CN116780533B - Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium - Google Patents

Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN116780533B
CN116780533B CN202311027499.0A CN202311027499A CN116780533B CN 116780533 B CN116780533 B CN 116780533B CN 202311027499 A CN202311027499 A CN 202311027499A CN 116780533 B CN116780533 B CN 116780533B
Authority
CN
China
Prior art keywords
cloud
forecasting
cloud image
result
module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311027499.0A
Other languages
Chinese (zh)
Other versions
CN116780533A (en
Inventor
宋美洋
刘鲁宁
郭炜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing East Environment Energy Technology Co ltd
Original Assignee
Beijing East Environment Energy Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing East Environment Energy Technology Co ltd filed Critical Beijing East Environment Energy Technology Co ltd
Priority to CN202311027499.0A priority Critical patent/CN116780533B/en
Publication of CN116780533A publication Critical patent/CN116780533A/en
Application granted granted Critical
Publication of CN116780533B publication Critical patent/CN116780533B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
    • G06V10/809—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of classification results, e.g. where the classifiers operate on the same input data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/13—Satellite images
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Remote Sensing (AREA)
  • Astronomy & Astrophysics (AREA)
  • Power Engineering (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The embodiment of the application provides a photovoltaic ultra-short-term forecasting method, a device, electronic equipment and a storage medium, wherein the photovoltaic ultra-short-term forecasting method comprises the following steps: acquiring foundation cloud image data and satellite cloud image data; inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result; inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result; inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture; inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image; and inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result.

