuation factor.

[0007] In a possible implementation, the following processing is executed: perform one-to-one data similarity identification on the first power generation data array and the second power generation data array to obtain a first similarity array; perform normalization processing on the first similarity array using a normalization function and matrixize the processing result to obtain a first interaction correlation matrix; use a convolutional network to calculate the second power generation data array and the first interaction correlation matrix to obtain a first interaction correlation power generation data array.

[0008] In a possible implementation, the following processing is executed: obtain a normalization function, where the normalization function is:

[0009]

[0010] where SFT[lim(x i , y i )] is the normalized value corresponding to the ith first similarity in the first similarity array, e is the base of the natural logarithm, n is the total number of first similarities in the first similarity array, x i is the ith first power generation data in the first power generation data array, y i is the ith second power generation data in the second power generation data array, and lim(x i , y i ) is the ith first similarity in the first similarity array.

[0011] In a possible implementation, centralized power generation data screening is performed on the predicted window power generation data array sequence to determine the target predicted window power generation data array, and the following processing is executed: randomly extract the first predicted window power generation data array from the predicted window power generation data array sequence and use it as the first screening center; construct the first screening neighborhood of the first screening center in the predicted window power generation data array sequence according to a preset similarity threshold; use the screening center iteration function to iteratively update the first screening center in the first screening neighborhood to obtain an iterative screening center; again, construct the iterative screening neighborhood of the iterative screening center in the predicted window power generation data array sequence according to the preset similarity threshold; again, use the screening center iteration function to iteratively update the iterative screening center in the iterative screening neighborhood to obtain an updated iterative screening center; and so on, until the similarity between two iterative screening centers obtained from two adjacent iterations is greater than or equal to the preset iterative similarity, stop the iteration, and use the predicted window power generation data array corresponding to the iterative screening center obtained in the last iteration as the target predicted window power generation data array.

[0012] In a possible implementation, the following processing is executed: The screening center iteration function is:

[0013]

[0014] where Point(z) is the iterative screening center, N(z) is the first screening neighborhood, z q is the q-th predicted window power generation data array in the first screening neighborhood, z is the first screening center, and K(z q - z) is the weight kernel function constructed based on the Gaussian function.

[0015] In a possible implementation, the power generation prediction network layer is called to identify the target predicted window power generation data array to obtain a power generation prediction result, and the following processing is executed: obtain multiple sample predicted window power generation data arrays and multiple sample power generation prediction results as training data; use the training data to perform supervised training on a framework constructed based on a convolutional neural network to learn the one-to-one mapping relationship between the predicted window power generation data array and the sample power generation prediction result until the training converges, and obtain the trained power generation prediction network layer.

[0016] The present application also provides a photovoltaic power station power generation prediction system combined with a dynamic update mechanism, including: a historical power generation data acquisition module for acquiring the power generation data of the photovoltaic module array in a target photovoltaic power station within a preset historical window to obtain a power generation data array sequence; a data fluctuation analysis module for performing iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor; a predicted power generation data extraction module for determining a power generation prediction window according to the magnitude of the data fluctuation factor, traversing and extracting the power generation data of the photovoltaic module array within the power generation prediction window to obtain a predicted window power generation data array sequence; a centralized screening module for centrally screening the power generation data of the predicted window power generation data array sequence to determine a target predicted window power generation data array; and a power generation prediction module for calling a power generation prediction network layer to identify the target predicted window power generation data array to obtain a power generation prediction result.

[0017] It is intended to first acquire the power generation data of the photovoltaic module array in a target photovoltaic power station within a preset historical window through the photovoltaic power station power generation prediction method and system combined with a dynamic update mechanism proposed in the present application to obtain a power generation data array sequence, then perform iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor, then determine a power generation prediction window according to the magnitude of the data fluctuation factor, traverse and extract the power generation data of the photovoltaic module array within the power generation prediction window to obtain a predicted window power generation data array sequence, then centrally screen the power generation data of the predicted window power generation data array sequence to determine a target predicted window power generation data array, and finally call a power generation prediction network layer to identify the target predicted window power generation data array to obtain a power generation prediction result, achieving the technical effect of improving the prediction accuracy of the photovoltaic power station power generation. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention will be briefly introduced below. Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the operations in the front or below do not necessarily need to be executed precisely in sequence. On the contrary, according to needs, they can be executed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or several operations can be removed from these processes.

