Distributed renewable energy power generation data prediction method, device, equipment and medium

Abstract

This application discloses a method, apparatus, equipment, and medium for predicting power generation data of distributed new energy sources. The method involves: acquiring the power station area where the distributed photovoltaic power station is located; acquiring multiple first two-dimensional images of the sky area taken by different drones from different angles at a target time; performing three-dimensional reconstruction to obtain three-dimensional models of multiple target cloud clusters in the sky area; projecting the power station area and the multiple target cloud cluster three-dimensional models onto a first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the multiple target cloud cluster three-dimensional models, wherein the first plane is perpendicular to the direction of sunlight; determining the overall cloud cluster influence parameters of the distributed photovoltaic power station based on the positional relationship between the cloud cluster projection areas and the power station projection areas of the distributed photovoltaic power station; and determining the predicted power generation of the distributed photovoltaic power station based on the overall cloud cluster influence parameters. This application can improve the accuracy of power generation prediction for distributed new energy sources.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence processing technology, specifically to a method, device, equipment, and medium for predicting power generation data of distributed new energy sources. Background Technology

[0002] With the development of the photovoltaic (PV) power generation industry, a large amount of PV power is being connected to the grid. However, PV power generation is subject to randomness and fluctuations due to weather conditions, making it difficult to achieve stable power output. Currently, the electricity generated in PV bases is typically consumed locally, that is, directly utilized in the surrounding areas, thereby improving the reliability of PV power supply, reducing transmission costs, and enhancing grid resilience. Cloud formations have a significant impact on PV power generation, resulting in low accuracy in forecasting distributed renewable energy generation. Summary of the Invention

[0003] This application provides a method, apparatus, equipment, and medium for predicting power generation data of distributed new energy sources, which can improve the accuracy of power generation prediction for distributed new energy sources.

[0004] Firstly, the distributed renewable energy power generation data prediction method provided in this application includes:

[0005] Obtain the location of the distributed photovoltaic power station;

[0006] Multiple first two-dimensional images of the sky region taken by different drones from different angles at a target time are acquired. The distributed photovoltaic power station corresponds to a drone swarm, and the drone swarm flies and takes pictures within an area centered on the distributed power station.

[0007] Three-dimensional reconstruction is performed based on multiple first two-dimensional images to obtain three-dimensional models of multiple target cloud clusters in the sky region;

[0008] Obtain the direction of sunlight at the target time;

[0009] The three-dimensional models of the power station area and the multiple target cloud clusters are respectively projected onto the first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the multiple target cloud clusters three-dimensional models, wherein the first plane is perpendicular to the sunlight direction;

[0010] The overall influence parameters of the cloud cluster on the distributed photovoltaic power station are determined based on the positional relationship between the cloud cluster projection area and the power station projection area.

[0011] The predicted power generation of the distributed photovoltaic power station is determined based on the overall cloud impact parameters of the distributed photovoltaic power station.

[0012] Optionally, determining the predicted power generation of the distributed photovoltaic power station based on the overall cloud impact parameters of the distributed photovoltaic power station includes:

[0013] Calculate the overall impact parameters of cloud clusters on the distributed photovoltaic power station at multiple different times within a preset time period;

[0014] Meteorological parameters of the distributed photovoltaic power station at multiple different times within a preset time period are obtained respectively, wherein the meteorological parameters include at least one of temperature, wind speed, and humidity;

[0015] By splicing together the overall impact parameters of cloud clusters at multiple different times and meteorological parameters at multiple different times, input parameters at multiple different times are obtained.

[0016] By inputting multiple input parameters at different times into a preset power generation prediction model, the predicted power generation of the distributed photovoltaic power station is obtained.

[0017] Optionally, determining the overall cloud impact parameters of the distributed photovoltaic power station based on the positional relationship between the cloud projection area and the power station projection area includes:

[0018] If the projected area of ​​the power station overlaps with the projected area of ​​the target cloud cluster in the 3D model, the influence parameter of the target cloud cluster 3D model on the cloud cluster of the distributed photovoltaic power station is determined based on the overlapping area and the projected area of ​​the target cloud cluster. If the projected area of ​​the power station and the projected area of ​​the cloud cluster do not overlap, the preset influence parameter value is determined as the influence parameter of the target cloud cluster 3D model on the cloud cluster of the distributed photovoltaic power station.

[0019] The average value of the influence parameters of each target cloud cluster 3D model on the individual cloud clusters of the distributed photovoltaic power station is determined as the overall influence parameter of the cloud clusters of the distributed photovoltaic power station.

[0020] Optionally, determining the cloud cluster individual influence parameters of the target cloud cluster 3D model on the distributed photovoltaic power station based on the cloud cluster projection area of ​​the overlapping region and the target cloud cluster 3D model includes:

[0021] Obtain the area ratio of the overlapping region to the projected area of ​​the power station;

[0022] Generate multiple first straight lines that pass through the target cloud 3D model and the overlapping area and are parallel to the direction of sunlight;

[0023] The first straight line is truncated on the surface of the three-dimensional model of the target cloud to generate the first line segment, resulting in the first line segment of multiple first straight lines;

[0024] The influence parameters of the target cloud cluster 3D model on the cloud cluster individual of the distributed photovoltaic power station are determined based on the area ratio of the region and the average length of multiple first line segments. The larger the average length of multiple first line segments, the larger the influence parameter of the cloud cluster individual; the larger the area ratio of the region, the larger the influence parameter of the cloud cluster individual.

[0025] Optionally, determining the influence parameters of the target cloud cluster 3D model on the individual cloud clusters of the distributed photovoltaic power station based on the area ratio of the region and the average length of multiple first line segments includes:

[0026] Obtain the centroid distance between the centroid of the overlapping region and the centroid of the cloud projection region;

[0027] The influence parameters of the target cloud cluster 3D model on the cloud cluster individual of the distributed photovoltaic power station are determined based on the area ratio of the region, the centroid spacing and the average length of multiple first line segments. The larger the average length of multiple first line segments, the larger the influence parameter of the cloud cluster individual; the larger the area ratio of the region, the larger the influence parameter of the cloud cluster individual; and the larger the centroid spacing, the smaller the influence parameter of the cloud cluster individual.