Description

Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium
Technical Field
The invention relates to the technical field of new energy power generation, in particular to a photovoltaic ultra-short-term forecasting method, a device, electronic equipment and a storage medium.
Background
With the increasing of the installation scale of new energy sources nationwide, the impact of the intermittence, randomness and fluctuation of new energy source power generation on the stability of the power grid is more obvious, and the requirements of power dispatching of each province on the accuracy of power prediction are more stringent; in a photovoltaic power generation system, the output power of the system is greatly dependent on the solar radiation amount received by a solar panel, and the solar radiation amount is related to the cloud amount of an area where a photovoltaic station is located, so that the integral cloud characteristics of the area where the photovoltaic station is located and the cloud shielding of a moving track are necessary to be predicted.
However, existing prediction of cloud amount for an area where a photovoltaic station is located generally uses a ground-based cloud image alone or uses a satellite cloud image alone for cloud image identification; according to the prediction method, the foundation cloud image has higher time resolution, and the image is updated once in a shorter time interval, so that the foundation cloud image can better identify the bottom characteristics of the cloud cluster, and the rapid change and evolution of the cloud body can be captured; however, due to the limitation of the geographic position of the ground cloud picture, the prediction accuracy of the cloud cluster moving track is lower; satellite clouds can provide a wider geographic coverage, including land and sea, which enables cloud movement predictions to be made over a larger range, and cloud evolution to be observed for different areas; however, due to the limited cloud of satellite cloud image space-time resolution, cloud occlusion prediction is poor; therefore, the scheme of carrying out cloud image identification by independently adopting the foundation cloud image or the satellite cloud image cannot simultaneously give consideration to high prediction precision for cloud movement tracks and high prediction precision for cloud shielding.
Therefore, it is currently needed to propose a photovoltaic ultra-short-term forecasting method, device, electronic equipment and storage medium, so as to solve the technical problem that a scheme for identifying cloud patterns by solely adopting a ground cloud pattern or a satellite cloud pattern in the related technology cannot simultaneously consider high prediction precision for cloud movement tracks and high precision prediction for cloud shielding.
Disclosure of Invention
The embodiment of the application provides a photovoltaic ultra-short-term forecasting method, a device, electronic equipment and a storage medium, which are used for solving the technical problem that a scheme for identifying a cloud pattern by singly adopting a ground cloud pattern or a satellite cloud pattern in the related technology cannot simultaneously consider high forecasting precision aiming at a cloud cluster moving track and high forecasting precision aiming at cloud cluster shielding.
According to a first aspect of the present application, there is provided a photovoltaic ultrashort-term forecasting method comprising: acquiring foundation cloud image data and satellite cloud image data; inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result; inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result; inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture; inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image; and inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result.
Optionally, the photovoltaic ultra-short term forecasting method further comprises: inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model for forecasting to obtain a second shielding forecasting result and shielding forecasting accuracy based on the satellite cloud picture; inputting the prediction result of the foundation cloud image into a pre-trained cloud occlusion prediction model for prediction to obtain a second movement prediction result and movement prediction accuracy based on the foundation cloud image; fusing the first movement prediction result based on the second movement prediction result and the movement prediction accuracy; and fusing the first shielding prediction result based on the second shielding prediction result and the shielding prediction accuracy.
Optionally, the method further comprises: the accuracy of the movement prediction is positively correlated with the fusion weight of the second movement prediction result; and the shielding prediction accuracy is positively correlated with the fusion weight of the second shielding prediction result.
Optionally, the photovoltaic ultra-short term prediction model includes a cloud chart irradiance correction module and a photovoltaic ultra-short term prediction module, and the inputting the first movement prediction result and the first shielding prediction result into the photovoltaic ultra-short term prediction model trained in advance, and obtaining the photovoltaic ultra-short term prediction result includes: inputting the first movement prediction result and the first shielding prediction result to the cloud picture irradiance correction module to respectively obtain irradiance after correction of the foundation cloud picture characteristics and irradiance after correction of the satellite cloud picture characteristics; performing weighted fusion based on irradiance after the foundation cloud image characteristic correction and deviation between irradiance and actual irradiance after the satellite cloud image characteristic correction to obtain a radiation forecast result; the irradiance after the foundation cloud picture characteristic correction and the irradiance after the satellite cloud picture characteristic correction are input to the photovoltaic ultra-short-term forecasting module to obtain a photovoltaic ultra-short-term forecasting result; the photovoltaic ultra-short-term forecasting module comprises a plurality of photovoltaic ultra-short-term forecasting sub-modules.
Optionally, the photovoltaic ultra-short term prediction model includes a cloud shadow analysis module, a cloud dynamic light intensity module and a radiation prediction module, and the inputting the first movement prediction result and the first shielding prediction result into the photovoltaic ultra-short term prediction model trained in advance, to obtain a photovoltaic ultra-short term prediction result includes: calculating altitude change information and azimuth change information of a radiation forecast place of the sun within a preset time period based on the cloud shadow analysis module; inputting the first movement prediction result and the first shielding prediction result to the cloud shadow analysis module, and calculating movement track information and movement duration information of cloud shadows based on the altitude change information and the azimuth change information; inputting the movement track information and the movement duration information into a cloud dynamic light intensity module, and calculating shielding influence information of cloud shadows on ground light intensity based on the Yun Dongtai light intensity module; and obtaining a radiation forecasting result by adopting a pre-trained cloud radiation forecasting model based on the shielding influence information.
Optionally, the ground cloud image forecasting model includes a ground cloud image feature extraction module and a ground cloud image motion trend determination module, and the inputting the ground cloud image data into the pre-trained ground cloud image forecasting model to obtain a ground cloud image forecasting result includes: inputting the foundation cloud image data into a pre-trained foundation cloud image forecast model, and extracting foundation cloud image motion characteristics and foundation cloud image morphological characteristics based on the foundation cloud image characteristic extraction module; the foundation cloud image movement characteristics comprise the position, the speed and the direction of a cloud body, and the foundation cloud image morphological characteristics comprise the shape and the size of the cloud body; and predicting based on the foundation cloud picture motion trend determining module, the foundation cloud picture motion characteristics and the foundation cloud picture morphological characteristics to obtain the foundation cloud picture motion trend based on the foundation cloud picture within a preset time length in the future.
Optionally, the satellite cloud image forecasting model includes a satellite cloud image feature extraction module, a satellite cloud image classification module and a satellite cloud image movement prediction module, and the inputting the ground cloud image data into a pre-trained ground cloud image forecasting model for forecasting, and obtaining a first shielding forecasting result based on the ground cloud image includes: inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model, and extracting satellite cloud image features based on the satellite cloud image feature extraction module, wherein the satellite cloud image features comprise the size, the shape and the texture of a cloud body; performing cloud classification based on the satellite cloud image characteristics and the satellite cloud image classification module to obtain classification results; and carrying out satellite cloud picture prediction based on the classification result and the satellite cloud picture movement prediction module to obtain satellite cloud picture movement trend based on the satellite cloud pictures in the future preset time length.
