Wind power-photovoltaic power combined prediction method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a wind power-photovoltaic power combined prediction method and device, equipment, a medium and a program product, and relates to the technical field of new energy power prediction. According to the method, the accuracy, stability and generalization of wind power and photovoltaic power combined prediction can be improved. The method comprises the following steps: performing clustering analysis on large-scale power clusters in each wind power plant and a photovoltaic electric field by using a fuzzy clustering algorithm, and quantitatively representing spatial correlation between each wind power plant and the photovoltaic electric field according to a clustering result; establishing a heterogeneous graph adjacency matrix according to a clustering result and a node mode, and representing a heterogeneous topological structure of the multi-mode wind power-photovoltaic power generation data by using the heterogeneous graph adjacency matrix and the strong correlation characteristic factors; according to the heterogeneous topological structure, establishing a time sequence feature mining model and a spatial correlation feature mining model, and constructing a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network; and the wind power-photovoltaic power combined prediction model is used for predicting power values of each wind power plant and each photovoltaic electric field.

Description

Technical Field

[0001] The present application relates to the technical field of new energy power prediction, and in particular to a wind power-photovoltaic power joint prediction method, device, computer equipment, computer-readable storage medium and computer program product. Background Art

[0002] As energy and power systems reduce carbon and cleanliness in response to climate change and energy crises, wind power and photovoltaic power generation are widely used in energy transformation due to their clean and sustainable characteristics. Since wind power generation is affected by external driving factors such as wind direction, temperature and air pressure, and photovoltaic power generation depends on variables such as solar radiation, relative humidity and temperature, how to accurately predict the power of wind power generation and photovoltaic power generation has become an urgent problem to be solved.

[0003] However, as the scale of regional renewable energy power installed capacity continues to increase, traditional renewable energy power prediction technology is limited by the single-station prediction model, making it difficult to achieve highly accurate and stable renewable energy power prediction; and for the joint prediction tasks of multiple clusters in the region, the final constructed prediction model and parameter size may be too complex and difficult to debug, and cannot be expanded to many stations. Summary of the invention

[0004] Based on this, it is necessary to provide a wind power-photovoltaic power joint prediction method, device, computer equipment, computer readable storage medium and computer program product to address the above technical problems.

[0005] In a first aspect, the present application provides a wind power-photovoltaic power joint prediction method, comprising:

[0006] According to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the strong correlation characteristic factors of various prediction tasks are selected by using the maximum correlation-minimum redundancy principle;

[0007] Performing cluster analysis on the large-scale power clusters in each of the wind farms and the photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results, and quantitatively characterizing the spatial correlation between each of the wind farms and the photovoltaic farm based on the clustering results;

[0008] Establishing a heterogeneous graph adjacency matrix according to the clustering results and the node modes, and characterizing the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors;

[0009] According to the heterogeneous topological structure, a time series feature mining model and a spatial correlation feature mining model are established respectively, and based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each of the wind farms and the photovoltaic farms in the target area.

[0010] In one embodiment, before selecting the strongly relevant characteristic factors of various prediction tasks by using the maximum correlation-minimum redundancy principle, the method further includes:

[0011] According to the historical meteorological data, a plurality of historical meteorological characteristics are obtained;

[0012] The method of selecting strongly relevant feature factors for various prediction tasks using the maximum correlation-minimum redundancy principle includes:

[0013] According to the time series of the first meteorological feature and the wind power sequence of the target electric field, a first mutual information value between the first meteorological feature and the wind power sequence is determined; according to the time series of the first meteorological feature and the time series of the second meteorological feature, a second mutual information value between the first meteorological feature and the second meteorological feature is determined; the first meteorological feature and the second meteorological feature are any two different meteorological features among the multiple historical meteorological features; based on the first mutual information value and the second mutual information value, a maximum correlation criterion is used to select candidate characteristic factors that are strongly correlated with the wind power sequence, and a minimum redundancy criterion is used to remove redundant characteristic factors from the candidate characteristic factors to obtain the strongly correlated characteristic factors.

[0014] In one embodiment, the clustering analysis of the large-scale power clusters in each of the wind farms and the photovoltaic farms using a fuzzy clustering algorithm to obtain clustering results includes:

[0015] While calculating the site location information of each of the wind farms and the photovoltaic farms using the fuzzy clustering algorithm, the annual power time series of the large-scale power cluster is used as the characteristic attribute of the wind power-photovoltaic site for spatial clustering; the central site of each of the large-scale power clusters obtained by clustering is selected as a representative site, and a silhouette coefficient is established based on the average distance between the representative sites; the clustering result is evaluated using the average value of the silhouette coefficient.

[0016] In one embodiment, before selecting the strongly correlated characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle, the method further includes:

[0017] The historical power data is normalized to be within a preset range; abnormal data under normal operating conditions in the historical power data and the historical meteorological data are identified, and the abnormal data is interpolated and filled using the quartile method.