[0028] Optionally, the step of performing three-dimensional reconstruction based on multiple first two-dimensional images to obtain three-dimensional models of multiple target cloud clusters in the sky region includes:

[0029] The SFM algorithm is used to reconstruct three-dimensional images from multiple first two-dimensional images to obtain initial point cloud data.

[0030] Select a predetermined ratio of first two-dimensional captured images from a plurality of first two-dimensional captured images as second two-dimensional captured images to obtain a plurality of second two-dimensional captured images, wherein the predetermined ratio is less than 1;

[0031] Cloud segmentation is performed on multiple second-two-dimensional captured images to obtain multiple cloud segmentation regions on the multiple second-two-dimensional captured images;

[0032] The point cloud data located within each of the cloud cluster segmentation regions in the initial point cloud data are defined as cloud cluster point cloud data;

[0033] Based on the cloud point cloud data, multiple three-dimensional models of the target cloud clusters are determined.

[0034] Optionally, generating multiple 3D models of the target cloud based on the cloud point cloud data includes:

[0035] Clustering the cloud point cloud data yields multiple first three-dimensional point clusters;

[0036] Based on the first three-dimensional point cluster, the three-dimensional model of the target cloud is determined, and multiple three-dimensional models of the target cloud corresponding to multiple first three-dimensional point clusters are obtained.

[0037] Secondly, the distributed renewable energy power generation data prediction device provided in this application includes:

[0038] The first acquisition module is used to acquire the power station area where the distributed photovoltaic power station is located;

[0039] The second acquisition module is used to acquire multiple first two-dimensional images of the sky area taken by different drones from different angles at the target time. The distributed photovoltaic power station corresponds to a drone cluster, and the drone cluster flies and takes pictures within the area centered on the distributed power station.

[0040] The three-dimensional reconstruction module is used to perform three-dimensional reconstruction based on multiple first two-dimensional captured images to obtain three-dimensional models of multiple target cloud clusters in the sky region;

[0041] The third acquisition module is used to acquire the direction of sunlight at the target time;

[0042] The projection module is used to project the power station area and the multiple target cloud cluster 3D models onto a first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the multiple target cloud cluster 3D models, wherein the first plane is perpendicular to the sunlight direction.

[0043] The first determining module is used to determine the overall influence parameters of the cloud cluster of the distributed photovoltaic power station based on the positional relationship between the cloud cluster projection area and the power station projection area of ​​the distributed photovoltaic power station.

[0044] The second determining module is used to determine the predicted power generation of the distributed photovoltaic power station based on the overall cloud impact parameters of the distributed photovoltaic power station.

[0045] Thirdly, the electronic device provided in this application includes a memory and a processor. The memory stores a computer program, and the processor is used to run the computer program in the memory to implement the steps in the distributed new energy power generation data prediction method provided in this application.

[0046] Fourthly, the computer-readable storage medium provided in this application stores multiple instructions that are suitable for a processor to load and implement the steps in the distributed new energy power generation data prediction method provided in this application.

[0047] Fifthly, the computer program product provided in this application includes a computer program or instructions that, when executed by a processor, implement the steps in the distributed new energy power generation data prediction method provided in this application.

[0048] In this application, compared to related technologies, the following methods are employed: First, the location of the distributed photovoltaic (PV) power station is obtained. Second, multiple first-dimensional images of the sky region are acquired from different angles by various drones at a target time. Each distributed PV power station corresponds to a drone swarm, with each swarm flying and taking pictures within an area centered on the distributed power station. Third, 3D reconstruction is performed based on the multiple first-dimensional images to obtain 3D models of multiple target cloud clusters in the sky region. Fourth, the direction of sunlight at the target time is obtained. Fifth, the power station region and the 3D models of the multiple target cloud clusters are projected onto a first plane to obtain the power station projection area corresponding to the power station region and the multiple cloud cluster projection areas corresponding to the 3D models of the multiple target cloud clusters. The first plane is perpendicular to the direction of sunlight. Fifth, the overall influence parameters of the clouds on the distributed PV power station are determined based on the positional relationship between the cloud cluster projection areas and the power station projection areas of the distributed PV power station. Sixth, the predicted power generation of the distributed PV power station is determined based on the overall influence parameters of the clouds on the distributed PV power station. This application can improve the accuracy of power generation prediction for distributed renewable energy sources. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a schematic diagram of a distributed new energy power generation data prediction system provided in an embodiment of this application;

[0051] Figure 2 This is a flowchart illustrating one embodiment of the distributed renewable energy power generation data prediction method provided in this application.

[0052] Figure 3 This is a schematic diagram of a three-dimensional model of a target cloud cluster and a projection of a power plant area in one embodiment of the distributed new energy power generation data prediction method provided in this application.

[0053] Figure 4 This is a schematic diagram of the cloud projection area and the power station projection area in one embodiment of the distributed new energy power generation data prediction method provided in this application;

[0054] Figure 5This is a schematic diagram of the structure of the distributed new energy power generation data prediction device provided in the embodiments of this application;

[0055] Figure 6 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0056] It should be noted that the principles of this application are illustrated by example in a suitable computing environment. The following description is based on the specific embodiments of this application that are illustrated, and should not be regarded as limiting other specific embodiments not detailed herein.

[0057] In the following description of this application, "some embodiments" are referred to, which describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same subset or different subset of all possible embodiments, and may be combined with each other without conflict.

[0058] In the following description of this application, the terms "first, second, third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0060] To improve the effectiveness of distributed renewable energy power generation data prediction, embodiments of this application provide a method, device, electronic device, computer-readable storage medium, and computer program product for predicting distributed renewable energy power generation data. The method can be executed by the distributed renewable energy power generation data prediction device or by an electronic device integrating the distributed renewable energy power generation data prediction device.

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0062] Please refer to Figure 1 This application also provides a distributed renewable energy power generation data prediction system, such as... Figure 1 As shown, the distributed new energy power generation data prediction system electronic device 100 integrates the distributed new energy power generation data prediction device provided in this application.

[0063] Among them, electronic device 100 can be any device equipped with a processor and having processing capabilities, such as mobile electronic devices with processors such as smartphones, tablets, PDAs, laptops, and smart speakers, or fixed electronic devices with processors such as desktop computers, televisions, servers, and industrial equipment.