According to a second aspect of the present application, there is provided a photovoltaic ultrashort-term forecasting device comprising: the data acquisition module is used for acquiring the ground cloud image data and the satellite cloud image data; the foundation cloud picture forecasting module is used for inputting the foundation cloud picture data into a pre-trained foundation cloud picture forecasting model to obtain a foundation cloud picture forecasting result; the satellite cloud image forecasting module is used for inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result; the mobile prediction result calculation module is used for inputting the satellite cloud image prediction result into a pre-trained cloud body mobile prediction model to predict, so as to obtain a first mobile prediction result based on the satellite cloud image; the shielding prediction result calculation module is used for inputting the foundation cloud image prediction result into a pre-trained cloud body shielding prediction model to predict, so as to obtain a first shielding prediction result based on the foundation cloud image; and the photovoltaic ultra-short-term forecasting module is used for inputting the first movement forecasting result and the first shielding forecasting result into a pre-trained photovoltaic ultra-short-term forecasting model to obtain a photovoltaic ultra-short-term forecasting result.
According to a third aspect of the present application, there is provided an electronic device comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the photovoltaic ultra-short term forecasting method of any one of the above.
According to a fourth aspect of the present application there is provided a computer readable storage medium having a computer program stored therein, wherein the computer program is arranged to perform the photovoltaic ultra-short term forecasting method as described in any one of the preceding claims when run.
The embodiment of the application provides a photovoltaic ultra-short-term forecasting method, which comprises the following steps of: acquiring foundation cloud image data and satellite cloud image data; inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result; inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result; inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture; inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image; inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result; according to the technical scheme, when the photovoltaic ultra-short-term radiation prediction is carried out, on one hand, aiming at the influence of cloud image shielding prediction on the photovoltaic ultra-short-term prediction, a foundation cloud image prediction result is input into a pre-trained cloud body shielding prediction model for prediction, so that shielding prediction with higher foundation cloud image precision is considered; on the other hand, the influence of cloud image movement prediction on photovoltaic ultra-short-term prediction is carried out, and meanwhile, the foundation cloud image prediction result is input into a cloud body shielding prediction model trained in advance for prediction, so that movement prediction with high satellite cloud image precision is considered, and the technical problem that a scheme of carrying out cloud image recognition by singly adopting foundation cloud images or satellite cloud images in the related technology cannot simultaneously consider high prediction precision on cloud image movement tracks and high precision prediction on cloud image shielding is solved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a schematic flow chart of an alternative photovoltaic ultra-short term forecasting method according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a photovoltaic ultra-short term forecast apparatus, according to an embodiment of the present application;
fig. 3 is a block diagram of an alternative electronic device in accordance with an embodiment of the present application.
Detailed Description
For a clearer understanding of the technical features, objects and effects of the present application, embodiments of the present application will now be described with reference to the drawings, in which like reference numerals refer to identical or structurally similar but functionally identical components throughout the separate views.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application, however, the present application may be practiced in other ways than those described herein, and therefore the scope of the present application is not limited to the specific embodiments disclosed below.
In the following description, various aspects of the present application will be described, however, it will be apparent to those skilled in the art that the present application may be practiced with only some or all of the structures or processes of the present application. For purposes of explanation, specific numbers, configurations and orders are set forth, it is apparent that the application may be practiced without these specific details. In other instances, well-known features will not be described in detail so as not to obscure the application.
According to an aspect of the present application, a photovoltaic ultra-short term forecasting method is provided, and fig. 1 is a schematic flow chart of an alternative photovoltaic ultra-short term forecasting method according to an embodiment of the present application, as shown in fig. 1, where the photovoltaic ultra-short term forecasting method includes:
s100, acquiring foundation cloud image data and satellite cloud image data.
For example, the ground cloud image data may be acquired by a ground based device such as an all-sky imager; for example, as an exemplary embodiment, the collecting of the ground cloud image as the ground cloud image data may be performed by the ground equipment at preset time intervals; wherein the preset time interval may be 5min, 10min, 15min, etc.; the satellite cloud image data can be downloaded by a weather website; for example, the satellite cloud map data may be the observed data of a wind cloud series meteorological satellite downloaded by the chinese meteorological office.
S200, inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result.
The foundation cloud image forecasting model is obtained by training historical foundation cloud image data and historical meteorological data and is used for forecasting the input foundation cloud image data to obtain a foundation cloud image forecasting result; the foundation cloud image forecasting result comprises the shape, speed and direction of the cloud cluster under the selected fixed representative speed or the shape, speed and direction of the cloud cluster under the selected self-adaptive representative speed.
Specifically, in the model training process, based on the foundation cloud image prediction model, the corresponding relation between the foundation cloud image data and the change trend of the data and cloud parameters such as cloud cover, cloud height, cloud thickness and the like of the cloud cluster is learned, and the corresponding relation between the cloud cluster parameters and meteorological parameters such as temperature, air pressure, humidity and wind speed is further learned. The learning of the corresponding relation between the foundation cloud image data and the cloud amount of the cloud cluster based on the foundation cloud image forecasting model can be realized by adjusting parameters in image contour and image filtering edge detection; specifically, as a possible implementation manner, in the training process, a cloud image is converted into a gray image, and then a detection threshold value is continuously adjusted until the image can be divided into a cloud part and a non-cloud part under a certain detection threshold value, wherein the outline of the cloud is the boundary of the two parts. As another possible implementation manner, a Canny algorithm can be used to detect the edges of the cloud image, and algorithm parameters of the Canny algorithm are continuously adjusted until the contours of the image can be obtained; as another possible implementation manner, the Sobel operator may also be used to perform edge detection on the image, and identify the cloud amount.
For example, in the training process, two adjacent cloud images are adopted for matching, the same cloud speed is found, and the corresponding relation between the two cloud images is determined; calculating the moving speed of cloud blocks in two adjacent cloud pictures by using a pixel matching method or a wavelet transformation method or a phase correlation method according to the matching result; wherein the displacement includes a horizontal direction sub-movement speed and a vertical direction sub-movement speed; and finally, calculating to obtain the velocity vector of the cloud block according to the displacement of the cloud block by using displacement field differentiation and other methods.
In the training process, as a possible implementation manner, a fixed speed range analysis can be selected for identifying the foundation cloud image, and as an example, a cloud body with a horizontal speed of 10-20 m/s can be selected for research, and a horizontal speed selection manner can be adopted. As another possible implementation, adaptive speed selection may also be performed: in this embodiment, the speed range is adaptively selected according to the motion characteristics of the cloud body, and according to the average speed and the speed distribution of the cloud body.
S300, inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result.
In this embodiment, the satellite cloud image prediction model may be a pre-trained deep learning prediction model; the satellite cloud image forecasting model is obtained by training historical satellite cloud image data and is used for forecasting the satellite cloud image forecasting result through the input satellite cloud image data at the current moment; the satellite cloud image forecasting result comprises cloud cluster physical parameters such as size, shape and texture of a selected cloud cluster and cloud cluster classification parameters such as high cloud, low cloud, rolling cloud and cloud accumulation of the cloud cluster.