[0018] In one of the embodiments, after constructing the wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model, it also includes:

[0019] The improved Beluga optimization algorithm is used to determine the optimal hyperparameters of the deep network model, and the optimal hyperparameters are substituted into the heterogeneous graph neural network to obtain an updated wind power-photovoltaic power joint prediction model; the updated wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each of the wind farms and the photovoltaic farms in the target area to obtain a wind power-photovoltaic power prediction value.

[0020] In one of the embodiments, the wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network is constructed based on the time series feature mining model and the spatial correlation feature mining model, including:

[0021] The time series data is modeled using a temporal convolutional network, and the time series features of the topological nodes are extracted from the time series data; the time series features are used as input, and the spatial correlation between the stations is modeled based on a graph attention network to obtain the spatial correlation features; based on the time series features and the spatial correlation features, the wind power-photovoltaic power joint prediction model is constructed.

[0022] In a second aspect, the present application also provides a wind power-photovoltaic power joint prediction device, comprising:

[0023] A data processing module is used to select strongly correlated characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle;

[0024] A correlation characterization module, used to perform cluster analysis on the large-scale power clusters in each of the wind farms and the photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results, and quantitatively characterize the spatial correlation between each of the wind farms and the photovoltaic farm based on the clustering results;

[0025] A structure characterization module, used to establish a heterogeneous graph adjacency matrix according to the clustering results and node modes, and characterize the heterogeneous topological structure of multi-modal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors;

[0026] A model building module is used to establish a time series feature mining model and a spatial correlation feature mining model according to the heterogeneous topology structure, and to build a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each of the wind farms and the photovoltaic farms in the target area.

[0027] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0028] According to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the maximum correlation-minimum redundancy principle is used to select strongly correlated characteristic factors of various prediction tasks; a fuzzy clustering algorithm is used to perform clustering analysis on the large-scale power clusters in each wind farm and photovoltaic farm to obtain clustering results, and according to the clustering results, the spatial correlation between each wind farm and photovoltaic farm is quantitatively characterized; according to the clustering results and node modes, a heterogeneous graph adjacency matrix is ​​established, and based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors, the heterogeneous topological structure of multi-modal wind power-photovoltaic power generation data is characterized; according to the heterogeneous topological structure, a time series feature mining model and a spatial correlation feature mining model are established respectively, and based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed; the wind power-photovoltaic power joint prediction model is used to predict the power value of each wind farm and photovoltaic farm in the target area.

[0029] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0030] According to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the maximum correlation-minimum redundancy principle is used to select strongly correlated characteristic factors of various prediction tasks; a fuzzy clustering algorithm is used to perform clustering analysis on the large-scale power clusters in each wind farm and photovoltaic farm to obtain clustering results, and according to the clustering results, the spatial correlation between each wind farm and photovoltaic farm is quantitatively characterized; according to the clustering results and node modes, a heterogeneous graph adjacency matrix is ​​established, and based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors, the heterogeneous topological structure of multi-modal wind power-photovoltaic power generation data is characterized; according to the heterogeneous topological structure, a time series feature mining model and a spatial correlation feature mining model are established respectively, and based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed; the wind power-photovoltaic power joint prediction model is used to predict the power value of each wind farm and photovoltaic farm in the target area.

[0031] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0032] According to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the maximum correlation-minimum redundancy principle is used to select strongly correlated characteristic factors of various prediction tasks; a fuzzy clustering algorithm is used to perform clustering analysis on the large-scale power clusters in each wind farm and photovoltaic farm to obtain clustering results, and according to the clustering results, the spatial correlation between each wind farm and photovoltaic farm is quantitatively characterized; according to the clustering results and node modes, a heterogeneous graph adjacency matrix is ​​established, and based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors, the heterogeneous topological structure of multi-modal wind power-photovoltaic power generation data is characterized; according to the heterogeneous topological structure, a time series feature mining model and a spatial correlation feature mining model are established respectively, and based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed; the wind power-photovoltaic power joint prediction model is used to predict the power value of each wind farm and photovoltaic farm in the target area.

[0033] The above-mentioned wind power-photovoltaic power joint prediction method, device, computer equipment, computer-readable storage medium and computer program product, first, use the maximum correlation-minimum redundancy principle to select strongly correlated characteristic factors for various prediction tasks; secondly, based on the fuzzy clustering algorithm, multiple wind farms and photovoltaic power plant sites are clustered into different groups; in addition, facing the problem of regional multi-cluster station new energy power prediction, by deeply exploring the spatiotemporal correlation characteristics between different stations, a heterogeneous graph adjacency matrix is ​​established and combined with strongly correlated characteristic factors, the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data is characterized, and a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed, and an improved Beluga optimization algorithm is introduced to solve the optimal hyperparameters of the deep network model, further improving the prediction performance and generalization ability of the model, thereby simplifying the model and parameter size, effectively improving the accuracy, stability and generalization of the joint prediction of wind power and photovoltaic power, providing strong support for the stable operation and optimized scheduling of the power system, and can be expanded to many stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 This is an application environment diagram of a wind power-photovoltaic power joint prediction method in an embodiment;