[0019] Figure 1 It is a schematic flowchart of the photovoltaic power station power generation prediction method combined with a dynamic update mechanism provided by the embodiment of the present application.

[0020] Figure 2 It is a schematic structural diagram of the photovoltaic power station power generation prediction system combined with a dynamic update mechanism provided by the embodiment of the present application.

[0021] Explanation of the reference numerals: historical power generation data acquisition module 10 , data fluctuation analysis module 20 , predicted power generation data extraction module 30 , centralized screening module 40 , power generation prediction module 50 . DETAILED DESCRIPTION

[0022] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0023] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0024] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server 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 modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0025] The present application embodiment provides a photovoltaic power station power generation prediction method combined with a dynamic update mechanism, such as Figure 1 As shown, the method includes:

[0026] Step S100 , collecting power generation data of a photovoltaic module array in a target photovoltaic power station within a preset historical window to obtain a power generation data array sequence.

[0027] Specifically, determine the photovoltaic module array within the target photovoltaic power station. The photovoltaic module array is a collection composed of multiple photovoltaic modules (such as solar panels) and is used to convert light energy into electrical energy. Set a preset historical window, which is a time range used to specify when to start and end collecting power generation data. Through the data acquisition system of the photovoltaic power station (such as the SCADA system), automatically collect the power generation data within this historical window to form a power generation data array sequence. This data includes, but is not limited to, the power generation amount, power generation time, light intensity, temperature, etc. of each photovoltaic module or array.

[0028] Step S200, perform iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor.

[0029] Specifically, for the obtained power generation data array sequence, perform iterative analysis in chronological order. Calculate the fluctuation situation (such as change rate, standard deviation, etc.) between adjacent data points in each iteration, thereby obtaining a data fluctuation factor. The data fluctuation factor reflects the degree of fluctuation of the power generation data over time.

[0030] In a possible implementation, perform iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor. Step S200 further includes step S210, extract the first power generation data array and the second power generation data array from the power generation data array sequence in chronological order. Specifically, first extract two consecutive power generation data arrays from the power generation data array sequence in chronological order, and name them the first power generation data array and the second power generation data array respectively. These two data arrays represent the power generation data within two adjacent time periods in the time series.

[0031] Step S220, perform iterative correlation on the first power generation data array and the second power generation data array to obtain a first interaction correlation matrix and a first interaction correlation power generation data array. Specifically, perform iterative correlation analysis on the first power generation data array and the second power generation data array. This analysis can be carried out based on correlation, covariance, or other statistical metrics to quantify the degree of interaction correlation between the two data arrays. The result of the analysis is an interaction correlation matrix (the first interaction correlation matrix), which reflects the correlation strength between the two data arrays. The lower the degree of data fluctuation, the deeper the degree of interaction correlation between the data. At the same time, based on this correlation analysis, generate a new power generation data array (the first interaction correlation power generation data array), which integrates the information of the original data arrays and emphasizes their correlation.

[0032] Step S230: Iteratively correlate the first interactive correlation power generation data array and the third power generation data array extracted from the power generation data array sequence in chronological order to obtain a second interactive correlation matrix and a second interactive correlation power generation data array; Step S240: Iteratively correlate the (N-2)th interactive correlation power generation data array and the Nth power generation data array extracted from the power generation data array sequence in chronological order to obtain the (N-1)th interactive correlation matrix, where N is the number of power generation data arrays in the power generation data array sequence.

[0033] Specifically, Steps S230 to S240 are iterative expansions of Step S220. In S230, the first interactive correlation power generation data array is iteratively correlated with the third power generation data array extracted in chronological order to generate a second interactive correlation matrix and a second interactive correlation power generation data array. This process is iterated until the entire power generation data array sequence is processed. In S240, the (N-2)th interactive correlation power generation data array and the Nth power generation data array are iteratively correlated to obtain the (N-1)th interactive correlation matrix. In this way, a series of interactive correlation matrices are obtained, which reflect the correlation degree of data in each time period of the entire power generation data array sequence.