[0064] In addition, such as Figure 1 As shown, the distributed new energy power generation data prediction system may also include a memory 200 for storing raw data, intermediate data and result data.

[0065] In this embodiment of the application, the memory 200 can be a cloud memory. Cloud storage is a new concept that is extended and developed from the concept of cloud computing. A distributed cloud storage system (hereinafter referred to as a storage system) refers to a storage system that uses cluster applications, grid technology and distributed storage file system functions to bring together a large number of storage devices of various types in the network (storage devices are also called storage nodes) through application software or application interfaces to work together to provide data storage and business access functions to the outside world.

[0066] Currently, the storage method of storage systems is as follows: Logical volumes are created. During the creation of a logical volume, physical storage space is allocated to each logical volume. This physical storage space may consist of a single storage device or the disks of several storage devices. Clients store data on a logical volume, which means storing the data on the file system. The file system divides the data into many parts, each part being an object. Each object contains not only the data but also additional information such as a data identifier (ID, ID entity). The file system writes each object to the physical storage space of that logical volume and records the storage location information of each object. Therefore, when a client requests access to data, the file system can allow the client to access the data based on the storage location information of each object.

[0067] The process by which a storage system allocates physical storage space to a logical volume is as follows: the physical storage space is pre-divided into strips according to the capacity estimate of the objects stored in the logical volume (this estimate often has a large margin relative to the actual capacity of the objects to be stored) and the grouping of Redundant Array of Independent Disks (RAID). A logical volume can be understood as a strip, thus allocating physical storage space to the logical volume.

[0068] It should be noted that, Figure 1 The schematic diagram of the distributed renewable energy power generation data prediction system shown is merely an example. The distributed renewable energy power generation data prediction system and scenario described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of the distributed renewable energy power generation data prediction system and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0069] The following sections provide detailed descriptions of each example. It should be noted that the sequence numbers of the following embodiments are not intended to limit the preferred order of the embodiments.

[0070] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating one embodiment of the distributed renewable energy power generation data prediction method provided in this application. Figure 2 As shown, the flow of the distributed renewable energy power generation data prediction method provided in this application is as follows:

[0071] 201. Obtain the power station area where the distributed photovoltaic power station is located.

[0072] In this embodiment of the application, the power station area is the area where the distributed photovoltaic power station is located.

[0073] 202. Acquire multiple first-dimensional images of the sky region taken by multiple different drones from different angles at the target time.

[0074] Each distributed photovoltaic power station corresponds to a drone swarm, with each drone swarm flying and taking pictures within an area centered on the distributed power station.

[0075] In this embodiment of the application, a ground station is provided in the distributed photovoltaic power station. The ground station is connected to a network of multiple drones in the drone swarm. The ground station controls multiple drones in the drone swarm to fly and take pictures within an area centered on the distributed power station.

[0076] Specifically, the process involves acquiring a third 2D image of the sky captured vertically upwards by a depth camera from a ground station, and a sky reference image captured vertically upwards by the same camera. The sky reference image is marked with the angles between each pixel and the vertical direction. The third 2D image is then used for detection to obtain the first cloud cluster segmentation region and the distance between each pixel in the first cloud cluster segmentation region and the distributed photovoltaic power station cloud cluster. Specifically, a cloud cluster segmentation model is pre-trained; this model can be the Unet model or other models, depending on the specific requirements. The third 2D image is then input into the cloud cluster segmentation model for detection to obtain the first cloud cluster segmentation region in the third 2D image. Current depth cameras can be categorized into three types based on their operating principles: TOF, RGB binocular, and structured light. The distance between each pixel in the first cloud cluster segmentation region and the distributed photovoltaic power station cloud cluster is obtained based on the depth information acquired by the depth camera.

[0077] Align the sky reference image and the third 2D image to obtain the angles between each pixel in the segmented region of the first cloud cluster on the third 2D image and the vertical direction. Based on the distance between each pixel in the segmented region of the first cloud cluster and the distributed photovoltaic power station cloud cluster, and the angles between each pixel in the segmented region of the first cloud cluster and the vertical direction, determine the physical distance between each pixel in the segmented region of the first cloud cluster and the image center of the third 2D image. The average value of these physical distances is then determined as the horizontal distance of the power station cloud cluster. The physical distance between a pixel and the image center of the third 2D image can be determined using trigonometric functions.

[0078] The latitude and longitude of the distributed photovoltaic power station are obtained. Based on the latitude and longitude of the distributed photovoltaic power station, a preset minimum solar irradiation angle and a preset cloud height are determined. Based on the preset cloud height and the preset minimum solar irradiation angle, a first distance threshold is determined. The first cloud segmentation region with a horizontal distance of less than the first distance threshold is determined as the second cloud segmentation region, resulting in multiple second cloud segmentation regions.

[0079] Multiple segmented regions of the second cloud clusters on the third two-dimensional image are projected onto the second plane where the distributed photovoltaic power station is located, resulting in multiple detection projection regions and centroids corresponding to the segmented regions of the second cloud clusters. A first polygon is determined based on the centroids of the multiple detection projection regions, wherein the centroids of all the detection projection regions are located within the first polygon, and the vertices of the first polygon are the centroids of the detection projection regions. The first polygon is enlarged according to a preset ratio to obtain a second polygon, wherein the shape and centroids of the first and second polygons are identical, and all the detection projection regions are located within the second polygon. The spatial region enclosed by the second polygon is defined as the activity area of ​​each drone in the drone swarm.

[0080] 203. Based on multiple first-dimensional images, perform three-dimensional reconstruction to obtain three-dimensional models of multiple target cloud clusters in the sky region.

[0081] In one specific embodiment, the SFM algorithm is used to perform 3D reconstruction on multiple first-dimensional captured images to obtain 3D models of multiple target cloud clusters in the sky region. In computer vision, 3D reconstruction refers to the process of reconstructing 3D information from single-view or multi-view images. Since the information in a single video is incomplete, 3D reconstruction requires the use of empirical knowledge. Multi-view 3D reconstruction (similar to human binocular positioning) is relatively easier. The method is to first calibrate the camera, that is, calculate the relationship between the camera's image coordinate system and the world coordinate system. Then, the 3D information is reconstructed using information from multiple 2D images.