[0036] Figure 2 A schematic diagram of a flow chart of a wind power-photovoltaic power joint prediction method in one embodiment;

[0037] Figure 3 A schematic diagram of an adjacency matrix of a wind power-photovoltaic heterogeneous graph in one embodiment;

[0038] Figure 4 A schematic diagram of a process of solving the optimal hyperparameter using the White Whale optimization algorithm in one embodiment;

[0039] Figure 5 A schematic diagram of a flow chart of a wind power-photovoltaic power joint prediction method in a specific embodiment;

[0040] Figure 6 A structural block diagram of a wind power-photovoltaic power joint prediction device in one embodiment;

[0041] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0043] The wind power-photovoltaic power joint prediction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal communicates with the server through the network. The data storage system can store the data that the server needs to process. The data storage system can be integrated on the server or placed on the cloud or other network servers.

[0044] Specifically, the wind power-photovoltaic power joint prediction method provided in the embodiment of the present application can be executed by a server.

[0045] Exemplarily, the server selects strongly correlated characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle; the server uses a fuzzy clustering algorithm to perform clustering analysis on the large-scale power clusters in each wind farm and photovoltaic farm to obtain clustering results, and quantitatively characterizes the spatial correlation between each wind farm and photovoltaic farm based on the clustering results; the server establishes a heterogeneous graph adjacency matrix based on the clustering results and node modes, and characterizes the heterogeneous topological structure of multimodal wind-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors; the server establishes a time series feature mining model and a spatial correlation feature mining model respectively according to the heterogeneous topological structure, and constructs a wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each wind farm and photovoltaic farm in the target area.

[0046] In such Figure 1 In the application environment shown, the terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, and tablet computers. The server can be implemented as an independent server or a server cluster consisting of multiple servers.

[0047] In one embodiment, Figure 2 As shown in the figure, a wind power-photovoltaic power joint prediction method is provided, which is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0048] Step S201 , based on the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the strong correlation characteristic factors of various prediction tasks are selected using the maximum correlation-minimum redundancy principle.

[0049] Among them, strongly correlated feature factors usually refer to a group of features with high correlation in data analysis and prediction, which may have a significant impact on the target variable (such as prediction results) in some fields.

[0050] Specifically, the server obtains the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, and selects strongly correlated characteristic factors of various prediction tasks based on the maximum correlation-minimum redundancy principle based on the historical power data and historical meteorological data.

[0051] Step S202: clustering analysis is performed on the large-scale power clusters in each wind farm and photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results. According to the clustering results, the spatial correlation between each wind farm and photovoltaic farm is quantitatively characterized.

[0052] The fuzzy clustering algorithm may be a FCM (Fuzzy C-means) clustering algorithm, which is mainly used to divide a data set into a number of fuzzy clusters.

[0053] Specifically, the server performs cluster analysis on large-scale power clusters through the FCM clustering algorithm, selects the central stations of various clusters obtained by clustering as representative stations, and establishes a silhouette coefficient to evaluate the clustering results. According to the obtained clustering results, the spatial correlation between the stations is quantitatively characterized.

[0054] Step S203, establishing a heterogeneous graph adjacency matrix according to the clustering results and the node modes, and characterizing the heterogeneous topological structure of the multi-modal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors.

[0055] Among them, the heterogeneous graph adjacency matrix can effectively represent the relationships between multiple types of nodes and edges in the graph, and it provides a flexible way to capture complex network structures.

[0056] Among them, heterogeneous topology refers to a network structure composed of different types of nodes and connection relationships. Compared with homogeneous topology, heterogeneous topology can more flexibly represent complex entities and their relationships.

[0057] Specifically, the server establishes a heterogeneous graph adjacency matrix according to the clustering results and node modes, and characterizes the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and strongly correlated characteristic factors.

[0058] Step S204, according to the heterogeneous topological structure, respectively establish a time series feature mining model and a spatial correlation feature mining model, and based on the time series feature mining model and the spatial correlation feature mining model, construct a wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each wind farm and photovoltaic farm in the target area.

[0059] Among them, the heterogeneous graph neural network is a deep learning model specially designed to process heterogeneous graphs. Through specially designed message passing and aggregation mechanisms, it can effectively learn and represent complex relationships in heterogeneous graphs.

[0060] Specifically, the server establishes a time series feature mining model and a spatial correlation feature mining model on the heterogeneous topology structure, and finally constructs a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network. The wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each wind farm and photovoltaic farm in the target area.