[0034] Step S250: Invoke the iterative data fluctuation analysis network layer to perform fluctuation analysis on the first interactive correlation matrix, the second interactive correlation matrix, the (N-2)th interactive correlation matrix, and the (N-1)th interactive correlation matrix to obtain the data fluctuation factor. Specifically, the iterative data fluctuation analysis network layer is invoked to process the sequence of interactive correlation matrices obtained in Steps S220 to S240. This network layer is a deep learning model that has been trained to identify the fluctuation patterns in the matrices. The iterative data fluctuation analysis network layer analyzes the input interactive correlation matrix and outputs a data fluctuation factor, which quantifies the fluctuation degree of the entire power generation data array sequence. This implementation method can capture the interactive correlation degree between data through iterative correlation analysis, and this correlation degree reflects the internal connection and stability between data. By processing these correlation matrices through the iterative data fluctuation analysis network layer, the fluctuation degree of the entire data sequence is accurately quantified. This fine quantification allows the size of the prediction window to be adjusted according to the actual fluctuation of the data, thereby improving the accuracy and stability of the prediction.

[0035] In a possible implementation, step S220 further includes step S221 of performing one-to-one data similarity identification on the first power generation data array and the second power generation data array to obtain a first similarity array. Specifically, a similarity calculation method is used to perform one-to-one data similarity identification on the first power generation data array and the second power generation data array, that is, similarity calculation is performed on the data points at corresponding positions in the two data arrays. The similarity can be calculated based on various distance metrics (such as Euclidean distance, Manhattan distance, cosine similarity, etc.) to quantify the similarity between the corresponding data points in the two data arrays, generating a similarity array, where each element represents the similarity of the data points at the corresponding positions in the first power generation data array and the second power generation data array.

[0036] Step S222 of normalizing the first similarity array using a normalization function and matrixifying the processing result to obtain a first interaction correlation matrix. Specifically, after obtaining the similarity array, a normalization function is used to process it to convert the similarity values to a unified scale for subsequent calculation and analysis. Among them, the normalization function can be linear normalization, logarithmic normalization, min-max normalization, etc. The processed result is converted into a matrix form, that is, the first interaction correlation matrix, which reflects the interaction correlation degree between the first power generation data array and the second power generation data array.

[0037] Step S223 of calculating the second power generation data array and the first interaction correlation matrix using a convolutional network to obtain a first interaction correlation power generation data array. Specifically, a convolutional neural network (CNN) is used to process the second power generation data array and the first interaction correlation matrix. CNN is a deep learning model. When applied to a two-dimensional data array, CNN extracts features, learns patterns, and makes predictions through structures such as convolutional layers, pooling layers, and fully connected layers. The second power generation data array and the first interaction correlation matrix are used as the input of CNN, and through the trained network for calculation, a first interaction correlation power generation data array is output. This new data array integrates the information of the second power generation data array and the correlation information of the first interaction correlation matrix. This implementation method can capture the local similarity between data points through one-to-one data similarity identification. The normalization process ensures that the similarity values are on a unified scale, facilitating subsequent analysis. Using a convolutional network for calculation can extract and utilize these similarities and correlation information to generate a new data array that integrates more useful features. This processing process not only improves the richness of data representation but also provides more accurate and comprehensive input data for subsequent power generation prediction, ultimately contributing to improving the accuracy and stability of power generation prediction.

[0038] In a possible implementation, step S222 further includes step S2221 of obtaining a normalization function, where the normalization function is:

[0039]

[0040] where SFT[lim(x i , y i )] is the normalized value corresponding to the i-th first similarity in the first similarity array, e is the base of the natural logarithm, n is the total number of first similarities in the first similarity array, x i is the i-th first power generation data in the first power generation data array, y i is the i-th second power generation data in the second power generation data array, and lim(x i , y i ) is the i-th first similarity in the first similarity array.

[0041] Specifically, the normalization function is used to convert each similarity value in the first similarity array into a unified scale between 0 and 1. lim(x i , y i ) is an element in the first similarity array, representing the degree of similarity between two data points x i and y i .