In the model training process, the corresponding relation between the satellite cloud image data and cloud cluster parameters such as the size, the shape, the texture and the like of the cloud cluster is learned based on the satellite cloud image forecasting model, and the corresponding relation between the satellite cloud image data and cloud cluster classification parameters such as high cloud, low cloud, rolling cloud, cloud accumulation and the like of the cloud cluster is learned based on the satellite cloud image forecasting model.
S400, inputting the satellite cloud image forecasting result into a pre-trained cloud body movement forecasting model to forecast, and obtaining a first movement forecasting result based on the satellite cloud image.
The cloud cover identification is usually carried out by using a foundation cloud cover alone or using a satellite cloud cover alone for the prediction of the cloud cover of the area where the photovoltaic station is located; for the prediction method, satellite cloud images can provide wider geographic coverage, including land and sea, so that cloud movement prediction can be performed in a wider range, and cloud evolution of different areas can be observed; however, due to the limited cloud of satellite cloud image space-time resolution, cloud occlusion prediction is poor; based on the method, the influence of the cloud picture movement prediction on the photovoltaic ultra-short-term prediction is considered, and the satellite cloud picture prediction result is considered to carry out movement prediction; based on the cloud image prediction result, the satellite cloud image prediction result is input into a pre-trained cloud body movement prediction model to be predicted, and a first movement prediction result based on the satellite cloud image is obtained.
S500, inputting the prediction result of the foundation cloud image into a pre-trained cloud occlusion prediction model for prediction to obtain a first occlusion prediction result based on the foundation cloud image;
the foundation cloud image has higher time resolution, and the image is updated once in a shorter time interval, so that the foundation cloud image can better identify the bottom characteristics of the cloud cluster and capture the rapid change and evolution of the cloud body; based on the method, aiming at the influence of cloud image shielding prediction on photovoltaic ultra-short-term prediction, the movement prediction is carried out by considering the foundation cloud image prediction result; based on the first occlusion prediction result, the foundation cloud image prediction result is input into a pre-trained cloud occlusion prediction model to be predicted, and the first occlusion prediction result based on the foundation cloud image is obtained.
S600, inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result.
In this embodiment, the photovoltaic ultrashort-term prediction model may be a combination of a plurality of algorithm modules, and specifically, the photovoltaic ultrashort-term prediction model may include a cloud pattern irradiance correction module and an ultrashort-term prediction module; the cloud picture irradiance correction module is used for correcting irradiance of the foundation cloud picture based on the first movement prediction result to obtain irradiance after correction of the characteristics of the foundation cloud picture, and correcting irradiance of the satellite cloud picture based on the first shielding prediction result to obtain irradiance after correction of the satellite cloud picture. The ultra-short-term radiation prediction module comprises a plurality of photovoltaic ultra-short-term prediction sub-modules, and is used for carrying out photovoltaic ultra-short-term prediction by adopting the plurality of photovoltaic ultra-short-term prediction sub-modules according to irradiance after the characteristic correction of the foundation cloud image and irradiance after the correction based on the satellite cloud image, so as to obtain a photovoltaic ultra-short-term prediction result.
As an exemplary embodiment, the cloud image irradiance correction module may include a cloud shadow analysis module, a cloud dynamic light intensity module, and a radiation prediction module, so as to obtain irradiance after the characteristic correction of the cloud image of the foundation and irradiance after the correction of the cloud image of the satellite sequentially through the cloud shadow analysis module and the cloud dynamic light intensity module based on the first movement prediction result and the first shielding prediction result.
The algorithm selection unit is used for performing algorithm selection based on respective accuracy of irradiance after correction of the ground cloud image characteristics and irradiance after correction of the satellite cloud image characteristics in the input data.
Based on the first movement prediction result and the first shielding prediction result are input into a pre-trained photovoltaic ultra-short-term prediction model, so that a photovoltaic ultra-short-term prediction result is obtained; illustratively, the photovoltaic ultrashort-term forecast result includes a photovoltaic output power forecast value.
According to the technical scheme, when the photovoltaic ultra-short-term radiation prediction is carried out, on one hand, aiming at the influence of cloud image shielding prediction on the photovoltaic ultra-short-term prediction, the foundation cloud image prediction result is input into a pre-trained cloud body shielding prediction model for prediction so as to consider shielding prediction with higher foundation cloud image precision; on the other hand, the influence of cloud image movement prediction on photovoltaic ultra-short-term prediction is carried out, and meanwhile, the foundation cloud image prediction result is input into a cloud body shielding prediction model trained in advance for prediction, so that movement prediction with high satellite cloud image precision is considered, and the technical problem that a scheme of carrying out cloud image recognition by singly adopting foundation cloud images or satellite cloud images in the related technology cannot simultaneously consider high prediction precision on cloud image movement tracks and high precision prediction on cloud image shielding is solved.
Although the satellite cloud image has lower precision in the aspect of cloud occlusion prediction and the ground cloud image has lower precision in the aspect of cloud movement prediction, the satellite cloud image data still has influence on the cloud occlusion prediction, and the ground cloud image data still has influence on the cloud movement prediction; in order to consider the influence of the satellite cloud image on cloud occlusion when making occlusion prediction and the influence of the ground cloud image on cloud movement when making movement prediction, as an exemplary embodiment, the photovoltaic ultra-short term forecasting method further includes: inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model for forecasting to obtain a second shielding forecasting result and shielding forecasting accuracy based on the satellite cloud picture; inputting the prediction result of the foundation cloud image into a pre-trained cloud occlusion prediction model for prediction to obtain a second movement prediction result and movement prediction accuracy based on the foundation cloud image; fusing the first movement prediction result based on the second movement prediction result and the movement prediction accuracy; and fusing the first shielding prediction result based on the second shielding prediction result and the shielding prediction accuracy.
In this embodiment, the first movement prediction result may be weighted and fused based on the second movement prediction result and the movement prediction accuracy, the second shielding prediction result may be weighted and fused based on the second shielding prediction result and the shielding prediction accuracy, and other optional methods such as root mean square error and average absolute error analysis may be used for fusion.
As an exemplary embodiment, the method further comprises: the accuracy of the movement prediction is positively correlated with the fusion weight of the second movement prediction result; and the shielding prediction accuracy is positively correlated with the fusion weight of the second shielding prediction result.
As an exemplary embodiment, the photovoltaic ultra-short term prediction model includes a cloud chart irradiance correction module and a photovoltaic ultra-short term prediction module, and the inputting the first movement prediction result and the first shielding prediction result into the photovoltaic ultra-short term prediction model trained in advance, to obtain a photovoltaic ultra-short term prediction result includes: inputting the first movement prediction result and the first shielding prediction result to the cloud picture irradiance correction module to respectively obtain irradiance corrected based on the characteristics of the ground cloud picture and irradiance corrected based on the characteristics of the satellite cloud picture; the irradiance after the foundation cloud picture characteristic correction and the irradiance after the satellite cloud picture characteristic correction are input to the photovoltaic ultra-short-term forecasting module to obtain a photovoltaic ultra-short-term forecasting result; the photovoltaic ultra-short-term forecasting module comprises a plurality of photovoltaic ultra-short-term forecasting sub-modules.