[0082] In another specific embodiment, three-dimensional reconstruction is performed based on multiple first two-dimensional images to obtain three-dimensional models of multiple target cloud clusters in the sky region, including:

[0083] (1) Use the SFM algorithm to perform three-dimensional reconstruction on multiple first two-dimensional images to obtain initial point cloud data.

[0084] Structure From Motion (SFM) is a technique for estimating three-dimensional structures from a sequence of multiple two-dimensional images containing visual motion information. Specifically, the SFM algorithm includes the following steps: extracting feature points from the images, extracting key points, and calculating descriptors; matching features between pairs of images; obtaining the camera spatial pose transformation between pairs; and performing sparse reconstruction based on the matched feature points to obtain point cloud data.

[0085] (2) Select a first two-dimensional image with a preset ratio from multiple first two-dimensional images as a second two-dimensional image to obtain multiple second two-dimensional images, wherein the preset ratio is less than 1.

[0086] The preset ratio can be 0.5 or other values, depending on the specific situation.

[0087] In one specific embodiment, a predetermined proportion of first two-dimensional captured images are selected from a plurality of first two-dimensional captured images as second two-dimensional captured images, resulting in a plurality of second two-dimensional captured images, including:

[0088] Step 1-1: The first two-dimensional captured image is determined as the image to be detected. Optical flow calculation is performed based on the image to be detected and the historical two-dimensional captured images of the drone that captured the image to be detected before capturing the image to be detected, so as to obtain the optical flow vector of each pixel on the image to be detected.

[0089] In the real world, everything is in motion, and the speed and direction of this motion may differ, thus forming a motion field. The motion of an object is projected onto an image as the movement of pixels. Optical flow is the instantaneous velocity of pixels moving on the imaging plane of a moving object in space. This instantaneous velocity of pixel movement is optical flow. The optical flow method is a method that uses the temporal changes of pixels in an image sequence and the correlation between adjacent frames to find the correspondence between the previous frame and the current frame, and to calculate the motion information of objects between adjacent frames.

[0090] Optical flow is a method that uses the temporal changes of pixels in an image sequence and the correlation between adjacent frames to find the correspondence between the previous and current frames, thereby calculating the motion information of objects between adjacent frames. The instantaneous rate of change of grayscale at a specific coordinate point on a two-dimensional image plane is typically defined as the optical flow vector.

[0091] Optical flow methods can be categorized based on their implementation: gradient-based methods, matching-based methods, energy-based methods, and phase-based methods. Among these, gradient-based methods include Farnback optical flow. The underlying assumptions of optical flow methods include: constant brightness between adjacent frames; temporal continuity or minimal motion changes between adjacent frames; and pixels within the same sub-image exhibiting identical motion.

[0092] Steps 1-2: Calculate the magnitude and direction of the optical flow vector of each pixel in the image to be detected.

[0093] Specifically, the optical flow vector magnitude of each pixel is calculated based on each optical flow vector (ui, vi). The formula for calculating the optical flow vector magnitude magi of the i-th pixel is as follows:

[0094] magi = sqrt(ui^2 + vi^2),

[0095] Specifically, the optical flow vector direction of each pixel is calculated based on each optical flow vector (ui, vi). The formula for calculating the optical flow vector direction angi of the i-th pixel is as follows:

[0096] angi = arctan(vi / ui),

[0097] Steps 1-3: Calculate the average value of the optical flow vector of each pixel to obtain the average amplitude.

[0098] Specifically, the formula for calculating the average amplitude μ is as follows:

[0099] μ = mean(mag),

[0100] Here, mag represents the set of optical flow vector magnitudes magi of each pixel in the image to be detected.

[0101] Steps 1-4: Calculate the standard deviation of the optical flow vector magnitude of each pixel to obtain the direction change parameters.

[0102] Specifically, the formula for calculating the direction change parameter σ is as follows:

[0103] σ = std(mag),

[0104] Where μ represents the average amplitude and σ represents the direction change parameter. A smaller μ indicates a smaller motion amplitude, and a smaller σ indicates that the magnitudes of the vectors are more uniform.

[0105] Steps 1-5: Calculate the variance of the optical flow vector direction of each pixel to obtain the parameter of the degree of change in the vector direction.

[0106] Specifically, the formula for calculating the parameter varθ, which represents the degree of change in vector direction, is as follows:

[0107] varθ = var(ang),

[0108] Where ang represents the set of optical flow vector directions angi of each pixel in the image to be detected. varθ represents the degree of dispersion of the principal directions; a smaller varθ indicates that the principal directions are more concentrated, while a larger varθ indicates that the principal directions are more dispersed.

[0109] Steps 1-6 determine the image vibration parameters of the image to be detected based on the average amplitude, direction change parameters, and vector direction change degree parameters, thus obtaining the image vibration parameters of each first two-dimensional captured image.

[0110] Among them, the image vibration parameter decreases as the average amplitude increases, the image vibration parameter increases as the direction change parameter increases, and the image vibration parameter decreases as the vector direction change parameter increases.

[0111] In one specific embodiment, the image vibration parameter G score It satisfies the following calculation formula,

[0112] G score =exp(-μ / σ)*exp(-varθ),

[0113] Where μ represents the average amplitude, σ represents the direction change parameter, varθ represents the vector direction change parameter, and G score This represents the vibration parameters of the image.

[0114] Among them, the image vibration parameter G score The value ranges from 0 to 1, with values ​​closer to 1 indicating a more stable image.

[0115] Steps 1-7: The first two-dimensional image with image vibration parameters greater than the preset vibration parameters is determined as the second two-dimensional image.

[0116] Furthermore, multiple images captured by the same drone can be obtained.

[0117] (3) Perform cloud segmentation on multiple second-dimensional images to obtain multiple cloud segmentation regions on multiple second-dimensional images.

[0118] Specifically, a pre-trained cloud segmentation model is used to segment clouds in multiple second-dimensional images, resulting in multiple cloud segmentation regions on the multiple second-dimensional images.

[0119] (4) The point cloud data located within each cloud cluster segmentation region in the initial point cloud data is identified as the cloud cluster point cloud data.

