A distributed photovoltaic cluster short-term power prediction method, system and medium

Abstract

A distributed photovoltaic cluster short-term power prediction method, system and medium, comprising: acquiring weather forecast grid point data at a prediction time and latitude and longitude information of each distributed photovoltaic power station in a to-be-tested distributed photovoltaic cluster; determining a resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid point data at the prediction time and the latitude and longitude information of each distributed photovoltaic power station in the to-be-tested distributed photovoltaic cluster; and bringing the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a distributed photovoltaic cluster short-term power prediction value; wherein the pre-trained mapping relationship model is obtained by training an RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values. The present application realizes the distributed photovoltaic cluster short-term power prediction value by converting the weather forecast grid point data into the resource forecast value of the distributed photovoltaic cluster, and improves the calculation efficiency of the distributed photovoltaic cluster short-term power prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of solar energy utilization, and in particular to a method, system and medium for predicting short-term power of a distributed photovoltaic cluster. Background Art

[0002] In recent years, distributed photovoltaics have become a key area of ​​new energy development. Distributed photovoltaics are characterized by their widespread and widespread presence, and their large-scale grid connection can impact the safe and stable operation of the power grid. The capacity of individual distributed photovoltaics is relatively small, making it difficult to ensure the quality of monitoring data.

[0003] Most current distributed photovoltaic cluster power prediction methods use the prediction results of a single distributed photovoltaic unit as the minimum unit, calculating regional prediction results through accumulation or statistical upscaling. These methods require high-quality monitoring data from model sites. As the cluster expands and the number of distributed photovoltaic units increases exponentially, the computational efficiency of these prediction models will be significantly reduced. Summary of the Invention

[0004] To address the problem that the computational efficiency of existing prediction models will be greatly reduced as the number of distributed photovoltaics increases exponentially with the expansion of clusters, the present invention proposes a short-term power prediction method for distributed photovoltaic clusters, including:

[0005] Obtain weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested;

[0006] Determining a resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the prediction time and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured;

[0007] Bringing the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster;

[0008] The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values.

[0009] Optionally, the training of the mapping relationship model includes:

[0010] Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values;

[0011] Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data;

[0012] Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value;

[0013] The RBF neural network is trained with the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network.

[0014] Optionally, the calculating of equivalent resource forecast information of a distributed photovoltaic cluster based on the historical numerical weather forecast grid data includes:

[0015] The average value of the data of the four grid points of each distributed photovoltaic power station is used as the resource forecast data of the grid where the distributed photovoltaic power station is located;

[0016] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0017] The resource forecast data of the grid where the distributed photovoltaic power station is located and the distributed photovoltaic installed capacity of the network are combined with the resource forecast value calculation formula of the cluster to obtain equivalent resource forecast information of the distributed photovoltaic cluster.

[0018] Optionally, the training of the RBF neural network using the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network includes:

[0019] Inputting the equivalent resource forecast information in the sample set into the hidden layer of the RBF neural network through the input layer of the RBF neural network;

[0020] The hidden layer neurons use the distance between the input vector and the center vector as the independent variable of the function and use the radial basis function as the activation function to map the input vector to the output layer of the RBF neural network;

[0021] The output value of the distributed photovoltaic cluster is used as the output of the output layer of the RBF neural network, and a mapping relationship between equivalent resource forecast information and the output value of the distributed photovoltaic cluster is constructed.

[0022] Optionally, determining the resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the prediction time and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured includes:

[0023] Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point;

[0024] The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located;

[0025] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0026] The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the network and the resource forecast value calculation formula of the cluster.

[0027] Optionally, the resource forecast value of the cluster is calculated as follows:

[0028]

[0029] Where T is the resource forecast value of the cluster, T k is the resource forecast value of the kth grid, Cap k is the total installed capacity of distributed photovoltaic power generation in the Kth grid, Cap is the total installed capacity of distributed photovoltaic clusters, k is the number of grids, and N is the number of grids.