In this embodiment, the photovoltaic ultra-short term prediction module includes a plurality of photovoltaic ultra-short term prediction sub-modules, and specifically, the photovoltaic ultra-short term prediction module includes an algorithm selection sub-module, a machine learning model sub-module, and a deep learning model sub-module; the machine learning model submodule comprises a plurality of machine learning submodules which are constructed and trained in advance based on historical ground cloud image data and historical satellite cloud image data, and particularly, when the machine learning model submodule is constructed and trained, artificial feature extraction and selection are carried out on irradiance after the correction of the characteristics of the historical ground cloud image, irradiance after the correction of the characteristics of the historical satellite cloud image and photovoltaic output result data corresponding to the irradiance, and the data are further divided into a training set, a verification set and a test set based on feature extraction and selection results; and constructing and training a plurality of machine learning sub-models based on AdaBoost, bagging, RT, GBRT, BRR machine learning algorithm by adopting a training set and a verification set, and calculating the weight coefficient of each machine learning sub-model by adopting a test set after the machine learning sub-models are converged.
The deep learning model submodule comprises a plurality of deep learning submodules which are constructed and trained in advance based on historical ground cloud image data and historical satellite cloud image data, and particularly, when the machine learning model submodule is constructed and trained, basic feature extraction is carried out on irradiance after the correction of the characteristics of the historical ground cloud image, irradiance after the correction of the characteristics of the historical satellite cloud image and photovoltaic output result data corresponding to the irradiance, and the data are further divided into a training set, a verification set and a test set based on basic feature extraction results; and constructing and training a plurality of deep learning submodels based on a RNN, LSTM, CNN deep learning algorithm by adopting a training set and a verification set, and calculating the fusion weight of each deep learning submodel by adopting a test set after the deep learning submodel converges.
The method comprises the steps of obtaining historical ground cloud image data, historical satellite cloud image data and historical actual measurement irradiance data after training of a photovoltaic ultra-short-term forecasting module is completed, obtaining irradiance after correcting the characteristics of the historical ground cloud image based on the historical ground cloud image data and the photovoltaic ultra-short-term forecasting model, and obtaining irradiance after correcting the characteristics of the historical satellite cloud image based on the historical satellite cloud image data and the photovoltaic ultra-short-term forecasting model; and carrying out error analysis on the irradiance after the correction of the characteristics of the historical ground cloud picture and the historical measured irradiance data to obtain a ground cloud picture irradiation error E1, carrying out error analysis on the irradiance after the correction of the characteristics of the historical satellite cloud picture and the historical measured irradiance data to obtain a satellite cloud picture irradiation error E2, and determining fusion weights of the irradiance after the correction of the characteristics of the ground cloud picture and the irradiance after the correction of the characteristics of the satellite cloud picture based on the E1 and the E2 when the irradiance is input to the photovoltaic ultra-short-term forecasting module to obtain a radiation forecasting result.
For example, the base cloud image irradiation error E1 and the satellite cloud image irradiation error E2 may be measured by using indexes such as root mean square error or average absolute error.
The irradiance after correction of the characteristics of the ground cloud image and the irradiance after correction of the characteristics of the satellite cloud image are input to the photovoltaic ultra-short-term prediction module, and when the obtained radiation prediction result is obtained, the inverse of the error is used as a fusion weight when the radiation prediction result is obtained, namely the ground cloud image radiation prediction result weight W1=E2/(E1+E2), and the satellite cloud image radiation prediction result weight W2=E1/(E1+E2). And finally, carrying out weighted average on the predicted data of the two predicted irradiance to obtain a final predicted result.
In order to determine the influence of cloud clusters on the radiance, as an exemplary embodiment, the cloud map irradiance correction module includes a cloud cluster shadow analysis module and a cloud dynamic light intensity module, and the inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short term prediction model, to obtain a photovoltaic ultra-short term prediction result includes: calculating altitude change information and azimuth change information of a radiation forecast place of the sun within a preset duration based on the cloud cluster shadow analysis module; inputting the first movement prediction result and the first shielding prediction result to the cloud cluster shadow analysis module, and calculating movement track information and movement duration information of cloud cluster shadows based on the altitude change information and the azimuth change information; inputting the movement track information and the movement duration information into a cloud dynamic light intensity module, and calculating shielding influence information of cloud shadows on ground light intensity based on the Yun Dongtai light intensity module; and obtaining a radiation forecasting result by adopting a pre-trained cloud radiation forecasting model based on the shielding influence information.
In the embodiment, the cloud cluster shadow analysis module is used for calculating altitude change information and azimuth change information of a radiation forecast place of the sun within a preset duration; and inputting the first movement prediction result and the first shielding prediction result to the cloud shadow analysis module, calculating the position and the form of the cloud shadow on the ground through the cloud shadow analysis module, and then calculating movement track information and movement duration information of the cloud shadow based on the altitude change information and the azimuth change information.
As a possible implementation manner, a cloud cluster shadow analysis model may be pre-constructed and trained based on historical ground cloud image data, historical satellite cloud image data, historical solar altitude data, historical solar azimuth data, cloud cluster shadow historical movement track data and cloud cluster shadow historical movement duration data, and the corresponding relationship between the historical ground cloud image data, the historical satellite cloud image data, the historical solar altitude data, the historical solar azimuth data, the cloud cluster shadow historical movement track data and the cloud cluster shadow historical movement duration data is learned in the training process.
In this embodiment, the movement track information and the movement duration information are input into a cloud dynamic light intensity module, and shielding influence information of cloud shadows on ground light intensity is calculated based on the Yun Dongtai light intensity module; specifically, based on a physical optical principle, a light transmission model is established to simulate the influence of cloud shadows on the ground light intensity. As a possible implementation manner, the simulation can be performed by adopting methods such as ray tracing, radiation transmission and the like; specifically, each point on the ground is taken as a starting point of light, and the influence of cloud shadows on the ground light intensity is calculated by simulating the transmission and scattering processes of the light in the atmosphere and the cloud layer.
As a possible implementation manner, a cloud cluster shielding model can be constructed and trained based on cloud shadow historical movement track data, cloud shadow historical movement duration data and historical ground light intensity data, and the corresponding relation between the cloud shadow historical movement track data, the cloud shadow historical movement duration data and the historical ground light intensity data is learned in the training process.
As an exemplary embodiment, the ground cloud image forecasting model includes a ground cloud image feature extraction module and a ground cloud image motion trend determination module, and the inputting the ground cloud image data into the pre-trained ground cloud image forecasting model to obtain a ground cloud image forecasting result includes: inputting the foundation cloud image data into a pre-trained foundation cloud image forecast model, and extracting foundation cloud image motion characteristics and foundation cloud image morphological characteristics based on the foundation cloud image characteristic extraction module; the foundation cloud image movement characteristics comprise the position, the speed and the direction of a cloud body, and the foundation cloud image morphological characteristics comprise the shape and the size of the cloud body; and predicting based on the foundation cloud picture motion trend determining module, the foundation cloud picture motion characteristics and the foundation cloud picture morphological characteristics to obtain the foundation cloud picture motion trend based on the foundation cloud picture within a preset time length in the future.
As an exemplary embodiment, the satellite cloud image prediction model includes a satellite cloud image feature extraction module, a satellite cloud image classification module and a satellite cloud image movement prediction module, and the inputting the ground-based cloud image data into a pre-trained ground-based cloud image prediction model for prediction, to obtain a first shielding prediction result based on the ground-based cloud image includes: inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model, and extracting satellite cloud image features based on the satellite cloud image feature extraction module, wherein the satellite cloud image features comprise the size, the shape and the texture of a cloud body; performing cloud classification based on the satellite cloud image characteristics and the satellite cloud image classification module to obtain classification results; and carrying out satellite cloud picture prediction based on the classification result and the satellite cloud picture movement prediction module to obtain satellite cloud picture movement trend based on the satellite cloud pictures in the future preset time length.