[0042] Step S300: Determine a power generation prediction window according to the magnitude of the data fluctuation factor, and traverse and extract the power generation data of the photovoltaic module array in the power generation prediction window to obtain a sequence of prediction window power generation data arrays.

[0043] Specifically, according to the magnitude of the obtained data fluctuation factor, dynamically determine the window size for the next power generation prediction. If the data fluctuation factor is large, it indicates that the power generation data is relatively unstable, so a shorter prediction window should be selected; conversely, if the data fluctuation factor is small, it indicates that the power generation data is relatively stable, and a longer prediction window can be selected. Within the determined prediction window, traverse and extract the power generation data of the photovoltaic module array to form a sequence of prediction window power generation data arrays.

[0044] Step S400: Perform centralized power generation data screening on the sequence of prediction window power generation data arrays to determine a target prediction window power generation data array.

[0045] Specifically, use data preprocessing techniques to screen the obtained sequence of prediction window power generation data arrays, remove data points that do not meet the requirements such as outliers and missing values, ensure the quality of the data used for prediction, and the screened data forms a target prediction window power generation data array.

[0046] In a possible implementation, centralized power generation data screening is performed on the predicted window power generation data array sequence to determine the target predicted window power generation data array. Step S400 further includes step S410 of randomly extracting a first predicted window power generation data array from the predicted window power generation data array sequence and using it as the first screening center. Specifically, a random number generator is used to randomly select a subsequence from the predicted window power generation data array sequence as the initial screening center (the first screening center), and the length of this subsequence is the same as the size of the predicted window.

[0047] Step S420, construct a first screening neighborhood of the first screening center in the predicted window power generation data array sequence according to a preset similarity threshold. Specifically, a similarity measurement method (such as Euclidean distance, cosine similarity, etc.) is used to calculate the similarity between each data array in the predicted window power generation data array sequence and the first screening center, and data arrays with a similarity greater than or equal to the preset threshold are screened out according to the preset similarity threshold (the similarity criterion for determining whether a data array belongs to the screening neighborhood) to form the first screening neighborhood.

[0048] Step S430, use the screening center iteration function to iteratively update the first screening center in the first screening neighborhood to obtain an iterative screening center. Specifically, the screening center iteration function (the function for updating the screening center) is used to update the first screening center within the first screening neighborhood, that is, according to the definition of the iteration function (such as weighted average, median, etc.), a certain statistic of the data arrays in the first screening neighborhood is calculated, and the data array corresponding to this statistic is used as the new iterative screening center.

[0049] Step S440, construct an iterative screening neighborhood of the iterative screening center in the predicted window power generation data array sequence again according to the preset similarity threshold; Step S450, use the screening center iteration function to iteratively update the iterative screening center in the iterative screening neighborhood to obtain an updated iterative screening center.

[0050] Specifically, repeat steps S420 and S430, but perform them on the new iterative screening center and its iterative screening neighborhood, that is, use the iterative screening center obtained in the previous iteration as the new screening center, recalculate the similarity and construct a new screening neighborhood, and then use the iteration function to update the screening center again.

[0051] Step S460, and so on, until the similarity between two iterative screening centers obtained in two adjacent iterations is greater than or equal to a preset iterative similarity, then stop the iteration, and use the power generation data array of the prediction window corresponding to the iterative screening center obtained in the last iteration as the target power generation data array of the prediction window. Specifically, after each iteration, calculate the similarity between the iterative screening centers of two adjacent iterations, and check whether the condition for stopping the iteration is met. When the similarity between the iterative screening centers obtained in two adjacent iterations is greater than or equal to the preset iterative similarity (the similarity criterion for determining when to stop the iteration), stop the iteration, and use the data array corresponding to the iterative screening center obtained in the last iteration as the target power generation data array of the prediction window. This implementation method randomly initializes the screening center, constructs a screening neighborhood according to the similarity threshold in each iteration, and then updates the screening center using an iterative function, which can gradually approach a more representative and stable data array as the target power generation data array of the prediction window. This method reduces the influence of noise and outliers on the prediction result and improves the accuracy and robustness of the prediction.