[0120] Since the initial point cloud data is reconstructed from multiple second-dimensional images, each 3D point cloud in the initial point cloud data can find a corresponding mapped pixel in the multiple second-dimensional images. Specifically, the mapped pixels of each 3D point cloud in the initial point cloud data corresponding to the second-dimensional images are obtained, and the 3D point clouds whose mapped pixels are located within the segmentation regions of each cloud cluster are determined as the cloud cluster point cloud data.

[0121] (5) Determine three-dimensional models of multiple target cloud clusters based on cloud cluster point cloud data.

[0122] In one specific embodiment, the cloud point cloud data is clustered to obtain multiple first three-dimensional point clusters; based on the first three-dimensional point clusters, a target cloud three-dimensional model is determined, resulting in multiple target cloud three-dimensional models corresponding to the multiple first three-dimensional point clusters. Specifically, the solid formed by the enclosed space of multiple three-dimensional point clouds in the first three-dimensional point clusters is determined as the target cloud three-dimensional model.

[0123] In another specific embodiment, image matching is performed on cloud segmentation regions in multiple second-dimensional captured images to obtain multiple cloud segmentation regions of the same cloud cluster in different second-dimensional captured images. These multiple cloud segmentation regions of the same cloud cluster in different second-dimensional captured images are defined as a segmentation region group, resulting in multiple segmentation region groups. In the initial point cloud data, the solid formed by multiple 3D point clouds within the segmentation regions of multiple cloud clusters located within a segmentation region group is defined as a first cloud cluster 3D model, resulting in multiple first cloud cluster 3D models corresponding to multiple segmentation region groups. The number of segmentation region groups is determined to a preset number, and the cloud point cloud data is clustered based on this preset number to obtain a preset number of first 3D point clusters. Second cloud cluster 3D models are determined based on these first 3D point clusters, resulting in a preset number of second cloud cluster 3D models corresponding to the preset number of first 3D point clusters. Specifically, the solid formed by multiple 3D point clouds within the first 3D point clusters is defined as the second cloud cluster 3D model. The first cloud cluster 3D model and the second cloud cluster 3D model are matched to obtain a first cloud cluster 3D model and a corresponding second cloud cluster 3D model. The intersection of the first and second cloud cluster 3D models is used to determine the target cloud cluster 3D model, resulting in multiple target cloud cluster 3D models. Specifically, the union of the first and second cloud cluster 3D models is used to determine the target cloud cluster 3D model.

[0124] Image matching refers to identifying corresponding points between two or more images using a specific matching algorithm. For example, in two-dimensional image matching, the correlation coefficients of windows of the same size in the target region and the search region are compared, and the center point of the window with the highest correlation coefficient in the search region is taken as the corresponding point. Essentially, it is an optimal search problem that applies matching criteria under the condition of primitive similarity.

[0125] Specifically, based on a preset number of clusters of cloud point cloud data, a preset number of first three-dimensional point clusters are obtained, including:

[0126] (1) Obtain the point cloud features of H three-dimensional points in the cloud data. The point cloud features include the three-dimensional coordinates and color values ​​of the three-dimensional points.

[0127] (2) Divide the H three-dimensional points into Q sets of three-dimensional points, where each set of three-dimensional points contains H / Q three-dimensional points.

[0128] (3) Cluster the Q sets of three-dimensional points respectively to obtain a preset number of second three-dimensional point clusters in each set of three-dimensional points.

[0129] In one specific embodiment, the k-means clustering algorithm is used to cluster the Q sets of three-dimensional points respectively, so as to obtain a preset number of second three-dimensional point clusters in each set of three-dimensional points.

[0130] (4) Determine each second three-dimensional point cluster as the target three-dimensional point cluster, and calculate the three-dimensional point cluster similarity between the target three-dimensional point cluster and the second three-dimensional point cluster in the Q three-dimensional point set respectively.

[0131] Specifically, the similarity between the cluster average vector of the target 3D point cluster and the cluster average vector of the second 3D point cluster is calculated; the similarity between the cluster average vector of the target 3D point cluster and the cluster average vector of the second 3D point cluster is determined as the 3D point cluster similarity between the target 3D point cluster and the second 3D point cluster.

[0132] (5) Merge the second three-dimensional point clusters with the highest similarity to the target three-dimensional point clusters in each set of three-dimensional points to obtain the first three-dimensional point clusters corresponding to the target three-dimensional point clusters, and obtain the first three-dimensional point clusters corresponding to the preset number of second three-dimensional point clusters.

[0133] 204. Obtain the direction of sunlight at the target time.

[0134] In this embodiment, the direction of sunlight can be obtained from the weather forecast of the location of the distributed photovoltaic power station. The direction of sunlight is as follows: Figure 3 As shown.

[0135] 205. Project the three-dimensional models of the power station area and multiple target cloud clusters onto the first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the three-dimensional models of multiple target cloud clusters.

[0136] The first plane is perpendicular to the direction of sunlight. The first plane, the power station area, the power station projection area, the target cloud 3D model, and the cloud projection area are shown below. Figure 3 and Figure 4 As shown.

[0137] 206. Determine the overall influence parameters of the cloud cluster on the distributed photovoltaic power station based on the positional relationship between the cloud cluster projection area and the power station projection area.

[0138] In this embodiment, the determination of the overall cloud impact parameters of the distributed photovoltaic power station based on the positional relationship between the cloud projection area and the power station projection area of ​​the distributed photovoltaic power station includes: if there is an overlap between the power station projection area and the cloud projection area of ​​the target cloud 3D model, then the determination of the individual cloud impact parameters of the target cloud 3D model on the distributed photovoltaic power station is based on the overlap area and the cloud projection area of ​​the target cloud 3D model; if there is no overlap between the power station projection area and the cloud projection area, then the preset impact parameter value is determined as the individual cloud impact parameter of the target cloud 3D model on the distributed photovoltaic power station; and the average value of the individual cloud impact parameters of each target cloud 3D model on the distributed photovoltaic power station is determined as the overall cloud impact parameter of the distributed photovoltaic power station.

[0139] The preset influence parameter value can be 1 or other values, depending on the specific situation. The power station projection area, cloud projection area, and overlapping area are as follows: Figure 4 As shown.

[0140] In this embodiment of the application, the influence parameters of the target cloud cluster's 3D model on individual cloud clusters in a distributed photovoltaic power station are determined based on the overlapping region and the cloud cluster projection region of the target cloud cluster's 3D model, including:

[0141] (1) Obtain the area ratio of the overlapping area to the projected area of ​​the power station.

[0142] In this embodiment of the application, the area ratio is the ratio of the overlapping area to the projected area of ​​the power station.

[0143] (2) Generate multiple first straight lines that pass through the target cloud 3D model and overlapping area and are parallel to the direction of sunlight.

[0144] In this embodiment, the centroid of the overlapping region is obtained, and multiple rays are generated starting from the centroid of the overlapping region. These rays intersect the boundary of the overlapping region at multiple intersection points, resulting in multiple second line segments starting from the centroid of the overlapping region and ending at multiple intersection points. Multiple equidistant points on the second line segments are obtained, and a straight line parallel to the direction of sunlight is drawn through each equidistant point on each second line segment, resulting in multiple first straight lines.

[0145] (3) Cut the first straight line on the surface of the target cloud three-dimensional model to generate the first line segment, and obtain the first line segment of multiple first straight lines.

[0146] (4) Determine the influence parameters of the target cloud cluster three-dimensional model on the cloud cluster of the distributed photovoltaic power station based on the regional area ratio and the average length of multiple first line segments.

[0147] Among them, the larger the average length of the multiple first line segments, the greater the influence parameter of the cloud cluster; the larger the area ratio, the smaller the influence parameter of the cloud cluster.

[0148] In one specific embodiment, the influence parameters of the target cloud cluster 3D model on the cloud cluster of the distributed photovoltaic power station are determined based on the regional area ratio and the average length of multiple first line segments. This includes: normalizing the average length of multiple first line segments to obtain a normalized average length; weighting the regional area ratio and the normalized average length to obtain a weighted average; and multiplying the weighted average with a preset influence parameter value and adding the preset influence parameter value to obtain the cloud cluster individual influence parameters.

[0149] In another specific embodiment, the influence parameters of the target cloud cluster 3D model on the individual cloud clusters of the distributed photovoltaic power station are determined based on the area ratio of the region and the average length of multiple first line segments, including:

[0150] (1) Obtain the centroid distance between the centroid of the overlapping region and the centroid of the cloud projection region.

[0151] (2) The influence parameters of the target cloud cluster three-dimensional model on the cloud cluster of the distributed photovoltaic power station are determined based on the regional area ratio, centroid spacing and the average length of multiple first line segments.

[0152] Among them, the larger the average length of multiple first line segments, the greater the influence parameter of a single cloud cluster; the larger the area ratio of the region, the greater the influence parameter of a single cloud cluster; and the larger the centroid spacing, the smaller the influence parameter of a single cloud cluster.

[0153] Specifically, the average length of multiple first line segments is normalized to obtain a normalized average length. The reciprocal of the average length of multiple first line segments is determined as the reciprocal of the average spacing. The reciprocal of the average spacing is normalized to obtain a normalized average of the reciprocal spacing. The normalized average of the reciprocal spacing, the area ratio, and the normalized average length are weighted to obtain a weighted average. The weighted average is multiplied by a preset influence parameter value and added to the preset influence parameter value to obtain the cloud cluster individual influence parameter.

[0154] Specifically, the cloud cluster's individual influence parameter P2 satisfies the following formula:

[0155] P2 = (a1*L + a2*A + a3*J + 1)*P1,

[0156] Wherein, P1 is the preset influence parameter value, which can be determined based on human experience; L is the normalized length average; a1 is the weight of the normalized length average L; A is the area ratio of the region; a2 is the weight of the area ratio of the region A; J is the inverse average of the normalized interval; and a3 is the weight of the inverse average of the normalized interval J.

[0157] 207. Determine the predicted power generation of a distributed photovoltaic power station based on the overall influence parameters of cloud clusters on the distributed photovoltaic power station.

[0158] In this embodiment of the application, determining the predicted power generation of a distributed photovoltaic power station based on the overall cloud impact parameters of the distributed photovoltaic power station includes:

[0159] (1) Calculate the overall impact parameters of cloud clusters at multiple different times within a preset time period for distributed photovoltaic power generation stations.

[0160] Specifically, different times were determined as target times to obtain the overall impact parameters of cloud clusters at multiple different times.

[0161] The multiple different times can be multiple time points arranged at preset time intervals. For example, if the preset time interval is 1 minute, the multiple time points are 12:01, 12:02, 12:03, 12:04, etc. For example, the overall impact parameters of the cloud clusters at multiple different times are 1, 1.1, 1.2, 1.2, and 1.5, respectively.

[0162] (2) Obtain meteorological parameters of the distributed photovoltaic power station at multiple different times within a preset time period, wherein the meteorological parameters include at least one of temperature, wind speed and humidity.

[0163] Specifically, meteorological parameters can be obtained through weather forecasts.

[0164] (3) The overall impact parameters of cloud clusters at multiple different times and meteorological parameters at multiple different times are spliced ​​together to obtain input parameters at multiple different times.

[0165] The output parameters for each time period include meteorological parameters and overall cloud impact parameters.

[0166] (4) Input multiple input parameters at different times into the preset power generation prediction model to obtain the predicted power generation of the distributed photovoltaic power station.

[0167] The preset power generation prediction model can be a pre-trained Long Short Term Memory (LSTM) network. Input parameters from multiple different times are input as a sequence of data into the preset power generation prediction model to obtain the predicted power generation of the distributed photovoltaic power station. LSTM is an improved recurrent neural network that can solve the problem of RNNs' inability to handle long-range dependencies.

[0168] To facilitate better implementation of the distributed renewable energy power generation data prediction method provided in this application embodiment, this application embodiment also provides a distributed renewable energy power generation data prediction device based on the above-described distributed renewable energy power generation data prediction method. The meanings of the terms used are the same as in the above-described distributed renewable energy power generation data prediction method; for specific implementation details, please refer to the descriptions in the above method embodiments.

[0169] Please refer to Figure 5 , Figure 5 This is a schematic diagram of the structure of a distributed renewable energy power generation data prediction device provided in an embodiment of this application. The distributed renewable energy power generation data prediction device may include:

[0170] The first acquisition module 701 is used to acquire the power station area where the distributed photovoltaic power station is located.