From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM (Read-Only Memory)/RAM (Random Access Memory), magnetic disk, optical disk), comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present application.
According to a third aspect of the embodiment of the present application, there is further provided a photovoltaic ultra-short-term prediction apparatus for implementing the above photovoltaic ultra-short-term prediction method, and fig. 2 is a schematic diagram of a photovoltaic ultra-short-term prediction apparatus according to an embodiment of the present application, as shown in fig. 2, where the photovoltaic ultra-short-term prediction apparatus includes:
a data acquisition module 201, configured to acquire ground cloud image data and satellite cloud image data;
the foundation cloud image forecasting module 202 is configured to input the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result;
the satellite cloud image forecasting module 203 inputs the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result;
the movement prediction result calculation module 204 is configured to input the satellite cloud image prediction result to a pre-trained cloud body movement prediction model for prediction, so as to obtain a first movement prediction result based on the satellite cloud image;
the occlusion prediction result calculation module 205 is configured to input the prediction result of the ground-based cloud image to a cloud occlusion prediction model trained in advance to perform prediction, so as to obtain a first occlusion prediction result based on the ground-based cloud image;
The photovoltaic ultra-short term prediction module 206 is configured to input the first movement prediction result and the first shielding prediction result into a photovoltaic ultra-short term prediction model trained in advance, so as to obtain a photovoltaic ultra-short term prediction result.
It should be noted that, the data acquisition module 201 in this embodiment may be used to perform the above step S100, the ground cloud image prediction module 202 in this embodiment may be used to perform the above step S200, the satellite cloud image prediction module 203 in this embodiment may be used to perform the above step S300, the movement prediction result calculation module 204 in this embodiment may be used to perform the above step S400, the occlusion prediction result calculation module 205 in this embodiment may be used to perform the above step S500, and the occlusion prediction result calculation module 206 in this embodiment may be used to perform the above step S600.
It should be noted that the above modules are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to what is disclosed in the above embodiments. It should be noted that, the above modules may be implemented in software as a part of the apparatus, or may be implemented in hardware, where the hardware environment includes a network environment.
According to a fourth aspect of the present application, there is provided an electronic device comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other via the communication bus, the memory being for storing a computer program; the processor is configured to execute the method in any of the embodiments described above by running the computer program stored on the memory.
Fig. 3 is a block diagram of an alternative electronic device, according to an embodiment of the application, as shown in fig. 3, including a processor 302, a communication interface 304, a memory 306, and a communication bus 308, wherein the processor 302, the communication interface 304, and the memory 306 communicate with each other via the communication bus 308, wherein,
a memory 306 for storing a computer program;
the processor 302 is configured to execute the computer program stored in the memory 306, and implement the following steps:
acquiring foundation cloud image data and satellite cloud image data;
inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result;
inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result;
Inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture;
inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image;
and inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result.
Alternatively, in the present embodiment, the above-described communication bus may be a PCI (Peripheral Component Interconnect, peripheral component interconnect standard) bus, or an EISA (Extended Industry Standard Architecture ) bus, or the like. The communication bus may be classified as an address bus, a data bus, a control bus, or the like. For ease of illustration, only one thick line is shown in fig. 3, but not only one bus or one type of bus.
The communication interface is used for communication between the electronic device and other devices.
The memory may include RAM or may include non-volatile memory (non-volatile memory), such as at least one disk memory. Optionally, the memory may also be at least one memory device located remotely from the aforementioned processor.
As an example, as shown in fig. 3, the above-mentioned memory 302 may include, but is not limited to, the above-mentioned data acquisition module 201, the ground cloud image prediction module 202, the satellite cloud image prediction module 203, the movement prediction result calculation module 204, the occlusion prediction result calculation module 205, and the photovoltaic ultra-short term prediction module 206, and may also include other module units in the above-mentioned embodiment, which are not described in detail in this example.
The processor may be a general purpose processor and may include, but is not limited to: CPU (Central Processing Unit ), NP (Network Processor, network processor), etc.; but also DSP (Digital Signal Processing, digital signal processor), ASIC (Application Specific Integrated Circuit ), FPGA (Field-Programmable Gate Array, field programmable gate array) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components.
Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments, and this embodiment is not described herein.
It will be appreciated by those skilled in the art that the structure shown in fig. 3 is only illustrative, and the device implementing the method according to any of the foregoing embodiments may be a terminal device, and the terminal device may be a smart phone (such as an Android mobile phone, an IOS mobile phone, etc.), a tablet computer, a palm computer, a mobile internet device (Mobile Internet Devices, MID), a PAD, etc. Fig. 3 is not limited to the structure of the electronic device. For example, the terminal device may also include more or fewer components (e.g., network interfaces, display devices, etc.) than shown in fig. 3, or have a different configuration than shown in fig. 3.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be implemented by a program for instructing a terminal device to execute in association with hardware, the program may be stored in a computer readable storage medium, and the storage medium may include: flash disk, ROM, RAM, magnetic or optical disk, etc.
As an exemplary embodiment, the application also provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is arranged to perform the method steps of any of the embodiments when run.
Alternatively, in the present embodiment, the above-described storage medium may be used for executing the program code of the method steps of the embodiment of the present application.
Alternatively, in this embodiment, the storage medium may be located on at least one network device of the plurality of network devices in the network shown in the above embodiment.
Alternatively, in the present embodiment, the storage medium is configured to store program code for performing the steps of:
acquiring foundation cloud image data and satellite cloud image data;
inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result;
Inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result;
inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture;
inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image;
and inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result.
Alternatively, specific examples in the present embodiment may refer to examples described in the above embodiments, which are not described in detail in the present embodiment.
Alternatively, in the present embodiment, the storage medium may include, but is not limited to: various media capable of storing program codes, such as a U disk, ROM, RAM, a mobile hard disk, a magnetic disk or an optical disk.
The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
The integrated units in the above embodiments may be stored in the above-described computer-readable storage medium if implemented in the form of software functional units and sold or used as separate products. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing one or more computer devices (which may be personal computers, servers or network devices, etc.) to perform all or part of the steps of the method described in the embodiments of the present application.
In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In several embodiments provided by the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, such as the division of the units, is merely a logical function division, and may be implemented in another manner, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution provided in the present embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application, which are intended to be comprehended within the scope of the present application.
In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application, which are intended to be comprehended within the scope of the present application.