[0061] In the above-mentioned wind power-photovoltaic power joint prediction method, firstly, the maximum correlation-minimum redundancy principle is used to select strongly correlated characteristic factors of various prediction tasks; secondly, based on the fuzzy clustering algorithm, multiple wind farms and photovoltaic power sites are clustered into different groups; in addition, facing the problem of regional multi-cluster station new energy power prediction, by deeply exploring the spatiotemporal correlation characteristics between different stations, a heterogeneous graph adjacency matrix is ​​established and combined with strongly correlated characteristic factors, the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data is characterized, and a wind power-photovoltaic power joint prediction model based on heterogeneous graph neural network is constructed. The improved Beluga optimization algorithm is introduced to solve the optimal hyperparameters of the deep network model, further improving the prediction performance and generalization ability of the model, thereby simplifying the model and parameter size, effectively improving the accuracy, stability and generalization of the joint prediction of wind power and photovoltaic power, providing strong support for the stable operation and optimized scheduling of the power system, and can be expanded to many stations.

[0062] In one embodiment, before selecting strongly correlated characteristic factors of various prediction tasks using the maximum correlation-minimum redundancy principle based on the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the following steps are also included:

[0063] The historical power data is normalized to normalize the historical power data to a preset range; abnormal data under normal operating conditions in the historical power data and historical meteorological data are identified, and the abnormal data is interpolated and filled using the quartile method.

[0064] Specifically, suppose the set of new energy stations in a region is ={1,2,…,N}, first, each station The original historical power time series is normalized to the interval [0,1] according to the following formula:

[0065] (1)

[0066] In the above formula, is the original output data of wind farm i at time t, p i,min and p i,max are the minimum and maximum values ​​of the power time series of the i-th wind farm, respectively, and p i,t is the output value of wind farm i at time t after data normalization, that is, 0≤p i,t ≤1.

[0067] Next, the historical power data and historical meteorological data are preprocessed, and the statistical method quartile method is used to interpolate and fill in abnormal data such as power missing data, 0 value data, negative data and excess data under normal operating conditions. The calculation formula of the quartile method can be expressed as:

[0068] (2)

[0069] In the above formula, Q3 is the 75th percentile of the samples in the data set, Q1 is the 25th percentile of the samples in the data set, and IQR is the interquartile range. For outliers in the data set, the following formula can be used to fill them:

[0070] (3)

[0071] In one embodiment, before selecting the strongly relevant characteristic factors of various prediction tasks using the maximum correlation-minimum redundancy principle, the following steps are also included:

[0072] According to the historical meteorological data, multiple historical meteorological characteristics are obtained;

[0073] In the above step S201, the maximum correlation-minimum redundancy principle is used to select the strongly correlated characteristic factors of various prediction tasks, which specifically includes the following steps:

[0074] According to the time series of the first meteorological feature and the wind power sequence of the target electric field, a first mutual information value between the first meteorological feature and the wind power sequence is determined; according to the time series of the first meteorological feature and the time series of the second meteorological feature, a second mutual information value between the first meteorological feature and the second meteorological feature is determined; the first meteorological feature and the second meteorological feature are any two different meteorological features among multiple historical meteorological features; based on the first mutual information value and the second mutual information value, a maximum correlation standard formula is used to select candidate characteristic factors that are strongly correlated with the wind power sequence, and a minimum redundancy standard formula is used to remove redundant characteristic factors from the candidate characteristic factors to obtain strongly correlated characteristic factors.

[0075] It should be noted that the historical meteorological data in the target area refers to the original observation data on the state of the atmosphere and climate, such as temperature, humidity, air pressure and rainfall, collected through various observation means (such as ground meteorological stations, meteorological satellites and radars); and the historical meteorological characteristics are the key information extracted from the historical meteorological data for analyzing weather phenomena, climate change and weather forecasts. These meteorological characteristics are usually indicators obtained after processing, aggregation or statistics of the original observation data, and are highly abstract. Common meteorological characteristics include: temperature extremes (such as the highest and lowest daily temperatures, which represent the characteristics of temperature fluctuations), humidity fluctuations (the changing trend of humidity over a certain period of time), pressure gradient (the rate of change of air pressure with space) and rainfall intensity (divided into light rain, moderate rain, heavy rain and other categories according to the amount of precipitation, duration, etc.).

[0076] For example, when the historical meteorological characteristics include temperature extremes, humidity fluctuations, pressure gradients, and rainfall intensity, if the first meteorological characteristic is temperature extremes, the second meteorological characteristic may be humidity fluctuations, pressure gradients, or rainfall intensity.

[0077] Specifically, based on the mutual information theory, the server uses the maximum relevance-minimum redundancy (mRMR) principle to select the strongly correlated feature factors in various prediction tasks and the key features with the greatest correlation and the least redundancy:

[0078] (4)

[0079] (5)

[0080] In the above formula, x i is the time series of meteorological feature i, p is the wind power series of the target electric field, H(.) represents the calculation of entropy, is the mutual information value between meteorological feature i and the target wind power sequence, is the mutual information value between meteorological feature i and meteorological feature j; then the maximum correlation standard maxD(S,y) is used to select the characteristic factors with greater correlation with the power sequence; based on the minimum redundancy standard min R(S), the redundant characteristic factors are removed:

[0081] (6)

[0082] (7)

[0083] In the above formula, D represents the features x in the initial data set S i and the mean of the mutual information between the power, R is the mutual information size between each feature in S, S m is the number of features selected in the dataset S, x i and x j are the time series of meteorological characteristics i and j respectively.

[0084] In one embodiment, in the above step S202, a fuzzy clustering algorithm is used to perform clustering analysis on large-scale power clusters in each wind farm and photovoltaic farm to obtain a clustering result, which specifically includes the following steps:

[0085] While calculating the site location information of each wind farm and photovoltaic farm using a fuzzy clustering algorithm, the annual power time series of the large-scale power cluster is used as the characteristic attribute of the wind power-photovoltaic site for spatial clustering. The central site of each large-scale power cluster obtained by clustering is selected as the representative site, and the silhouette coefficient is established based on the average distance between the representative sites. The clustering results are evaluated using the average value of the silhouette coefficient.

[0086] Specifically, in order to quantitatively characterize the spatial correlation between different stations, the FCM (Fuzzy C-means) algorithm is used to calculate the station location information while taking the annual power time series of wind power-photovoltaic stations as the characteristic attributes of wind power-photovoltaic stations for spatial clustering. The objective function is optimized to obtain the membership of each sample point relative to all cluster centers, and the data point is assigned to the class with the maximum membership to complete the accurate classification of the sample set.

[0087] Assume that the cluster contains N new energy stations P={p1,p2,…,p n} is divided into K categories, where K categories are {q1,q2,…,q k FCM is achieved by minimizing the sum of weighted distances from the grouped station power sequence to the center of all station power sequences:

[0088] (8)

[0089] In the above formula, p i is the i-th data point, qj is the center of the jth cluster, u ij It indicates the membership of the i-th data point to the j-th cluster. m is a weighted index, usually m>1, which is used to control the fuzziness of clustering.

[0090] In order to find the minimum value of the objective function, it is necessary to iteratively calculate the membership u ij and the cluster center q j , the process can be expressed as:

[0091] (9)

[0092] (10)

[0093] FCM clustering is an unsupervised algorithm that uses the silhouette coefficient (SC) to evaluate clustering results. The formula is as follows:

[0094] (11)

[0095] In the above formula, m(i) is the average distance between the output station i and other stations, and n(i) is the average distance between the output station and all stations belonging to other classes. Finally, the average value of the silhouette coefficient of all data points is used to evaluate the entire clustering result. The value of the silhouette coefficient ranges from -1 to 1. The larger the value, the better the clustering effect.

[0096] In one embodiment, in the above step S204, based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed, which specifically includes the following steps:

[0097] The time series data modeled by temporal convolutional network is used to extract the time series features of topological nodes in the time series data. The time series features are used as input, and the spatial correlation between stations is modeled based on the graph attention network to obtain the spatial correlation features. Based on the time series features and spatial correlation features, a wind power-photovoltaic power joint prediction model is constructed.

[0098] Specifically, the server constructs a wind power-photovoltaic power joint prediction model based on the FCM clustering results and the modes of the nodes, which integrates heterogeneous graph neural networks. The model mainly includes the wind power-photovoltaic heterogeneous graph topology structure and the wind power-photovoltaic joint spatiotemporal correlation feature mining model based on heterogeneous graph neural networks.

[0099] In order to prevent the propagation of different modal data group characteristics from being confused, a wind power-photovoltaic heterogeneous graph adjacency matrix is ​​established, so that the heterogeneous graph can well contain the internal mode types. The heterogeneous graph adjacency matrix A is a square matrix, whose elements A ij; represents the connection relationship between node i and node j. In the heterogeneous graph, the modal dimension of node data is added on the basis of the traditional isomorphic graph to increase the diversity of data type representation. At the same time, in order to prevent the irrelevant prediction information of other modal data from interfering with the learning of the current modal mapping relationship, the edge weights of the adjacency matrix are processed heterogeneously. Its structure is as follows: Figure 3 shown.

[0100] Heterogeneous graph topology is used to perform type-aware weighted aggregation on the features of different types of nodes and their adjacent nodes, and mapping is performed through multi-layer perceptrons designed for different node types, thereby enhancing the ability to represent complex and diverse relationships in the graph.

[0101] Then, a wind power-photovoltaic power joint spatiotemporal correlation feature mining model based on heterogeneous graph neural network is established. The temporal convolutional network (TCN) is used to model time series data and extract sample time series features of topological nodes. The causal convolution process can be expressed as:

[0102] (12)

[0103] In the above formula, x t is the node feature input at time step t, h t is the hidden state or output feature at time step t, W k is the weight matrix of the kth convolution kernel. K is the number of convolution kernels, It is to introduce nonlinear activation function.

[0104] The time series features h after TCN processing t , each row represents the node feature of a time step. t As input, the spatial correlation between stations is modeled based on the Graph Attention Network (GAT):

[0105] (13)

[0106] In the above formula, represents the new spatial feature representation of node v after GAT processing, N(v) represents the set of neighboring station nodes of node v, is the attention coefficient between node stations u and v, indicating the contribution of station u’s power sequence to the feature of v, c v is the number of neighboring stations of node v. W is the weight matrix used for feature transformation.

[0107] Combining the temporal and spatial correlation features extracted by TCN and GAT, a joint prediction model based on heterogeneous graph neural network is constructed. The model can be expressed as:

[0108] (14)

[0109] In the above formula, H (l+1) is the node feature representation of the lth layer, TCN and GAT represent the process of temporal and spatial feature extraction respectively.

[0110] In one of the embodiments, after constructing a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network based on a time series feature mining model and a spatial correlation feature mining model, the following steps are also included:

[0111] The improved Beluga optimization algorithm is used to determine the optimal hyperparameters of the deep network model, and the optimal hyperparameters are substituted into the heterogeneous graph neural network to obtain an updated wind power-photovoltaic power joint prediction model; the updated wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each wind farm and photovoltaic farm in the target area to obtain the wind power-photovoltaic power prediction value.

[0112] Specifically, for the above-mentioned wind power-photovoltaic power joint prediction model integrating heterogeneous graph neural network, the improved Beluga optimization algorithm is introduced to determine the optimal hyperparameters of the deep network model. The traditional Beluga optimization algorithm mainly consists of three processes: exploration, development and whale fall. The formula of the optimization algorithm can be expressed as:

[0113] (15)

[0114] (16)

[0115] (17)

[0116] In the above formula, T(.), K(.) and J(.) represent the exploration, development and whale fall processes of the algorithm respectively. i t represents the position of the i-th beluga whale at time t, b i is the best position currently found, and rand is a random number. Represents the Levy flight process, which is used to enhance the global search capability of the algorithm. rand is the location of a randomly selected beluga whale.

[0117] Based on the original algorithm, the study improved it by adding the white whale disease strategy to speed up the convergence of the algorithm. The process can be expressed as:

[0118] (18)

[0119] In the above formula, rand1 and rand2 are both random numbers. At this time, calculate the balance factor B f and the probability of the white whale falling W f , when B f >W f When , the beluga whale enters the disease stage, otherwise it enters the whale fall stage. The detailed process of solving the optimal hyperparameters of the deep network model based on the improved beluga optimization algorithm is as follows: Figure 4 shown.

[0120] Through the above steps, the improved Beluga optimization algorithm can be effectively used to determine the optimal hyperparameters of the deep network model, and these hyperparameters can be applied to the heterogeneous graph neural network model. The trained model is used to infer the test data, and the prediction results output by the model are aggregated to finally obtain the power prediction values ​​of the wind farms and photovoltaic farms in the region.

[0121] In one embodiment, Figure 5 As shown, a wind power-photovoltaic power joint prediction method is provided, which specifically includes the following steps:

[0122] Step S501, obtain the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, normalize the historical power data to normalize the historical power data to a preset range; identify abnormal data under normal operating conditions in the historical power data and historical meteorological data, and use the quartile method to interpolate and fill the abnormal data.

[0123] Step S502: determine a first mutual information value between the first meteorological feature and the wind power sequence according to the time series of the first meteorological feature and the wind power sequence of the target electric field; determine a second mutual information value between the first meteorological feature and the second meteorological feature according to the time series of the first meteorological feature and the time series of the second meteorological feature; based on the first mutual information value and the second mutual information value, select candidate characteristic factors that are strongly correlated with the wind power sequence using the maximum correlation standard, and remove redundant characteristic factors from the candidate characteristic factors using the minimum redundancy standard to obtain strongly correlated characteristic factors.

[0124] Step S503, while calculating the site location information of each wind farm and photovoltaic farm using a fuzzy clustering algorithm, use the annual power time series of the large-scale power cluster as the characteristic attribute of the wind power-photovoltaic site for spatial clustering; select the central site of each large-scale power cluster obtained by clustering as a representative site, and establish a silhouette coefficient based on the average distance between the representative sites; and use the average value of the silhouette coefficient to evaluate the clustering result.

[0125] Step S504: quantitatively characterize the spatial correlation between each wind farm and photovoltaic farm according to the clustering result.

[0126] Step S505, establishing a heterogeneous graph adjacency matrix according to the clustering results and node modes, and characterizing the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors; and establishing a time series feature mining model and a spatial correlation feature mining model according to the heterogeneous topological structure.

[0127] Step S506, using the time series data modeled by the time convolution network, extracting the time series features of the topological nodes in the time series data; using the time series features as input, modeling the spatial correlation between the stations based on the graph attention network to obtain the spatial correlation features; and constructing a wind power-photovoltaic power joint prediction model based on the time series features and the spatial correlation features.

[0128] Step S507, using the improved Beluga optimization algorithm to determine the optimal hyperparameters of the deep network model, substituting the optimal hyperparameters into the heterogeneous graph neural network to obtain an updated wind power-photovoltaic power joint prediction model; using the updated wind power-photovoltaic power joint prediction model to predict the power values ​​of each wind farm and photovoltaic farm in the target area, to obtain a wind power-photovoltaic power prediction value.

[0129] The beneficial effects brought by the above embodiments are as follows:

[0130] This application proposes a wind power-photovoltaic power joint prediction method that takes into account spatiotemporal correlation, and constructs a wind power-photovoltaic power generation spatiotemporal coupling model based on a heterogeneous graph convolutional neural network, which effectively improves the accuracy, stability and generalization of renewable energy output prediction, and provides decision support for the coordinated scheduling and control of wind power-photovoltaic combined power generation systems.

[0131] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0132] Based on the same inventive concept, the embodiment of the present application also provides a wind power-photovoltaic power joint prediction device for implementing the wind power-photovoltaic power joint prediction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more wind power-photovoltaic power joint prediction device embodiments provided below can refer to the limitations of the wind power-photovoltaic power joint prediction method above, and will not be repeated here.

[0133] In an exemplary embodiment, Figure 6 As shown, a wind power-photovoltaic power joint prediction device is provided, comprising:

[0134] The data processing module 601 is used to select strong correlation characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle;

[0135] The correlation characterization module 602 is used to perform cluster analysis on the large-scale power clusters in each wind farm and photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results, and quantitatively characterize the spatial correlation between each wind farm and photovoltaic farm based on the clustering results;

[0136] The structure characterization module 603 is used to establish a heterogeneous graph adjacency matrix according to the clustering results and the node modes, and characterize the heterogeneous topological structure of the multi-modal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strong correlation characteristic factors;

[0137] The model building module 604 is used to establish a time series feature mining model and a spatial correlation feature mining model according to the heterogeneous topological structure, and to build a wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each wind farm and photovoltaic farm in the target area.

[0138] In one embodiment, the wind power-photovoltaic power joint prediction device also includes a feature extraction module, which is used to obtain multiple historical meteorological features based on historical meteorological data; the data acquisition module 601 is also used to determine the first mutual information value between the first meteorological feature and the wind power sequence based on the time series of the first meteorological feature and the wind power sequence of the target electric field; determine the second mutual information value between the first meteorological feature and the second meteorological feature based on the time series of the first meteorological feature and the time series of the second meteorological feature; based on the first mutual information value and the second mutual information value, select candidate feature factors that are strongly correlated with the wind power sequence using the maximum correlation standard, and use the minimum redundancy standard to remove redundant feature factors from the candidate feature factors to obtain strongly correlated feature factors.

[0139] In one embodiment, the correlation characterization module 602 is also used to use a fuzzy clustering algorithm to calculate the site location information of each wind farm and photovoltaic farm, and to use the annual power time series of the large-scale power cluster as the characteristic attribute of the wind power-photovoltaic site for spatial clustering; select the central site of each large-scale power cluster obtained by clustering as a representative site, and establish a silhouette coefficient based on the average distance between each representative site; and use the average value of the silhouette coefficient to evaluate the clustering results.

[0140] In one embodiment, the wind power-photovoltaic power joint prediction device also includes a data processing module, which is used to normalize the historical power data to normalize the historical power data to a preset range; identify abnormal data under normal operating conditions in the historical power data and historical meteorological data, and use the quartile method to interpolate and fill the abnormal data.

[0141] In one embodiment, the wind power-photovoltaic power joint prediction device also includes a model updating module, which is used to use the improved Beluga optimization algorithm to determine the optimal hyperparameters of the deep network model, substitute the optimal hyperparameters into the heterogeneous graph neural network, and obtain an updated wind power-photovoltaic power joint prediction model; use the updated wind power-photovoltaic power joint prediction model to predict the power values ​​of each wind farm and photovoltaic farm in the target area to obtain a wind power-photovoltaic power prediction value.

[0142] In one embodiment, the model building module 604 is also used to extract time series features of topological nodes from time series data using time convolutional network modeling; using the time series features as input, based on the graph attention network, modeling the spatial correlation between sites to obtain spatial correlation features; and constructing a wind power-photovoltaic power joint prediction model based on the time series features and the spatial correlation features.

[0143] Each module in the above-mentioned wind power-photovoltaic power joint prediction device can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0144] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 7As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store historical power data and historical meteorological data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a wind power-photovoltaic power joint prediction method is implemented.

[0145] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0146] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0147] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0148] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0149] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0150] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0151] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0152] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A wind power-photovoltaic power joint prediction method, characterized in that: The method comprises: According to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area, the strong correlation characteristic factors of various prediction tasks are selected by using the maximum correlation-minimum redundancy principle; Performing cluster analysis on the large-scale power clusters in each of the wind farms and the photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results, and quantitatively characterizing the spatial correlation between each of the wind farms and the photovoltaic farm based on the clustering results; Establishing a heterogeneous graph adjacency matrix according to the clustering results and the node modes, and characterizing the heterogeneous topological structure of multimodal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors; According to the heterogeneous topological structure, a time series feature mining model and a spatial correlation feature mining model are established respectively, and based on the time series feature mining model and the spatial correlation feature mining model, a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network is constructed; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each of the wind farms and the photovoltaic farms in the target area.

2. The method according to claim 1, characterized in that Before selecting the strongly relevant characteristic factors of various prediction tasks by using the maximum correlation-minimum redundancy principle, the method further includes: According to the historical meteorological data, a plurality of historical meteorological characteristics are obtained; The method of selecting strongly relevant feature factors for various prediction tasks using the maximum correlation-minimum redundancy principle includes: Determining a first mutual information value between the first meteorological feature and the wind power sequence according to a time series of the first meteorological feature and a wind power sequence of the target electric field; Determine a second mutual information value between the first meteorological feature and the second meteorological feature according to the time series of the first meteorological feature and the time series of the second meteorological feature; the first meteorological feature and the second meteorological feature are any two different meteorological features among the plurality of the historical meteorological features; Based on the first mutual information value and the second mutual information value, a maximum correlation standard is used to select candidate characteristic factors that are strongly correlated with the wind power sequence, and a minimum redundancy standard is used to remove redundant characteristic factors from the candidate characteristic factors to obtain the strongly correlated characteristic factors.

3. The method according to claim 1, characterized in that The clustering analysis of the large-scale power clusters in each of the wind farms and the photovoltaic farms using a fuzzy clustering algorithm to obtain clustering results includes: While calculating the site location information of each of the wind farms and the photovoltaic farms using the fuzzy clustering algorithm, the annual power time series of the large-scale power cluster is used as a characteristic attribute of the wind power-photovoltaic site for spatial clustering; Selecting the central site of each of the large-scale power clusters obtained by clustering as a representative site, and establishing a silhouette coefficient according to the average distance between each of the representative sites; The clustering result is evaluated using the average value of the silhouette coefficient.

4. The method according to claim 1, characterized in that: Before selecting the strongly correlated characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle, the method further includes: Normalizing the historical power data to normalize the historical power data to a preset range; Abnormal data under normal operating conditions in the historical power data and the historical meteorological data are identified, and the abnormal data are interpolated and filled using the quartile method.

5. The method according to claim 1, characterized in that After constructing the wind power-photovoltaic power joint prediction model based on the heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model, the method further includes: Determine the optimal hyperparameters of the deep network model using the improved Beluga optimization algorithm, substitute the optimal hyperparameters into the heterogeneous graph neural network, and obtain an updated wind power-photovoltaic power joint prediction model; The updated wind power-photovoltaic power joint prediction model is used to predict the power values ​​of the wind farms and the photovoltaic farms in the target area to obtain a wind power-photovoltaic power prediction value.

6. The method according to any one of claims 1 to 5, characterized in that: The method of constructing a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model includes: Using time series data modeled by a temporal convolutional network, extracting time series features of topological nodes from the time series data; Taking the time series features as input, the spatial correlation between stations is modeled based on the graph attention network to obtain spatial correlation features; Based on the time series characteristics and the spatial correlation characteristics, the wind power-photovoltaic power joint prediction model is constructed.

7. A wind power-photovoltaic power joint prediction device, characterized in that: The device comprises: A data processing module is used to select strongly correlated characteristic factors of various prediction tasks according to the historical power data of each wind farm and photovoltaic farm in the target area and the historical meteorological data in the target area using the maximum correlation-minimum redundancy principle; A correlation characterization module, used to perform cluster analysis on the large-scale power clusters in each of the wind farms and the photovoltaic farm using a fuzzy clustering algorithm to obtain clustering results, and quantitatively characterize the spatial correlation between each of the wind farms and the photovoltaic farm based on the clustering results; A structure characterization module, used to establish a heterogeneous graph adjacency matrix according to the clustering results and node modes, and characterize the heterogeneous topological structure of multi-modal wind power-photovoltaic power generation data based on the heterogeneous graph adjacency matrix and the strongly correlated characteristic factors; A model building module is used to establish a time series feature mining model and a spatial correlation feature mining model according to the heterogeneous topology structure, and to build a wind power-photovoltaic power joint prediction model based on a heterogeneous graph neural network based on the time series feature mining model and the spatial correlation feature mining model; the wind power-photovoltaic power joint prediction model is used to predict the power values ​​of each of the wind farms and the photovoltaic farms in the target area.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images