[0052] In a possible implementation, step S430 further includes step S431, and the screening center iterative function is:

[0053]

[0054] where Point(z) is the iterative screening center, N(z) is the first screening neighborhood, z q is the q-th power generation data array of the prediction window in the first screening neighborhood, z is the first screening center, and K(z q -z) is the weight kernel function constructed based on the Gaussian function.

[0055] Specifically, the weight kernel function constructed based on the Gaussian function calculates the weight according to a certain distance (such as the Euclidean distance) between the data array and the first screening center. The larger the weight, the more similar the data array is to the first screening center, and the greater the contribution to updating the screening center. For each data array in the first screening neighborhood, calculate its distance from the first screening center, and use the weight kernel function to calculate the corresponding weight. Sum the products of each data array in the first screening neighborhood and its corresponding weight to obtain the weighted sum. Divide the weighted sum by the total sum of the weights (i.e., normalization) to obtain the new screening center. This new screening center is the iterative screening center. This implementation method uses the weight kernel function constructed based on the Gaussian function to update the screening center. The Gaussian function has smoothness, which can reduce the influence of noise and outliers on the screening center, making the updated screening center more stable, ensuring that the selection of the screening center is gradually optimized during the iteration process, thus obtaining a more accurate and stable target power generation data array of the prediction window, and further improving the accuracy and reliability of the photovoltaic power station power generation prediction.

[0056] Step S500: Invoke the power generation prediction network layer to identify the power generation data array of the target prediction window, and obtain the power generation prediction result.

[0057] Specifically, machine learning or deep learning techniques are adopted. First, a power generation prediction network layer (such as a neural network, support vector machine, etc.) is constructed. This network layer has been trained to be able to identify power generation data and predict power generation. Then, the obtained power generation data array of the target prediction window is input into this network layer for identification, and the power generation prediction result is output. In the embodiment of the present application, the power generation data of the photovoltaic module array in the target photovoltaic power station within the preset historical window is collected to obtain a sequence of power generation data arrays. The iterative data fluctuation analysis is performed on the sequence of power generation data arrays in chronological order to obtain the data fluctuation factor. The power generation prediction window is determined according to the size of the data fluctuation factor. The power generation data of the photovoltaic module array is extracted by traversing within the power generation prediction window to obtain a sequence of power generation data arrays of the prediction window. The centralized power generation data screening is performed on the sequence of power generation data arrays of the prediction window to determine the power generation data array of the target prediction window. The power generation prediction network layer is invoked to identify the power generation data array of the target prediction window, and the power generation prediction result is obtained, etc. Technical means have achieved the technical effect of improving the power generation prediction accuracy of the photovoltaic power station.

[0058] In a possible implementation manner, when invoking the power generation prediction network layer to identify the power generation data array of the target prediction window and obtain the power generation prediction result, step S500 further includes step S510: Obtain multiple sample power generation data arrays of prediction windows and multiple sample power generation prediction results as training data. Specifically, from the historical power generation data, the power generation data is extracted according to the power generation prediction window to form multiple sample power generation data arrays of prediction windows. Each sample power generation data array of the prediction window contains the power generation data of the photovoltaic module array within a certain specific time period. The sample power generation prediction result is the actual power generation data corresponding to the sample power generation data array of the prediction window. The sample power generation data array of the prediction window and the corresponding sample power generation prediction result are used to train the power generation prediction network layer.

[0059] Step S520: Use the training data to perform supervised training on the framework constructed based on the convolutional neural network, and learn the one-to-one mapping relationship between the power generation data array of the prediction window and the predicted power generation amount of the sample, until the training converges, and obtain the trained power generation prediction network layer. Specifically, design a CNN model, which includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer, etc. The input layer receives the sample prediction window power generation data array, and the output layer outputs the predicted power generation amount. The mean square error (MSE) of the loss function is used to measure the difference between the model prediction result and the actual sample power generation prediction result, and the stochastic gradient descent (SGD) is used to minimize the loss function, so as to update the weight parameters of the CNN model. Batch the training data and input it into the CNN model, calculate the prediction result through forward propagation, calculate the gradient through backward propagation, and update the model weights. Repeat this process until the loss function converges or reaches the preset number of training epochs. This implementation method adopts a supervised training method based on CNN. CNN has strong feature extraction capabilities and can automatically learn useful features from power generation data. Through supervised training with a large amount of sample data, the model can learn the complex mapping relationship between power generation data and power generation amount, so as to accurately predict new data.

[0060] In the above, with reference to Figure 1 The photovoltaic power station power generation prediction method combined with the dynamic update mechanism according to the embodiment of the present invention is described in detail. Next, with reference to Figure 2 Describe the photovoltaic power station power generation prediction system combined with the dynamic update mechanism according to the embodiment of the present invention.

[0061] The photovoltaic power station power generation prediction system combined with the dynamic update mechanism according to the embodiment of the present invention is used to solve the technical problem of low prediction accuracy existing in the prior art, and achieve the technical effect of improving the prediction accuracy of the photovoltaic power station power generation amount. The photovoltaic power station power generation prediction system combined with the dynamic update mechanism includes: a historical power generation data acquisition module 10, a data fluctuation analysis module 20, a predicted power generation data extraction module 30, a centralized screening module 40, and a power generation amount prediction module 50.

[0062] The historical power generation data acquisition module 10 is used to acquire the power generation data of the photovoltaic module array in the target photovoltaic power station within a preset historical window, and obtain a power generation data array sequence; the data fluctuation analysis module 20 is used to perform iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor; the predicted power generation data extraction module 30 is used to determine a power generation prediction window according to the magnitude of the data fluctuation factor, and traverse and extract the power generation data of the photovoltaic module array within the power generation prediction window to obtain a predicted window power generation data array sequence; the centralized screening module 40 is used to perform centralized power generation data screening on the predicted window power generation data array sequence to determine a target predicted window power generation data array; the power generation prediction module 50 is used to call the power generation prediction network layer to identify the target predicted window power generation data array to obtain a power generation prediction result.

[0063] Next, the specific configuration of the data fluctuation analysis module 20 will be described in detail. As described above, iterative data fluctuation analysis is performed on the power generation data array sequence in chronological order to obtain a data fluctuation factor. The data fluctuation analysis module 20 may further include: a power generation data array extraction unit for extracting a first power generation data array and a second power generation data array from the power generation data array sequence in chronological order; an iterative association unit for performing iterative association on the first power generation data array and the second power generation data array to obtain a first interaction association matrix and a first interaction association power generation data array, performing iterative association on the first interaction association power generation data array and a third power generation data array extracted from the power generation data array sequence in chronological order to obtain a second interaction association matrix and a second interaction association power generation data array, and performing iterative association on the N-2th interaction association power generation data array and the Nth power generation data array extracted from the power generation data array sequence in chronological order to obtain the N-1th interaction association matrix, where N is the number of power generation data arrays in the power generation data array sequence; a fluctuation analysis unit for calling an iterative data fluctuation analysis network layer to perform fluctuation analysis on the first interaction association matrix, the second interaction association matrix, the N-2th interaction association matrix, and the N-1th interaction association matrix to obtain the data fluctuation factor.

[0064] Among them, the iterative association unit may further include: a data similarity recognition subunit for performing one-to-one data similarity recognition on the first power generation data array and the second power generation data array to obtain a first similarity array; a normalization processing subunit for performing normalization processing on the first similarity array using a normalization function and matrixifying the processing result to obtain a first interaction association matrix; a calculation subunit for using a convolutional network to calculate the second power generation data array and the first interaction association matrix to obtain a first interaction association power generation data array.

[0065] Among them, the normalization processing subunit may further include: a normalization function obtaining component for obtaining a normalization function, where the normalization function is:

[0066]

[0067] Among them, SFT[lim(x i , y i )] is the normalized value corresponding to the i-th first similarity in the first similarity array, e is the base of the natural logarithm, n is the total number of first similarities in the first similarity array, x i is the i-th first power generation data in the first power generation data array, y i is the i-th second power generation data in the second power generation data array, lim(x i , y i ) is the i-th first similarity in the first similarity array.

[0068] Next, the specific configuration of the centralized screening module 40 will be described in detail. As described above, for the sequence of predicted window power generation data arrays, centralized power generation data screening is performed to determine the target predicted window power generation data array. The centralized screening module 40 may further include: a first screening center obtaining unit for randomly extracting a first predicted window power generation data array from the sequence of predicted window power generation data arrays and using it as the first screening center; a first screening neighborhood constructing unit for constructing a first screening neighborhood of the first screening center in the sequence of predicted window power generation data arrays according to a preset similarity threshold; a screening center iteration unit for iteratively updating the first screening center in the first screening neighborhood by using a screening center iteration function to obtain an iterative screening center, and then constructing an iterative screening neighborhood of the iterative screening center in the sequence of predicted window power generation data arrays again according to the preset similarity threshold, and then using the screening center iteration function to iteratively update the iterative screening center in the iterative screening neighborhood to obtain an updated iterative screening center, and so on, until the similarity between two iterative screening centers obtained by two adjacent iterations is greater than or equal to a preset iterative similarity, at which point the iteration stops, and the predicted window power generation data array corresponding to the iterative screening center obtained in the last iteration is used as the target predicted window power generation data array.

[0069] Among them, the screening center iteration unit may further include: a screening center iteration function obtaining subunit for obtaining a screening center iteration function, and the screening center iteration function is:

[0070]

[0071] Among them, Point(z) is the iterative screening center, N(z) is the first screening neighborhood, z qFor the power generation data array of the q-th prediction window in the first screening neighborhood, z is the first screening center, and K(z q -z) is a weight kernel function constructed based on the Gaussian function.

[0072] Next, the specific configuration of the power generation prediction module 50 will be described in detail. As described above, the power generation prediction network layer is called to identify the target prediction window power generation data array to obtain the power generation prediction result. The power generation prediction module 50 may further include: a training data acquisition unit for acquiring a plurality of sample prediction window power generation data arrays and a plurality of sample power generation prediction results as training data; a power generation prediction network layer acquisition unit for using the training data to perform supervised training on a framework constructed based on a convolutional neural network, learning the one-to-one mapping relationship between the prediction window power generation data array and the sample power generation prediction result until the training converges, and obtaining the trained power generation prediction network layer.

[0073] The photovoltaic power station power generation prediction system combining a dynamic update mechanism provided by an embodiment of the present invention can execute the photovoltaic power station power generation prediction method combining a dynamic update mechanism provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method.

[0074] Although various references are made to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or the server. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0075] The above specific embodiments do not constitute a limitation to the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be executed in a different order from that in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A photovoltaic power plant power generation prediction method combined with a dynamic update mechanism, characterized in that: The method comprises: Collecting power generation data of the photovoltaic module array in the target photovoltaic power station within a preset historical window to obtain a power generation data array sequence; Performing iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor; Determine a power generation prediction window according to the size of the data fluctuation factor, traverse and extract power generation data of the photovoltaic component array in the power generation prediction window, and obtain a power generation data array sequence of the prediction window; Centrally screening the power generation data of the prediction window power generation data array sequence to determine the target prediction window power generation data array; The power generation prediction network layer is called to identify the power generation data array of the target prediction window to obtain the power generation prediction result.

2. The photovoltaic power station power generation prediction method combined with a dynamic update mechanism as claimed in claim 1, characterized in that: Iterative data fluctuation analysis is performed on the power generation data array sequence in chronological order to obtain data fluctuation factors, including: Extracting a first power generation data array and a second power generation data array from the power generation data array sequence in chronological order; Iteratively correlating the first power generation data array and the second power generation data array to obtain a first interactive correlation matrix and a first interactively correlated power generation data array; Iteratively correlating the first interactively correlated power generation data array and a third power generation data array extracted in chronological order from the power generation data array sequence to obtain a second interactively correlated matrix and a second interactively correlated power generation data array; Iteratively correlating the N-2th interactively correlated power generation data array with the Nth power generation data array extracted in chronological order from the power generation data array sequence to obtain the N-1th interactively correlated matrix, wherein N is the number of power generation data arrays in the power generation data array sequence; The iterative data fluctuation analysis network layer is called to perform fluctuation analysis on the first interaction correlation matrix, the second interaction correlation matrix, the N-2th interaction correlation matrix and the N-1th interaction correlation matrix to obtain the data fluctuation factor.

3. The photovoltaic power station power generation prediction method combined with a dynamic update mechanism as claimed in claim 2, characterized in that: include: Performing one-to-one data similarity identification on the first power generation data array and the second power generation data array to obtain a first similarity array; Performing normalization processing on the first similarity array using a normalization function, and matrixing the processing result to obtain a first interaction correlation matrix; The second power generation data array and the first interactive correlation matrix are calculated using a convolutional network to obtain a first interactive correlation power generation data array.

4. The photovoltaic power station power generation prediction method combined with a dynamic update mechanism as claimed in claim 3, characterized in that: include: Obtain a normalization function, wherein the normalization function is: Among them, SFT[lim(x i ,y i )] is the normalized value corresponding to the i-th first similarity in the first similarity array, e is the base of the natural logarithm, n is the total number of first similarities in the first similarity array, x i is the i-th first power generation data in the first power generation data array, y i is the i-th second power generation data in the second power generation data array, lim(x i ,y i ) is the i-th first similarity in the first similarity array.

5. The photovoltaic power plant power generation prediction method combined with a dynamic update mechanism as claimed in claim 1, characterized in that: Centrally screening the power generation data of the prediction window power generation data array sequence to determine the target prediction window power generation data array comprises: Randomly extracting a first prediction window power generation data array from the prediction window power generation data array sequence and using it as a first screening center; According to a preset similarity threshold, constructing a first screening neighborhood of the first screening center in the prediction window power generation data array sequence; Iteratively updating the first screening center in the first screening neighborhood using a screening center iteration function to obtain an iterative screening center; Again, according to the preset similarity threshold, constructing an iterative screening neighborhood of the iterative screening center in the prediction window power generation data array sequence; Iteratively updating the iterative screening center in the iterative screening neighborhood using the screening center iteration function again to obtain an updated iterative screening center; And so on, until the similarity of two iteration screening centers obtained in two adjacent iterations is greater than or equal to the preset iteration similarity, the iteration is stopped, and the prediction window power generation data array corresponding to the iteration screening center obtained in the last iteration is used as the target prediction window power generation data array.

6. The photovoltaic power station power generation prediction method combined with a dynamic update mechanism as claimed in claim 5, characterized in that: The screening center iteration function is: Among them, Point(z) is the iterative screening center, N(z) is the first screening neighborhood, and z q is the power generation data array of the qth prediction window in the first screening neighborhood, z is the first screening center, K(z q -z) is a weight kernel function built based on the Gaussian function.

7. The photovoltaic power station power generation prediction method combined with a dynamic update mechanism as claimed in claim 1, characterized in that: The power generation prediction network layer is called to identify the power generation data array of the target prediction window to obtain the power generation prediction result, including: Acquire multiple sample prediction window power generation data arrays and multiple sample power generation prediction results as training data; The framework built based on the convolutional neural network is supervised and trained using the training data to learn the one-to-one mapping relationship between the prediction window power generation data array and the sample power generation prediction results until the training converges, thereby obtaining the trained power generation prediction network layer.

8. A photovoltaic power station power generation prediction system combined with a dynamic update mechanism is characterized in that: The system is used to implement the photovoltaic power station power generation prediction method combined with a dynamic update mechanism according to any one of claims 1 to 7, and the system includes: A historical power generation data acquisition module is used to collect power generation data of the photovoltaic module array in the target photovoltaic power station within a preset historical window to obtain a power generation data array sequence; A data fluctuation analysis module, used to perform iterative data fluctuation analysis on the power generation data array sequence in chronological order to obtain a data fluctuation factor; A predicted power generation data extraction module is used to determine a power generation prediction window according to the size of the data fluctuation factor, traverse and extract the power generation data of the photovoltaic component array in the power generation prediction window, and obtain a prediction window power generation data array sequence; A centralized screening module, used for performing centralized power generation data screening on the prediction window power generation data array sequence to determine a target prediction window power generation data array; The power generation prediction module is used to call the power generation prediction network layer to identify the power generation data array of the target prediction window to obtain the power generation prediction result.

Images