[0030] Optionally, the resource forecast value of the cluster includes one or more of the following: total irradiance, temperature, air pressure, and humidity.

[0031] In another aspect, the present invention further provides a distributed photovoltaic cluster short-term power prediction system, comprising:

[0032] An acquisition module is used to obtain the weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested;

[0033] A determination module, configured to determine a resource forecast value of a distributed photovoltaic cluster based on the weather forecast grid data at the prediction moment and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured;

[0034] A prediction module is used to bring the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster;

[0035] The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values.

[0036] Optionally, a training module is also included for:

[0037] Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values;

[0038] Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data;

[0039] Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value;

[0040] The RBF neural network is trained with the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network.

[0041] Optionally, the determining module is specifically configured to:

[0042] Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point;

[0043] The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located;

[0044] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0045] The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the network and the resource forecast value calculation formula of the cluster.

[0046] In another aspect, the present application further provides a computing device, comprising: one or more processors;

[0047] a processor for executing one or more programs;

[0048] When the one or more programs are executed by the one or more processors, the above-mentioned method for short-term power prediction of a distributed photovoltaic cluster is implemented.

[0049] On the other hand, the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed, the method for short-term power prediction of a distributed photovoltaic cluster as described above is implemented.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] The present invention provides a method for short-term power prediction of a distributed photovoltaic cluster, comprising obtaining weather forecast grid point data at the prediction time and latitude and longitude information of each distributed photovoltaic power station in a distributed photovoltaic cluster to be tested; determining a resource forecast value for the distributed photovoltaic cluster based on the weather forecast grid point data at the prediction time and the latitude and longitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested; and entering the resource forecast value for the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value for the distributed photovoltaic cluster; wherein the pre-trained mapping relationship model is obtained by training an RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values. The present invention achieves a short-term power forecast value for the distributed photovoltaic cluster by converting weather forecast grid point data into a resource forecast value for the distributed photovoltaic cluster, thereby improving the computational efficiency of the short-term power prediction of the distributed photovoltaic cluster.

[0052] The present invention constructs a mapping relationship model between equivalent resource forecast information and distributed photovoltaic cluster output values. The mapping relationship model has low dependence on single-point monitoring data and high algorithm stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a flow chart of a method for short-term power prediction of a distributed photovoltaic cluster according to the present invention;

[0054] Figure 2 This is a flow chart of a method for short-term power prediction of a distributed photovoltaic cluster in an embodiment of the present invention;

[0055] Figure 3 A schematic diagram of grid matching according to the present invention;

[0056] Figure 4 Schematic diagram of the prediction effect in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] This paper proposes a method for short-term power forecasting for distributed photovoltaic clusters. This method converts gridded numerical weather forecasts into equivalent resource forecast information for distributed photovoltaic clusters and constructs a mapping model between historical equivalent resource forecast information and the total output of distributed photovoltaic clusters. This method significantly improves the computational efficiency of short-term power forecasting for distributed photovoltaic clusters, while also maintaining a low dependency on single-point monitoring data and high algorithm stability.

[0058] Example 1:

[0059] A short-term power prediction method for distributed photovoltaic clusters, such as Figure 1 Shown, including:

[0060] S1: Obtain the weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested;

[0061] S2: determining a resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the prediction time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be measured;

[0062] S3: Substitute the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster;

[0063] The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values.

[0064] The following combination Figure 2 The present invention provides a method for predicting the short-term power of a distributed photovoltaic cluster.

[0065] Before S1, it also includes the training of the mapping relationship model.

[0066] The training of the mapping relationship model specifically includes:

[0067] Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values;

[0068] Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data;

[0069] Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value;

[0070] The RBF neural network is trained with the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network.

[0071] Preferably, historical numerical weather forecast grid point data and corresponding distributed photovoltaic cluster output values ​​are obtained. In this embodiment, obtaining numerical weather forecast grid point data for a period of history is taken as an example for introduction.

[0072] Preferably, calculating the equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data includes:

[0073] The average value of the data of the four grid points of each distributed photovoltaic power station is used as the resource forecast data of the grid where the distributed photovoltaic power station is located;

[0074] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0075] The resource forecast data of the grid where the distributed photovoltaic power station is located and the distributed photovoltaic installed capacity of the network are combined with the resource forecast value calculation formula of the cluster to obtain equivalent resource forecast information of the distributed photovoltaic cluster.

[0076] Preferably, the average value of the data of the four grid points of each distributed photovoltaic power station is used as the resource forecast data of the grid where the distributed photovoltaic power station is located, specifically including:

[0077] T k =(T g1 +T g2 +T g3 +T g4 ) / 4,

[0078] Among them, T g1 , T g2 , T g3 , T g4 It is the resource forecast data of the four grid points of the grid.

[0079] The cluster resource forecast value calculation formula is as follows:

[0080]

[0081] Where T is the resource forecast value of the cluster, T k is the resource forecast value of the kth grid, Cap k is the total installed capacity of distributed photovoltaic power generation in the Kth grid, Cap is the total installed capacity of distributed photovoltaic clusters, k is the number of grids, and N is the number of grids.

[0082] Preferably, the RBF neural network is trained using the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network, including:

[0083] Inputting the equivalent resource forecast information in the sample set into the hidden layer of the RBF neural network through the input layer of the RBF neural network;

[0084] The hidden layer neurons use the distance between the input vector and the center vector as the independent variable of the function and use the radial basis function as the activation function to map the input vector to the output layer of the RBF neural network;

[0085] The output value of the distributed photovoltaic cluster is used as the output of the output layer of the RBF neural network, and a mapping relationship between equivalent resource forecast information and the output value of the distributed photovoltaic cluster is constructed.

[0086] The RBF neural network model training process includes:

[0087] 1) The input layer is based on the calculated historical equivalent resource forecast information of the distributed photovoltaic cluster, including total irradiance, temperature, air pressure, humidity and other factors.

[0088] 2) The hidden layer neurons use the distance between the input vector and the center vector (Euclidean distance) as the independent variable of the function and use the radial basis function as the activation function. The farther the neuron input is from the center of the radial basis function, the lower the degree of neuron activation.

[0089] 3) The output layer represents the output power of the distributed photovoltaic cluster. This model uses the concept of direct prediction to construct a direct mapping from multiple meteorological factors to output power. The mapping from the hidden layer space to the output space is a linear relationship. That is, the network output is the linearly weighted sum of the outputs of each hidden layer neuron. The weights are obtained through training.

[0090] S1 specifically includes:

[0091] Obtain the grid longitude and latitude information of the weather forecast data, the weather forecast grid data at the forecast time, and obtain the longitude and latitude information of each distributed photovoltaic site in the area.

[0092] Weather forecast grid data include: total irradiance, temperature, air pressure, humidity and other numerical weather forecast data.

[0093] The step S2 of determining the resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be measured specifically includes:

[0094] Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point;

[0095] The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located;

[0096] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0097] The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the network and the resource forecast value calculation formula of the cluster.

[0098] Among them, the grid points of the weather forecast are matched based on the latitude and longitude information of each distributed photovoltaic power station point, such as Figure 3 As shown, specifically including:

[0099] According to the longitude and latitude information of the distributed photovoltaic site, the distributed photovoltaic site is matched to the nearest grid. For example, the numerical weather forecast of the area where the distributed photovoltaic cluster is located is marked as N grids. By comparing the longitude and latitude of the distributed photovoltaic site with the longitude and latitude of the numerical weather forecast grid, it is determined that the distributed photovoltaic site belongs to the kth grid.

[0100] The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located, specifically including:

[0101] The resource forecast value calculation method for a single grid is:

[0102] T k =(T g1 +T g2 +T g3 +T g4 ) / 4,

[0103] Among them, T g1 , T g2 , T g3 , T g4 It is the resource forecast data of the four grid points of the grid.

[0104] The sum of the installed capacities of all distributed photovoltaic sites in each grid is calculated to obtain the total distributed photovoltaic installed capacity of the network, which specifically includes:

[0105] Calculate the distributed photovoltaic site installed capacity and Cap in the non-empty grid. The non-empty grid here refers to the grid with distributed photovoltaic. By accumulating, the distributed photovoltaic installed capacity and Cap of the kth grid are obtained. k .

[0106] The cluster resource forecast value calculation formula is as follows:

[0107]

[0108] Where T is the resource forecast value of the cluster, T k is the resource forecast value of the kth grid, Cap k is the total installed capacity of distributed photovoltaic power generation in the Kth grid, Cap is the total installed capacity of distributed photovoltaic clusters, k is the number of grids, and N is the number of grids.

[0109] The resource data here include total irradiance, temperature, air pressure, humidity and other forecast information that can be provided by numerical weather forecasts.

[0110] Preferably, the resource forecast value of the distributed photovoltaic cluster is brought into a pre-trained mapping relationship model in S3 to obtain the short-term power forecast value of the distributed photovoltaic cluster, specifically including:

[0111] Based on the future numerical weather forecast grid data, the equivalent resource forecast data is calculated, the input of the mapping relationship model is used, and the short-term forecast results of the future distributed photovoltaic cluster output are output.

[0112] According to the grid forecast data of the numerical weather forecast for the next several days, combined with the established grid form, the equivalent resource forecast data at the forecast time is calculated, which is used as the input of the trained RBF neural network model to calculate the future short-term power forecast value.

[0113] The distributed photovoltaic cluster short-term power prediction method provided by the present invention is used to calculate the distributed photovoltaic cluster power prediction result for a certain day. Figure 4 As shown in the result diagram, it can be seen that the equivalent resource prediction has a good correlation with the actual output, and the method proposed in the present invention can achieve a good prediction effect.

[0114] Example 2:

[0115] The present invention based on the same inventive concept also provides a distributed photovoltaic cluster short-term power prediction system, comprising:

[0116] An acquisition module is used to obtain the weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested;

[0117] A determination module, configured to determine a resource forecast value of a distributed photovoltaic cluster based on the weather forecast grid data at the prediction moment and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured;

[0118] A prediction module is used to bring the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster;

[0119] The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values.

[0120] A distributed photovoltaic cluster short-term power prediction system further includes: a module training module for:

[0121] Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values;

[0122] Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data;

[0123] Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value;

[0124] The RBF neural network is trained with the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network.

[0125] Preferably, the determining module is specifically configured to:

[0126] Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point;

[0127] The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located;

[0128] Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the network;

[0129] The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the network and the resource forecast value calculation formula of the cluster.

[0130] The resource forecast value calculation formula of the cluster is as follows:

[0131]

[0132] Where T is the resource forecast value of the cluster, T k is the resource forecast value of the kth grid, Cap k is the total installed capacity of distributed photovoltaic power generation in the Kth grid, Cap is the total installed capacity of distributed photovoltaic clusters, k is the number of grids, and N is the number of grids.

[0133] Preferably, the resource forecast value of the cluster includes one or more of the following: total irradiance, temperature, air pressure, and humidity.

[0134] Example 3:

[0135] Based on the same inventive concept, the present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a distributed photovoltaic cluster short-term power prediction method in the above embodiment.

[0136] Example 4:

[0137] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It can be understood that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the steps of a distributed photovoltaic cluster short-term power prediction method in the above embodiment.

[0138] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0139] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0142] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A distributed photovoltaic cluster short-term power prediction method, characterized in that: include: Obtain weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested; Determining a resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the prediction time and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured; Bringing the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster; The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value; The training of the mapping relationship model includes: Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values; Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data; Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value; The equivalent resource forecast information in the sample set is used as input, and the output value of the distributed photovoltaic cluster is used as output to train the RBF neural network, thereby obtaining a trained RBF neural network; The calculating of equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data includes: The average value of the data of the four grid points of each distributed photovoltaic power station is used as the resource forecast data of the grid where the distributed photovoltaic power station is located; Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the grid; The resource forecast data of the grid where the distributed photovoltaic power station is located and the distributed photovoltaic installed capacity of the grid are combined with the resource forecast value calculation formula of the cluster to obtain equivalent resource forecast information of the distributed photovoltaic cluster.

2. The method according to claim 1, wherein The RBF neural network is trained using the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network, including: Inputting the equivalent resource forecast information in the sample set into the hidden layer of the RBF neural network through the input layer of the RBF neural network; The hidden layer neurons use the distance between the input vector and the center vector as the independent variable of the function and use the radial basis function as the activation function to map the input vector to the output layer of the RBF neural network; The output value of the distributed photovoltaic cluster is used as the output of the output layer of the RBF neural network, and a mapping relationship between equivalent resource forecast information and the output value of the distributed photovoltaic cluster is constructed.

3. The method according to claim 1, wherein The determining of the resource forecast value of the distributed photovoltaic cluster based on the weather forecast grid data at the prediction time and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured includes: Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point; The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located; Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the grid; The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the grid, and the resource forecast value calculation formula of the cluster.

4. The method according to claim 3, wherein The resource forecast value calculation formula of the cluster is as follows: Where, Forecast value of cluster resources, is the resource forecast value of the kth grid, For the K The distributed photovoltaic installations of the grid and For distributed photovoltaic cluster installation and k is the grid number, N is the number of grids.

5. The method according to claim 4, wherein The resource forecast value of the cluster includes one or more of the following: total irradiance, temperature, air pressure, and humidity.

6. A system for implementing the distributed photovoltaic cluster short-term power prediction method according to any one of claims 1 to 5, characterized in that: include: An acquisition module is used to obtain the weather forecast grid data at the forecast time and the longitude and latitude information of each distributed photovoltaic power station in the distributed photovoltaic cluster to be tested; A determination module, configured to determine a resource forecast value of a distributed photovoltaic cluster based on the weather forecast grid data at the prediction moment and the longitude and latitude information of each distributed photovoltaic power station point in the distributed photovoltaic cluster to be measured; A prediction module is used to bring the resource forecast value of the distributed photovoltaic cluster into a pre-trained mapping relationship model to obtain a short-term power forecast value of the distributed photovoltaic cluster; The pre-trained mapping relationship model is obtained by training the RBF neural network based on historical equivalent resource forecast information and corresponding distributed photovoltaic cluster output values.

7. The system according to claim 6, wherein: Also included are training modules for: Obtain historical numerical weather forecast grid data and corresponding distributed photovoltaic cluster output values; Calculating equivalent resource forecast information of the distributed photovoltaic cluster based on the historical numerical weather forecast grid data; Constructing a sample set based on the equivalent resource forecast information and the corresponding distributed photovoltaic cluster output value; The RBF neural network is trained with the equivalent resource forecast information in the sample set as input and the output value of the distributed photovoltaic cluster as output to obtain a trained RBF neural network.

8. The system according to claim 6, wherein: The determining module is specifically configured to: Matching the weather forecast grid points based on the latitude and longitude information of each distributed photovoltaic power station point; The average value of the data of the four grid points of the grid where each distributed photovoltaic site is located is used as the resource forecast data of the grid where the distributed photovoltaic site is located; Sum the installed capacities of all distributed photovoltaic sites in each grid to obtain the total distributed photovoltaic installed capacity of the grid; The resource forecast value of the distributed photovoltaic cluster is obtained based on the resource forecast data of the grid where the distributed photovoltaic power station is located, the distributed photovoltaic installed capacity of the grid, and the resource forecast value calculation formula of the cluster.

9. A computer device, characterized in that: include: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, a distributed photovoltaic cluster short-term power prediction method according to any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed, a distributed photovoltaic cluster short-term power prediction method according to any one of claims 1 to 5 is implemented.

Images