Claims (8)

1. The photovoltaic ultra-short-term forecasting method is characterized by comprising the following steps of:
Acquiring foundation cloud image data and satellite cloud image data;
inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model to obtain a foundation cloud image forecasting result;
inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result;
inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model to forecast, so as to obtain a first movement forecasting result based on the satellite cloud picture;
inputting the foundation cloud image forecasting result into a pre-trained cloud body shielding forecasting model for forecasting to obtain a first shielding forecasting result based on the foundation cloud image;
inputting the first movement prediction result and the first shielding prediction result into a pre-trained photovoltaic ultra-short-term prediction model to obtain a photovoltaic ultra-short-term prediction result; wherein,
the photovoltaic ultra-short term forecasting model comprises a cloud picture irradiance correction module and a photovoltaic ultra-short term forecasting module, the first movement forecasting result and the first shielding forecasting result are input into the photovoltaic ultra-short term forecasting model trained in advance, and the obtaining of the photovoltaic ultra-short term forecasting result comprises the following steps:
Inputting the first movement prediction result and the first shielding prediction result to the cloud picture irradiance correction module to respectively obtain irradiance after correction of the foundation cloud picture characteristics and irradiance after correction of the satellite cloud picture characteristics;
the irradiance after the foundation cloud picture characteristic correction and the irradiance after the satellite cloud picture characteristic correction are input to the photovoltaic ultra-short-term forecasting module to obtain a photovoltaic ultra-short-term forecasting result; the photovoltaic ultra-short-term forecasting module comprises a plurality of photovoltaic ultra-short-term forecasting sub-modules;
the cloud picture irradiance correction module comprises a cloud shadow analysis module and a cloud dynamic light intensity module, the first movement prediction result and the first shielding prediction result are input into a photovoltaic ultra-short-term prediction model trained in advance, and the obtaining of the photovoltaic ultra-short-term prediction result comprises the following steps:
calculating altitude change information and azimuth change information of a radiation forecast place of the sun within a preset time period based on the cloud shadow analysis module;
inputting the first movement prediction result and the first shielding prediction result to the cloud shadow analysis module, and calculating movement track information and movement duration information of cloud shadows based on the altitude change information and the azimuth change information;
Inputting the movement track information and the movement duration information into a cloud dynamic light intensity module, and calculating shielding influence information of cloud shadows on ground light intensity based on the Yun Dongtai light intensity module;
and obtaining a radiation forecasting result by adopting a pre-trained cloud radiation forecasting model based on the shielding influence information.
2. The photovoltaic ultrashort-term forecasting method of claim 1, further comprising:
inputting the satellite cloud picture forecasting result into a pre-trained cloud body movement forecasting model for forecasting to obtain a second shielding forecasting result and shielding forecasting accuracy based on the satellite cloud picture;
inputting the prediction result of the foundation cloud image into a pre-trained cloud occlusion prediction model for prediction to obtain a second movement prediction result and movement prediction accuracy based on the foundation cloud image;
fusing the first movement prediction result based on the second movement prediction result and the movement prediction accuracy;
and fusing the first shielding prediction result based on the second shielding prediction result and the shielding prediction accuracy.
3. The photovoltaic ultrashort-term forecasting method of claim 2, further comprising:
The accuracy of the movement prediction is positively correlated with the fusion weight of the second movement prediction result;
and the shielding prediction accuracy is positively correlated with the fusion weight of the second shielding prediction result.
4. The photovoltaic ultra-short term forecasting method according to claim 1, wherein the ground cloud image forecasting model comprises a ground cloud image feature extraction module and a ground cloud image movement trend determination module, and the step of inputting the ground cloud image data into the pre-trained ground cloud image forecasting model to obtain a ground cloud image forecasting result comprises the following steps:
inputting the foundation cloud image data into a pre-trained foundation cloud image forecast model, and extracting foundation cloud image motion characteristics and foundation cloud image morphological characteristics based on the foundation cloud image characteristic extraction module; the foundation cloud image movement characteristics comprise the position, the speed and the direction of a cloud body, and the foundation cloud image morphological characteristics comprise the shape and the size of the cloud body;
and predicting based on the foundation cloud picture motion trend determining module, the foundation cloud picture motion characteristics and the foundation cloud picture morphological characteristics to obtain the foundation cloud picture motion trend based on the foundation cloud picture within a preset time length in the future.
5. The photovoltaic ultra-short term forecasting method according to claim 1, wherein the satellite cloud image forecasting model comprises a satellite cloud image feature extraction module, a satellite cloud image classification module and a satellite cloud image movement forecasting module, the step of inputting the foundation cloud image data into a pre-trained foundation cloud image forecasting model for forecasting, and the step of obtaining a first shielding forecasting result based on the foundation cloud image comprises the following steps:
inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model, and extracting satellite cloud image features based on the satellite cloud image feature extraction module, wherein the satellite cloud image features comprise the size, the shape and the texture of a cloud body;
performing cloud classification based on the satellite cloud image characteristics and the satellite cloud image classification module to obtain classification results;
and carrying out satellite cloud picture prediction based on the classification result and the satellite cloud picture movement prediction module to obtain satellite cloud picture movement trend based on the satellite cloud pictures in the future preset time length.
6. A photovoltaic ultrashort-term forecasting device, characterized in that it comprises:
the data acquisition module is used for acquiring the ground cloud image data and the satellite cloud image data;
The foundation cloud picture forecasting module is used for inputting the foundation cloud picture data into a pre-trained foundation cloud picture forecasting model to obtain a foundation cloud picture forecasting result;
the satellite cloud image forecasting module is used for inputting the satellite cloud image data into a pre-trained satellite cloud image forecasting model to obtain a satellite cloud image forecasting result;
the mobile prediction result calculation module is used for inputting the satellite cloud image prediction result into a pre-trained cloud body mobile prediction model to predict, so as to obtain a first mobile prediction result based on the satellite cloud image;
the shielding prediction result calculation module is used for inputting the foundation cloud image prediction result into a pre-trained cloud body shielding prediction model to predict, so as to obtain a first shielding prediction result based on the foundation cloud image;
the photovoltaic ultra-short-term forecasting module is used for inputting the first movement forecasting result and the first shielding forecasting result into a photovoltaic ultra-short-term forecasting model trained in advance to obtain a photovoltaic ultra-short-term forecasting result; the photovoltaic ultra-short-term forecasting model comprises a cloud picture irradiance correction module and a photovoltaic ultra-short-term forecasting module, wherein the photovoltaic ultra-short-term forecasting module is further used for inputting the first movement forecasting result and the first shielding forecasting result into the cloud picture irradiance correction module to respectively obtain irradiance after correction of the characteristics of the foundation cloud picture and irradiance after correction of the characteristics of the satellite cloud picture; the irradiance after the foundation cloud picture characteristic correction and the irradiance after the satellite cloud picture characteristic correction are input to the photovoltaic ultra-short-term forecasting module to obtain a photovoltaic ultra-short-term forecasting result; the photovoltaic ultra-short-term forecasting module comprises a plurality of photovoltaic ultra-short-term forecasting sub-modules;
The cloud picture irradiance correction module comprises a cloud shadow analysis module and a cloud dynamic light intensity module, the first movement prediction result and the first shielding prediction result are input into a photovoltaic ultra-short-term prediction model trained in advance, and the obtaining of the photovoltaic ultra-short-term prediction result comprises the following steps: calculating altitude change information and azimuth change information of a radiation forecast place of the sun within a preset time period based on the cloud shadow analysis module; inputting the first movement prediction result and the first shielding prediction result to the cloud shadow analysis module, and calculating movement track information and movement duration information of cloud shadows based on the altitude change information and the azimuth change information; inputting the movement track information and the movement duration information into a cloud dynamic light intensity module, and calculating shielding influence information of cloud shadows on ground light intensity based on the Yun Dongtai light intensity module; and obtaining a radiation forecasting result by adopting a pre-trained cloud radiation forecasting model based on the shielding influence information.
7. An electronic device, the electronic device comprising:
at least one processor; the method comprises the steps of,
a memory communicatively coupled to the at least one processor; wherein,
The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the photovoltaic ultra-short term forecasting method of any one of claims 1 to 5.
8. A computer-readable storage medium, characterized in that the storage medium has stored therein a computer program, wherein the computer program is arranged to execute the photovoltaic ultrashort-term forecasting method according to any one of claims 1 to 5 when run.
CN202311027499.0A 2023-08-16 2023-08-16 Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium Active CN116780533B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311027499.0A CN116780533B (en) 2023-08-16 2023-08-16 Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311027499.0A CN116780533B (en) 2023-08-16 2023-08-16 Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN116780533A CN116780533A (en) 2023-09-19
CN116780533B true CN116780533B (en) 2023-11-03