[0171] The second acquisition module 702 is used to acquire multiple first two-dimensional images of the sky area taken by different drones from different angles at the target time. Among them, the distributed photovoltaic power station corresponds to a drone cluster, and the drone cluster flies and takes pictures in the area centered on the distributed power station.

[0172] The 3D reconstruction module 703 is used to perform 3D reconstruction based on multiple first 2D captured images to obtain 3D models of multiple target cloud clusters in the sky region.

[0173] The third acquisition module 704 is used to acquire the direction of sunlight at the target time;

[0174] The projection module 705 is used to project the three-dimensional models of the power station area and multiple target cloud clusters onto the first plane respectively, so as to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the three-dimensional models of multiple target cloud clusters. The first plane is perpendicular to the direction of sunlight.

[0175] The first determining module 706 is used to determine the overall influence parameters of the cloud cluster of the distributed photovoltaic power station based on the positional relationship between the cloud cluster projection area and the power station projection area of ​​the distributed photovoltaic power station.

[0176] The second determining module 707 is used to determine the predicted power generation of the distributed photovoltaic power station based on the overall influence parameters of the cloud clusters of the distributed photovoltaic power station.

[0177] For details on the implementation of each of the above modules, please refer to the previous examples, which will not be repeated here.

[0178] This application also provides an electronic device, including a memory and a processor, wherein the processor executes the steps in the distributed new energy power generation data prediction method provided in this embodiment by calling a computer program stored in the memory.

[0179] Please refer to Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0180] The electronic device may include components such as a processor 101 with one or more processing cores, a memory 102 with one or more computer-readable storage media, a power supply 103, and an input unit 104. Those skilled in the art will understand that the electronic device structure shown in the figures does not constitute a limitation on the electronic device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0181] The processor 101 is the control center of the electronic device, connecting various parts of the device via various interfaces and lines. It executes software programs and / or modules stored in the memory 102, and calls data stored in the memory 102, to perform various functions and process data. Optionally, the processor 101 may include one or more processing cores; alternatively, the processor 101 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 101.

[0182] The memory 102 can be used to store software programs and modules. The processor 101 executes various functional applications and data processing by running the software programs and modules stored in the memory 102. The memory 102 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 102 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 102 may also include a memory controller to provide the processor 101 with access to the memory 102.

[0183] The electronic device also includes a power supply 103 that supplies power to the various components. Optionally, the power supply 103 can be logically connected to the processor 101 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 103 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0184] The electronic device may also include an input unit 104, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0185] Although not shown, the electronic device may also include a display unit, an image acquisition component, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 101 in the electronic device will load one or more executable codes corresponding to computer programs into the memory 102 according to the following instructions, and the processor 101 will execute the steps in the distributed new energy power generation data prediction method provided in this application, such as:

[0186] The process involves: acquiring the location of the distributed photovoltaic (PV) power station; obtaining multiple first-dimensional (2D) images of the sky region taken by different drones from different angles at a target time, where each PV power station corresponds to a drone swarm, and each drone swarm flies and takes pictures within an area centered on the PV power station; performing 3D reconstruction based on the multiple first-dimensional images to obtain 3D models of multiple target cloud clusters in the sky region; obtaining the direction of sunlight illumination at the target time; projecting the 3D models of the power station region and the multiple target cloud clusters onto a first plane to obtain the power station projection area corresponding to the power station region and the multiple cloud cluster projection areas corresponding to the multiple target cloud cluster 3D models, where the first plane is perpendicular to the direction of sunlight illumination; determining the overall cloud cluster influence parameters of the distributed PV power station based on the positional relationship between the cloud cluster projection areas and the power station projection areas of the distributed PV power station; and determining the predicted power generation of the distributed PV power station based on the overall cloud cluster influence parameters of the distributed PV power station.

[0187] It should be noted that the electronic device provided in this application embodiment and the distributed new energy power generation data prediction method in the above embodiment belong to the same concept. The specific implementation process can be found in the above related embodiments, and will not be repeated here.

[0188] This application also provides a computer-readable storage medium storing a computer program thereon. When the computer program stored thereon is executed on the processor of the electronic device provided in the embodiments of this application, the processor of the electronic device performs the steps in the distributed new energy power generation data prediction method provided in this application. The storage medium may be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0189] This application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform various optional implementations of the aforementioned distributed renewable energy power generation data prediction method.

[0190] The above provides a detailed description of a distributed new energy power generation data prediction method, apparatus, equipment, and medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

[0191] It should be noted that when the above embodiments of this application are applied to specific products or technologies, and user-related data is involved, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

Claims

1. A method for predicting power generation data from distributed new energy sources, characterized in that, The method for predicting power generation data from distributed new energy sources includes: Obtain the location of the distributed photovoltaic power station; Multiple first two-dimensional images of the sky region taken by different drones from different angles at a target time are acquired, wherein the distributed photovoltaic power station corresponds to a drone swarm, and each drone swarm flies and takes pictures within an area centered on the distributed photovoltaic power station; Three-dimensional reconstruction is performed based on multiple first two-dimensional images to obtain three-dimensional models of multiple target cloud clusters in the sky region; Obtain the direction of sunlight at the target time; The three-dimensional models of the power station area and the multiple target cloud clusters are respectively projected onto the first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the multiple target cloud clusters three-dimensional models, wherein the first plane is perpendicular to the sunlight direction; The overall cloud impact parameters of the distributed photovoltaic power station are determined based on the positional relationship between the cloud projection area and the power station projection area. Specifically, if the power station projection area overlaps with the cloud projection area of ​​the target cloud 3D model, the individual cloud impact parameters of the target cloud 3D model on the distributed photovoltaic power station are determined based on the overlapping area and the cloud projection area of ​​the target cloud 3D model. If the power station projection area and the cloud projection area do not overlap, a preset impact parameter value is determined as the individual cloud impact parameter of the target cloud 3D model on the distributed photovoltaic power station. The average value of the individual cloud impact parameters of each target cloud 3D model on the distributed photovoltaic power station is determined as the overall cloud impact parameter of the distributed photovoltaic power station. The predicted power generation of the distributed photovoltaic power station is determined based on the overall cloud impact parameters of the distributed photovoltaic power station. This involves calculating the overall cloud impact parameters of the distributed photovoltaic power station at multiple different times within a preset time period; acquiring meteorological parameters of the distributed photovoltaic power station at multiple different times within the preset time period, wherein the meteorological parameters include at least one of temperature, wind speed, and humidity; concatenating the overall cloud impact parameters and the meteorological parameters at multiple different times to obtain multiple input parameters at different times; and inputting the multiple input parameters at different times into a preset power generation prediction model to obtain the predicted power generation of the distributed photovoltaic power station.

2. The method for predicting power generation data of distributed new energy sources according to claim 1, characterized in that, The determination of the cloud cluster individual cloud cluster impact parameters of the target cloud cluster 3D model on the distributed photovoltaic power station based on the cloud cluster projection area of ​​the overlapping region and the target cloud cluster 3D model includes: Obtain the area ratio of the overlapping region to the projected area of ​​the power station; Generate multiple first straight lines that pass through the target cloud 3D model and the overlapping area and are parallel to the direction of sunlight; The first straight line is truncated on the surface of the three-dimensional model of the target cloud to generate the first line segment, resulting in the first line segment of multiple first straight lines; The influence parameters of the target cloud cluster 3D model on the cloud cluster individual of the distributed photovoltaic power station are determined based on the area ratio of the region and the average length of multiple first line segments. The larger the average length of multiple first line segments, the larger the influence parameter of the cloud cluster individual; the larger the area ratio of the region, the larger the influence parameter of the cloud cluster individual.

3. The method for predicting power generation data of distributed new energy sources according to claim 2, characterized in that, The method of determining the influence parameters of the target cloud cluster 3D model on the individual cloud clusters of the distributed photovoltaic power station based on the area ratio of the region and the average length of multiple first line segments includes: Obtain the centroid distance between the centroid of the overlapping region and the centroid of the cloud projection region; The influence parameters of the target cloud cluster 3D model on the cloud cluster individual of the distributed photovoltaic power station are determined based on the area ratio of the region, the centroid spacing and the average length of multiple first line segments. The larger the average length of multiple first line segments, the larger the influence parameter of the cloud cluster individual; the larger the area ratio of the region, the larger the influence parameter of the cloud cluster individual; and the larger the centroid spacing, the smaller the influence parameter of the cloud cluster individual.

4. The method for predicting power generation data of distributed new energy sources according to claim 1, characterized in that, The process of reconstructing three-dimensional images based on multiple first two-dimensional images to obtain three-dimensional models of multiple target cloud clusters in the sky region includes: The SFM algorithm is used to reconstruct three-dimensional images from multiple first two-dimensional images to obtain initial point cloud data. Select a predetermined ratio of first two-dimensional captured images from a plurality of first two-dimensional captured images as second two-dimensional captured images to obtain a plurality of second two-dimensional captured images, wherein the predetermined ratio is less than 1; Cloud segmentation is performed on multiple second-two-dimensional captured images to obtain multiple cloud segmentation regions on the multiple second-two-dimensional captured images; The point cloud data located within each of the cloud cluster segmentation regions in the initial point cloud data are defined as cloud cluster point cloud data; Based on the cloud point cloud data, multiple three-dimensional models of the target cloud clusters are determined.

5. The method for predicting power generation data of distributed new energy sources according to claim 4, characterized in that, The generation of multiple 3D models of the target cloud based on the cloud point cloud data includes: Clustering the cloud point cloud data yields multiple first three-dimensional point clusters; Based on the first three-dimensional point cluster, the three-dimensional model of the target cloud is determined, and multiple three-dimensional models of the target cloud corresponding to multiple first three-dimensional point clusters are obtained.

6. A distributed renewable energy power generation data prediction device, characterized in that, include: The first acquisition module is used to acquire the power station area where the distributed photovoltaic power station is located; The second acquisition module is used to acquire multiple first two-dimensional images of the sky area taken by different drones from different angles at the target time. The distributed photovoltaic power station corresponds to a drone swarm, and the drone swarm flies and takes pictures within the area centered on the distributed photovoltaic power station. The three-dimensional reconstruction module is used to perform three-dimensional reconstruction based on multiple first two-dimensional captured images to obtain three-dimensional models of multiple target cloud clusters in the sky region; The third acquisition module is used to acquire the direction of sunlight at the target time; The projection module is used to project the power station area and the multiple target cloud cluster 3D models onto a first plane to obtain the power station projection area corresponding to the power station area and the multiple cloud cluster projection areas corresponding to the multiple target cloud cluster 3D models, wherein the first plane is perpendicular to the sunlight direction. The first determining module is used to determine the overall cloud impact parameters of the distributed photovoltaic power station based on the positional relationship between the cloud projection area and the power station projection area of ​​the distributed photovoltaic power station; wherein, if the power station projection area overlaps with the cloud projection area of ​​the target cloud 3D model, the target cloud 3D model's individual cloud impact parameters on the distributed photovoltaic power station are determined based on the overlapping area and the target cloud 3D model's individual cloud projection area; if the power station projection area and the cloud projection area do not overlap, a preset impact parameter value is determined as the target cloud 3D model's individual cloud impact parameters on the distributed photovoltaic power station; and the average value of the individual cloud impact parameters of each target cloud 3D model on the distributed photovoltaic power station is determined as the overall cloud impact parameters of the distributed photovoltaic power station. The second determining module is used to determine the predicted power generation of the distributed photovoltaic power station based on the overall cloud impact parameters of the distributed photovoltaic power station. Specifically, it calculates the overall cloud impact parameters of the distributed photovoltaic power station at multiple different times within a preset time period; acquires meteorological parameters of the distributed photovoltaic power station at multiple different times within the preset time period, wherein the meteorological parameters include at least one of temperature, wind speed, and humidity; concatenates the overall cloud impact parameters and the meteorological parameters at multiple different times to obtain multiple input parameters at different times; and inputs the multiple input parameters at different times into a preset power generation prediction model to obtain the predicted power generation of the distributed photovoltaic power station.

7. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor running the computer program in the memory to perform the steps in the distributed new energy power generation data prediction method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to execute the steps in the distributed renewable energy power generation data prediction method according to any one of claims 1 to 5.

Images