Family

ID=88011837

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311027499.0A Active CN116780533B (en) 2023-08-16 2023-08-16 Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN116780533B (en)

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104156733A (en) * 2014-08-12 2014-11-19 中国人民解放军理工大学 Foundation cloud form identification method based on multiscale structure characteristics
CN107644416A (en) * 2017-09-15 2018-01-30 中国科学院寒区旱区环境与工程研究所 A kind of real-time dynamic cloud amount inversion method based on ground cloud atlas
CN109543721A (en) * 2018-11-05 2019-03-29 中国科学院寒区旱区环境与工程研究所 A kind of solar irradiance ultra-short term forecasting procedure under fine with occasional clouds weather condition
CN112507793A (en) * 2020-11-05 2021-03-16 上海电力大学 Ultra-short-term photovoltaic power prediction method
CN113128793A (en) * 2021-05-19 2021-07-16 中国南方电网有限责任公司 Photovoltaic power combination prediction method and system based on multi-source data fusion
CN113378459A (en) * 2021-06-02 2021-09-10 兰州交通大学 Photovoltaic power station ultra-short-term power prediction method based on satellite and internet of things information
CN114510513A (en) * 2021-12-31 2022-05-17 国网浙江省电力有限公司绍兴供电公司 Short-term meteorological forecast data processing method for ultra-short-term photovoltaic power prediction
KR20220078963A (en) * 2020-12-04 2022-06-13 주식회사 효성 Deep learning based photovoltaic power generation forecasting method using satellite images and apparatus using the same
CN114707688A (en) * 2021-12-31 2022-07-05 国网浙江省电力有限公司绍兴供电公司 Photovoltaic power ultra-short-term prediction method based on satellite cloud chart and space-time neural network
CN114898228A (en) * 2022-06-14 2022-08-12 中国科学院西北生态环境资源研究院 Solar total irradiance inversion method based on satellite cloud picture and random forest model
CN115546657A (en) * 2022-09-20 2022-12-30 湖南防灾科技有限公司 Regional earth surface radiation intensity prediction method based on cloud cluster motion trail
CN116245248A (en) * 2023-03-15 2023-06-09 国网河北省电力有限公司保定供电分公司 Cloud optical thickness-based distributed light Fu Muxian ultra-short-term power prediction method
CN116542383A (en) * 2023-05-12 2023-08-04 国网安徽省电力有限公司电力科学研究院 Distributed photovoltaic system output prediction method based on small fluctuation weather satellite cloud image

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102584854B1 (en) * 2021-09-03 2023-10-13 경북대학교 산학협력단 Prediction system for solor power generation

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104156733A (en) * 2014-08-12 2014-11-19 中国人民解放军理工大学 Foundation cloud form identification method based on multiscale structure characteristics
CN107644416A (en) * 2017-09-15 2018-01-30 中国科学院寒区旱区环境与工程研究所 A kind of real-time dynamic cloud amount inversion method based on ground cloud atlas
CN109543721A (en) * 2018-11-05 2019-03-29 中国科学院寒区旱区环境与工程研究所 A kind of solar irradiance ultra-short term forecasting procedure under fine with occasional clouds weather condition
CN112507793A (en) * 2020-11-05 2021-03-16 上海电力大学 Ultra-short-term photovoltaic power prediction method
KR20220078963A (en) * 2020-12-04 2022-06-13 주식회사 효성 Deep learning based photovoltaic power generation forecasting method using satellite images and apparatus using the same
CN113128793A (en) * 2021-05-19 2021-07-16 中国南方电网有限责任公司 Photovoltaic power combination prediction method and system based on multi-source data fusion
CN113378459A (en) * 2021-06-02 2021-09-10 兰州交通大学 Photovoltaic power station ultra-short-term power prediction method based on satellite and internet of things information
CN114510513A (en) * 2021-12-31 2022-05-17 国网浙江省电力有限公司绍兴供电公司 Short-term meteorological forecast data processing method for ultra-short-term photovoltaic power prediction
CN114707688A (en) * 2021-12-31 2022-07-05 国网浙江省电力有限公司绍兴供电公司 Photovoltaic power ultra-short-term prediction method based on satellite cloud chart and space-time neural network
CN114898228A (en) * 2022-06-14 2022-08-12 中国科学院西北生态环境资源研究院 Solar total irradiance inversion method based on satellite cloud picture and random forest model
CN115546657A (en) * 2022-09-20 2022-12-30 湖南防灾科技有限公司 Regional earth surface radiation intensity prediction method based on cloud cluster motion trail
CN116245248A (en) * 2023-03-15 2023-06-09 国网河北省电力有限公司保定供电分公司 Cloud optical thickness-based distributed light Fu Muxian ultra-short-term power prediction method
CN116542383A (en) * 2023-05-12 2023-08-04 国网安徽省电力有限公司电力科学研究院 Distributed photovoltaic system output prediction method based on small fluctuation weather satellite cloud image

Also Published As

Publication number Publication date
CN116780533A (en) 2023-09-19

Similar Documents

Publication Publication Date Title
Onishi et al. Super-resolution simulation for real-time prediction of urban micrometeorology
US20200090357A1 (en) Method and system for generating synthetic point cloud data using a generative model
CN111161349B (en) Object posture estimation method, device and equipment
WO2015104281A1 (en) Solar irradiance forecasting
CN109214366A (en) Localized target recognition methods, apparatus and system again
CN107767454A (en) A kind of three-dimensional mobile fast modeling method of outdoor scene, apparatus and system
CN108182695B (en) Target tracking model training method and device, electronic equipment and storage medium
CN110826581B (en) Animal number identification method, device, medium and electronic equipment
CN113537561A (en) Ultra-short-term solar radiation prediction method and system based on foundation cloud picture
CN111047088A (en) Prediction image acquisition method and device, computer equipment and storage medium
CN114185114B (en) Correction method and device for rainfall forecast falling area, electronic equipment and medium
CN115563732A (en) Spraying track simulation optimization method and device based on virtual reality
CN112036381A (en) Visual tracking method, video monitoring method and terminal equipment
CN117079132A (en) Remote sensing image target detection method based on Gaussian distance loss
CN116799796A (en) Photovoltaic power generation power prediction method, device, equipment and medium
CN115239508A (en) Scene planning adjustment method, device, equipment and medium based on artificial intelligence
CN117409083B (en) Cable terminal identification method and device based on infrared image and improved YOLOV5
CN116780533B (en) Photovoltaic ultra-short-term forecasting method and device, electronic equipment and storage medium
CN111428191B (en) Antenna downtilt angle calculation method and device based on knowledge distillation and storage medium
CN116051777B (en) Super high-rise building extraction method, apparatus and readable storage medium
CN116029440A (en) Ultra-short-term power prediction method and device for photovoltaic power station
CN115760908A (en) Insulator tracking method and device based on capsule network perception characteristics
CN115618714A (en) Solar radiation probability forecasting method and related equipment
CN115859765A (en) Method, device, equipment and storage medium for predicting city expansion
KR101326644B1 (en) Full-body joint image tracking method using evolutionary exemplar-based particle filter

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant