CN117691592A - Photovoltaic output prediction method and device, electronic equipment and storage medium - Google Patents
Photovoltaic output prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN117691592A CN117691592A CN202311724642.1A CN202311724642A CN117691592A CN 117691592 A CN117691592 A CN 117691592A CN 202311724642 A CN202311724642 A CN 202311724642A CN 117691592 A CN117691592 A CN 117691592A
- Authority
- CN
- China
- Prior art keywords
- historical
- cloud
- image
- photovoltaic output
- satellite cloud
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000003860 storage Methods 0.000 title claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 15
- 238000001914 filtration Methods 0.000 claims description 38
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 15
- 150000003839 salts Chemical class 0.000 claims description 9
- 238000009499 grossing Methods 0.000 claims description 8
- 238000003709 image segmentation Methods 0.000 claims description 8
- 230000000903 blocking effect Effects 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 9
- 238000009826 distribution Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 235000002566 Capsicum Nutrition 0.000 description 5
- 239000006002 Pepper Substances 0.000 description 5
- 241000722363 Piper Species 0.000 description 5
- 235000016761 Piper aduncum Nutrition 0.000 description 5
- 235000017804 Piper guineense Nutrition 0.000 description 5
- 235000008184 Piper nigrum Nutrition 0.000 description 5
- 230000008901 benefit Effects 0.000 description 5
- 238000013527 convolutional neural network Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000010248 power generation Methods 0.000 description 3
- 230000001932 seasonal effect Effects 0.000 description 3
- 230000002146 bilateral effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Photovoltaic Devices (AREA)
Abstract
The application discloses a photovoltaic output prediction method, a device, electronic equipment and a storage medium, wherein the photovoltaic output prediction method comprises the following steps: acquiring a plurality of first historical satellite cloud pictures in a first preset time period in a first target area; extracting first historical cloud cluster shielding features in a first historical satellite cloud picture; according to the first historical cloud cover characteristics, predicting the cloud cover characteristics in a second preset time period; acquiring first weather information in a second preset time period; and inputting the cloud cover characteristics and the first atmospheric information into a preset model to obtain a photovoltaic output predicted value, wherein the preset model is obtained based on ConvTransformer model training. Therefore, the prediction accuracy of the photovoltaic output is improved.
Description
Technical Field
The application relates to the technical field of photovoltaic output prediction, in particular to a photovoltaic output prediction method, a device, electronic equipment and a storage medium.
Background
The photovoltaic has the advantages of small scale, high practicability, easy combination with buildings and the like, has been popularized in China on a large scale in recent years, however, the large-scale photovoltaic is accessed into the power distribution network in a multipoint and unordered mode, so that the power distribution network is faced with a series of problems of electric energy quality, harmonic waves, economic operation and the like, and the safe and stable operation of the power grid is seriously affected. Therefore, the prediction of the photovoltaic output is scientifically carried out, and the method is an important guarantee means for realizing reliable, efficient and economic grid-connected operation of the photovoltaic.
The photovoltaic power generation has the characteristics of randomness and strong volatility under the influence of weather, for example, the fluctuation of the photovoltaic power output is particularly large in cloudy and overcast weather and other weather. The cloud cluster is one of main factors influencing the solar irradiance, and can shade solar irradiation, so that the photovoltaic output is influenced, and how to solve the problem that the photovoltaic output caused by the shade of the cloud cluster is difficult to predict is a hot spot problem of current research.
It can be seen that the existing photovoltaic output has the technical problem that accurate prediction is difficult.
Disclosure of Invention
The embodiment of the application aims to provide a photovoltaic output prediction method, a device, electronic equipment and a storage medium, which are used for solving the technical problem that the photovoltaic output is difficult to accurately predict in the prior art.
To achieve the above object, a first aspect of the present application provides a photovoltaic output prediction method, including:
acquiring a plurality of first historical satellite cloud pictures in a first preset time period in a first target area;
extracting first historical cloud cluster shielding features in a first historical satellite cloud picture;
according to the first historical cloud cover characteristics, predicting the cloud cover characteristics in a second preset time period;
acquiring first weather information in a second preset time period;
and inputting the cloud cover characteristics and the first atmospheric information into a preset model to obtain a photovoltaic output predicted value, wherein the preset model is obtained based on ConvTransformer model training.
In an embodiment of the present application, extracting a first historical cloud cluster occlusion feature in a first historical satellite cloud image includes:
image segmentation is carried out on the first historical satellite cloud image by using a gradient vector flow model, so that a plurality of historical cloud image in the first historical satellite cloud image are obtained;
and extracting features of the plurality of historical cloud cluster images by using the gray level co-occurrence matrix to obtain first historical cloud cluster shielding features in the first historical satellite cloud image.
In an embodiment of the present application, before extracting the first historical cloud cluster occlusion feature in the first historical satellite cloud chart, the method further includes:
the first historical satellite cloud image is preprocessed.
In an embodiment of the present application, preprocessing a first historical satellite cloud image includes:
carrying out gray histogram equalization on the first historical satellite cloud picture to obtain an equalized first historical satellite cloud picture;
and carrying out smooth filtering processing on the equalized first historical satellite cloud image.
In the embodiment of the present application, smoothing filtering processing is performed on the equalized first historical satellite cloud image, including:
removing Gaussian noise in the equalized first historical satellite cloud picture by means of mean filtering;
and removing the salt and pepper noise in the equalized first historical satellite cloud picture by using median filtering.
In an embodiment of the present application, the first historical cloud occlusion feature includes a historical cloud thickness and a historical cloud texture.
In an embodiment of the present application, the training step of the preset model includes:
acquiring a plurality of second historical satellite cloud pictures, second weather information and corresponding photovoltaic output historical values in a third preset time period in a second target area;
extracting a second historical cloud cluster shielding characteristic in a second historical satellite cloud picture;
training according to the second weather information, the second historical cloud cover characteristic and the photovoltaic output historical value to obtain a preset model based on the ConvTransformer model.
A second aspect of the present application provides a photovoltaic output predicting device, comprising:
a memory configured to store instructions; and
a processor configured to invoke instructions from a memory and when executing the instructions is capable of implementing the photovoltaic output prediction method according to any of the first aspects.
A third aspect of the present application provides an electronic device, including:
the photovoltaic output predicting device according to the second aspect.
A fourth aspect of the present application provides a machine-readable storage medium having stored thereon instructions for causing a machine to perform the photovoltaic output prediction method according to any one of the first aspect.
According to the technical scheme, the first historical cloud cluster shielding characteristic is obtained according to the first historical satellite cloud image, and the photovoltaic output is predicted by using the preset model based on the ConvTransformer model in combination with the first atmospheric information, so that the influence of high temperature, overcast and rainy, cloud cluster shielding or other bad weather on the photovoltaic output prediction can be reduced, and the prediction accuracy of the photovoltaic output is effectively improved.
Additional features and advantages of embodiments of the present application will be set forth in the detailed description that follows.
Drawings
The accompanying drawings are included to provide a further understanding of embodiments of the present application and are incorporated in and constitute a part of this specification, illustrate embodiments of the present application and together with the description serve to explain, without limitation, the embodiments of the present application. In the drawings:
FIG. 1 schematically illustrates a flow diagram of a photovoltaic output prediction method according to an embodiment of the present application;
fig. 2 schematically shows a schematic structural diagram of a photovoltaic output predicting device according to an embodiment of the present application.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the specific implementations described herein are only for illustrating and explaining the embodiments of the present application, and are not intended to limit the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present application based on the embodiments herein.
It should be noted that, in the embodiment of the present application, directional indications (such as up, down, left, right, front, and rear … …) are referred to, and the directional indications are merely used to explain the relative positional relationship, movement conditions, and the like between the components in a specific posture (as shown in the drawings), and if the specific posture is changed, the directional indications are correspondingly changed.
In addition, if there is a description of "first", "second", etc. in the embodiments of the present application, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions of the embodiments may be combined with each other, but it is necessary to base that the technical solutions can be realized by those skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be regarded as not exist and not within the protection scope of the present application.
Fig. 1 schematically shows a flow diagram of a photovoltaic output prediction method according to an embodiment of the present application. As shown in fig. 1, an embodiment of the present application provides a photovoltaic output prediction method, which may include the following steps.
Step 110, acquiring a plurality of first historical satellite cloud pictures in a first preset time period in a first target area;
it can be appreciated that, since the photovoltaic output prediction method provided by the embodiment of the present application needs to be predicted by combining with the cloud cover feature, a cloud image needs to be obtained first, and then the cloud cover feature is obtained from the cloud image. The common cloud image can comprise a ground cloud image and a satellite cloud image, wherein the ground cloud image is an image acquired in real time by an all-sky imager on the basis of the ground, and the satellite cloud image is an image obtained by observing the whole earth through a satellite. Compared with a satellite cloud image, the time and space resolution of the ground cloud image is higher, but the full-sky imager is high in price and is not suitable for a photovoltaic power station with low input cost. Therefore, the embodiment of the application selects the satellite cloud image, firstly, a plurality of first historical satellite cloud images in a first preset time period in a first target area are acquired by using the satellite, and an image source is provided for extracting first historical cloud occlusion features from the first historical satellite cloud images.
It should be noted that the first target area and the first preset time period may be set according to actual requirements, which is not limited in the embodiment of the present application.
Step 120, extracting first historical cloud cluster shielding features in a first historical satellite cloud picture;
it is understood that photovoltaic output refers to the product of current and voltage produced by a photovoltaic cell or photovoltaic module under certain conditions, and is used to represent the electrical energy production capacity of a photovoltaic site under given conditions. The photovoltaic power generation has the characteristics of high randomness and volatility under the influence of weather, for example, in cloudy and overcast weather, fluctuation of photovoltaic output is particularly large, and cloud clusters are one of main factors influencing the magnitude of solar irradiance and can shield solar irradiation, so that the photovoltaic output is influenced. Therefore, in the embodiment of the present application, the first historical cloud cluster shielding feature is extracted from the obtained first historical satellite cloud image, and there are various ways of extracting the first historical cloud cluster shielding feature, for example, a deep learning method, an image segmentation method, or other texture feature extraction methods, and the accuracy of photovoltaic output prediction can be improved by combining the first historical cloud cluster shielding feature to predict the photovoltaic output.
In the embodiment of the present application, step 120 includes:
step 121, performing image segmentation on a first historical satellite cloud image by using a gradient vector flow model to obtain a plurality of historical cloud cluster images in the first historical satellite cloud image;
and 122, extracting features of the plurality of historical cloud cluster images by using the gray level co-occurrence matrix to obtain first historical cloud cluster shielding features in the first historical satellite cloud image.
It is understood that the gradient vector flow (Gradient Vector Flow, GVF) model is a computer vision model for edge detection and image segmentation. It is an improvement based on an active contour model, aiming at overcoming some limitations of sensitivity and convergence speed in the traditional active contour model. The active contour model converges the contour to the edge of the object of interest by defining an initial contour in the image and minimizing the energy. However, conventional active contour models may converge to false edges or stagnate at local minima when edges are unclear or noise is present, while gradient vector flow models introduce gradient vector flow as force fields, helping the active contour model to better adapt to complex edges in the image. The method generates a smooth force field which attracts the contour to the edge of the target by calculating the gradient field of the image, thereby improving the perception and attraction capability of the active contour model to the edge.
Specifically, firstly, determining regions to be segmented in a first historical satellite cloud picture, extracting edges of each region to be segmented by using a gradient vector flow model, wherein the extracted results comprise real edges and pseudo edges, the real edges are edges which can be used for representing real cloud clusters, the pseudo edges are edges which cannot be used for representing real cloud clusters, therefore, the pseudo edges need to be removed, the real edges are connected end to end, the edges are extracted again by using the gradient vector flow model, the pseudo edges are removed again, the real edges are connected end to end, and the steps are repeated until the regions to be segmented do not comprise the pseudo edges any more and the real edges are all connected into a plurality of complete historical cloud cluster images.
The gradient vector flow model is used for carrying out image segmentation on the first historical satellite cloud image to obtain a plurality of historical cloud image in the first historical satellite cloud image, which is beneficial to reducing background and other interference information in the first historical satellite cloud image, and can accurately capture the edge information of the first historical satellite cloud image, and the clear edge features can be used for more accurately defining the area of the historical cloud image, so that the structure and texture features of the cloud are more concentrated instead of the surrounding environment when the first historical cloud shielding features are extracted later, and the accuracy of extracting the first historical cloud shielding features is improved.
It will be appreciated that Gray-Level Co-occurrence Matrix, GLCM, which is a statistical method for describing image texture features, is commonly used in the fields of image classification, texture recognition, feature extraction, etc., and is capable of capturing spatial relationships and distribution features between pixel Gray levels in an image. The gray level co-occurrence matrix describes texture features based on the relative position relation among pixel gray levels in an image, and a matrix is constructed by counting the occurrence frequency and the spatial distribution condition of specific pixel pairs in the image, wherein elements in the matrix represent the probability distribution condition of the specific gray level pairs in the image under given directions and distances, can be used for calculating a series of statistical features, and can be used for quantitatively describing the texture features of the image by analyzing the statistical features of the gray level co-occurrence matrix, so as to be used for image classification, identification or feature extraction.
Specifically, taking any point (x, y) in the historical cloud image and another point (x+a, y+b) deviating from the point distance a, setting the gray value corresponding to the point as (g 1, g 2), enabling the point (x, y) to move on the whole screen of the historical cloud image, obtaining various (g 1, g 2) values, setting the number of gray value levels as L, and enabling the combination of the (g 1, g 2) to share L 2 A kind of module is assembled in the module and the module is assembled in the module. And counting the occurrence times of each value (g 1, g 2) for the whole picture, and then arranging the values into a square matrix, namely a gray level co-occurrence matrix, wherein the gray level co-occurrence matrix can be used for calculating the pixel mean value, entropy and correlation in the historical cloud image, and the first historical cloud shielding characteristic in the first historical satellite cloud image can be obtained by analyzing the pixel mean value, entropy and correlation in the historical cloud image.
By utilizing the gray level co-occurrence matrix, the first historical cloud cover characteristics in the first historical satellite cloud picture can be accurately extracted, and the first historical cloud cover characteristics are quantized, so that the cloud cover characteristics in a second preset time period can be predicted.
In the embodiment of the present application, before step 120, the method further includes:
the first historical satellite cloud image is preprocessed.
It will be appreciated that due to the large spatial resolution of the first historical satellite cloud image that was initially acquired, insufficient detail information on the image may result in insufficiently accurate and detailed extracted features, and that low resolution images may have more noise or pixel level errors that may affect the accuracy of the image segmentation, resulting in interference or distortion of the extracted features and thus affecting the accuracy of the subsequent photovoltaic output predictions. Therefore, before the first historical cloud cluster shielding feature in the first historical satellite cloud image is extracted, the first historical satellite cloud image is preprocessed, for example, the operations such as image enhancement, strong enemy resolution, filtering and smoothing are used, so that the image feature of the first historical satellite cloud image can be enhanced, noise and ambiguity can be reduced, and a more reliable data base can be provided for subsequent photovoltaic output prediction.
In an embodiment of the present application, preprocessing a first historical satellite cloud image includes:
carrying out gray histogram equalization on the first historical satellite cloud picture to obtain an equalized first historical satellite cloud picture;
and carrying out smooth filtering processing on the equalized first historical satellite cloud image.
It will be appreciated that gray histogram equalization is an image processing technique for enhancing image contrast. In the image recognition process, important characteristics of the image can be highlighted by increasing gray contrast, and gray histogram equalization is to change pixel point distribution on each gray level to ensure that the pixel points have the same pixel point number, so that the image is distributed uniformly in the whole gray value dynamic change range, the brightness distribution state of the image is improved, and the visual effect of the image is enhanced.
In the embodiment of the application, the cloud cluster or the cloud layer in the cloud map is not obvious enough due to the reasons of illumination conditions and the like, and the contrast of the image can be enhanced by gray histogram equalization, so that the boundaries of the cloud cluster and other backgrounds or objects are clearer; and the texture information of the cloud cluster in the satellite cloud picture is important for the extraction of the subsequent first historical cloud cluster shielding characteristics, but sometimes the texture information is not obvious due to the fact that images with low contrast or narrow gray scale range can be caused, and the gray scale histogram equalization can enable the texture information to be more outstanding, so that the subsequent first historical cloud cluster shielding characteristics can be conveniently extracted.
It can be appreciated that since the first historical satellite cloud image is inevitably interfered by noise during the transmission and reception processes, and the noise advantage affects the quality of the first historical satellite cloud image, the image needs to be denoised after important features of the salient image are equalized by using the gray histogram. In the embodiment of the application, smoothing filtering processing is performed on the equalized first historical satellite cloud image. Smoothing filtering is an image processing technique aimed at reducing noise in an image and blurring details in an image, and is often used to remove noise, reduce variations in an image, smooth brightness variations of an image, and the like. The core idea is to process pixels in an image to reduce sharp variations between pixel values, smooth filtering typically operates based on a neighborhood or window of pixels in the image, statistics or calculations are performed on the pixel values within the window, and then these statistics are used to replace the value of the center pixel to reduce noise effects, blur details, reduce variations in the image, etc. Common smoothing filtering methods include mean filtering, gaussian filtering, median filtering, bilateral filtering and the like, different filtering methods are suitable for different types of noise, and proper filtering methods are selected and need to be determined according to specific image characteristics and processing requirements.
In the embodiment of the present application, smoothing filtering processing is performed on the equalized first historical satellite cloud image, including:
removing Gaussian noise in the equalized first historical satellite cloud picture by means of mean filtering;
and removing the salt and pepper noise in the equalized first historical satellite cloud picture by using median filtering.
It is understood that mean filtering and median filtering are two common smoothing filtering techniques. The average filtering is to replace the value of the central pixel by taking the average value of the pixel values in the neighborhood around the pixel, so that each pixel is replaced by the average value of the surrounding pixels, the average filtering is simple and easy to realize, and continuous noise such as Gaussian noise in an image can be effectively reduced; the median filtering replaces a central pixel by selecting a median of pixel values in a pixel neighborhood, is suitable for removing discrete noise such as salt and pepper noise, can keep edges and details of an image, and has a good effect on keeping image characteristics.
In the first historical satellite cloud picture, common noise is Gaussian noise or spiced salt noise, and average filtering and median filtering have good adaptability to the Gaussian noise and the spiced salt noise, and are respectively suitable for removing the Gaussian noise and the spiced salt noise. Compared with mean filtering and median filtering, the Gaussian filtering and bilateral filtering are more complex and time-consuming in calculation, so that when a large-scale satellite cloud image is actually processed, gaussian noise in the equalized first historical satellite cloud image is removed by means of the mean filtering, and salt and pepper noise in the equalized first historical satellite cloud image is removed by means of the median filtering, so that the quality of an image is improved, the influence of noise is reduced, the integral characteristics of the image are improved, and the extraction of the shielding characteristics of a subsequent first historical cloud cluster is facilitated.
In an embodiment of the present application, the first historical cloud occlusion feature includes a historical cloud thickness and a historical cloud texture.
Specifically, the first historical cloud cover feature includes a historical cloud thickness and a historical cloud texture, and when the first historical cloud cover feature is extracted by using the gray level co-occurrence matrix, the gray level co-occurrence matrix can calculate the pixel mean value, entropy and correlation in the historical cloud image. The historical cloud thickness may be described in terms of a pixel mean of the historical cloud image, which represents the overall brightness or gray level of the historical cloud image, with thicker historical cloud images typically having a higher pixel mean and thinner historical cloud images likely having a lower pixel mean. The historical cloud texture may then be described in terms of entropy and correlation in the historical cloud image. Entropy refers to the degree of information confusion of an image region, and for a historical cloud image region with complex textures, the entropy value will be higher. The correlation reflects the statistical relationship between pixels in the historical cloud image, and is higher for image areas with more regular textures.
The historical cloud thickness provides information about the cloud density and degree of occlusion, while the historical cloud texture provides visual features about the internal structure and distribution of the cloud. By combining the historical cloud thickness with the historical cloud texture, the first historical cloud occlusion feature is better described.
Step 130, predicting cloud occlusion characteristics in a second preset time period according to the first historical cloud occlusion characteristics;
it can be appreciated that, since the first historical cloud occlusion feature obtained in the above embodiment is a historical value, the historical value cannot be directly used to predict the photovoltaic output at the future time, and therefore, the cloud occlusion feature within the second preset time period needs to be predicted according to the first historical cloud occlusion feature, and then the corresponding photovoltaic output predicted value is predicted by using the cloud occlusion feature within the second preset time period.
Specifically, a differential integrated moving average autoregressive (Autoregressive Integrated Moving Average, ARIMA) model, which is one method used to process non-stationary, trending and seasonal time series data, may be used to predict cloud occlusion characteristics over a second predetermined period of time. It combines an Autoregressive (AR) model, an Integrated (I) part, and a Moving Average (MA) model, which are able to capture trends, seasonal, and other periodicity in the data. Wherein the autoregressive model predicts the current value using past observations, thereby capturing the autocorrelation in the time series; the integration part represents the difference operation on the original data, and by means of the difference, the non-stationary time sequence can be converted into a stationary sequence, which helps to eliminate the trend or seasonal influence; the moving average model predicts the current value using a linear combination of past error terms, representing the relationship between the current value and the error, which in part helps to capture irregularities or randomness in the time series. The cloud blocking characteristics in the second preset time period with higher prediction accuracy can be obtained by inputting the first historical cloud blocking characteristics into a differential integration moving average autoregressive model trained in advance, so that the corresponding photovoltaic output predicted value can be predicted according to the cloud blocking characteristics.
Step 140, acquiring first weather information in a second preset time period;
it will be appreciated that the photovoltaic output is affected not only by the cloud cover characteristics, but also by various weather information, such as solar irradiance, temperature, humidity, wind speed, air pressure, etc. When the solar irradiance is higher, the photovoltaic panel on the photovoltaic station is subjected to stronger illumination, the photovoltaic output is larger, namely more electric energy is generated; the high temperature also reduces the conversion efficiency of the electric energy, so that the photovoltaic output is reduced; meanwhile, the humidity, the wind speed and the air pressure can also have certain influence on the power generation performance of the photovoltaic station, the high humidity can influence the working efficiency of electronic elements in the photovoltaic panel, the wind speed can influence the heat dissipation effect of the photovoltaic panel, and the air pressure change can influence the water vapor content in the atmosphere and the propagation of solar radiation. Therefore, the embodiment of the application acquires the first weather information in the second preset time period, the first weather information can be acquired from the numerical forecasting results (Numerical Weather Prediction, NWP), the first weather information comprises solar irradiance, temperature, humidity, wind speed, air pressure and the like, and the photovoltaic output is predicted by combining the first weather information, so that the influence of high temperature, overcast and rainy or other bad weather on the photovoltaic output prediction can be reduced, and the prediction precision is effectively improved.
And 150, inputting cloud blocking characteristics and first atmospheric information into a preset model to obtain a photovoltaic output predicted value, wherein the preset model is obtained based on ConvTransformer model training.
It will be appreciated that the conv transducer model is a hybrid architecture that combines convolutional neural networks (Convolutional Neural Network, CNN) and transducer models, which combines the advantages of both models, and is suitable for many types of tasks, such as sequence data and image data. The ConvTransformer model can capture more global information through the combination of the convolutional neural network and the Transformer model, the photovoltaic output data usually has strong timeliness, the traditional convolutional neural network cannot capture global information well when processing the time sequence data, and the ConvTransformer model can capture long-term dependence and global modes in the time sequence data well through combining with the Transformer model, so that the model can capture complex modes and trends in the photovoltaic output data well; and the ConvTransformer model adopts a parallel computing structure, so that the ConvTransformer model can be used for high-efficiency training on a large-scale distributed cluster, and can be used for rapidly and efficiently processing large-scale photovoltaic output data.
Specifically, the cloud cover characteristics and the first atmospheric information obtained in the embodiment are input into a preset model obtained based on ConvTransformer model training, so that a photovoltaic output predicted value with high accuracy can be predicted.
In an embodiment of the present application, the training step of the preset model includes:
acquiring a plurality of second historical satellite cloud pictures, second weather information and corresponding photovoltaic output historical values in a third preset time period in a second target area;
extracting a second historical cloud cluster shielding characteristic in a second historical satellite cloud picture;
training according to the second weather information, the second historical cloud cover characteristic and the photovoltaic output historical value to obtain a preset model based on the ConvTransformer model.
Specifically, a plurality of second historical satellite cloud charts, second weather information and corresponding photovoltaic output historical values in a third preset time period in a second target area are firstly obtained, wherein the second target area and the first target area can be the same or different, and are set according to actual requirements; the satellite cloud image can be obtained from a satellite, the second weather information can be obtained from a numerical forecasting result, and the photovoltaic output historical value can be obtained from weather station historical monitoring data near the second target area. Carrying out image preprocessing on the second historical satellite cloud image by using radian histogram equalization, removing Gaussian noise in the equalized second historical satellite cloud image by using mean value filtering, removing salt and pepper noise in the equalized second historical satellite cloud image by using median filtering, carrying out image segmentation on the equalized and filtered second historical satellite cloud image by using a gradient vector flow model, extracting second historical cloud occlusion features of the segmented second historical satellite cloud image by using a gray level co-occurrence matrix, inputting second weather information, the second historical cloud occlusion features and photovoltaic output historical values into an initial model based on a ConvTransformer model, and training the initial model to obtain a preset model based on the ConvTransformer model.
The preset model based on the ConvTransformer model is obtained through training according to the second weather information, the second historical cloud occlusion characteristics and the photovoltaic output historical value, so that the model can be ensured to have stronger light output prediction capability.
According to the photovoltaic output prediction method, the first historical cloud cluster shielding characteristic is obtained through the first historical satellite cloud image, the photovoltaic output is predicted by using the preset model based on the ConvTransformer model in combination with the first weather information, and the influence of high temperature, overcast and rainy, cloud cluster shielding or other bad weather on the photovoltaic output prediction can be reduced, so that the prediction precision of the photovoltaic output is effectively improved.
Fig. 2 schematically shows a schematic structural diagram of a photovoltaic output predicting device according to an embodiment of the present application. As shown in fig. 2, in an embodiment of the present application, there is provided a photovoltaic output predicting apparatus, including:
a memory 210 configured to store instructions; and
the processor 220 is configured to call instructions from the memory 210 and when executing the instructions is capable of implementing the photovoltaic output prediction method according to any of the first aspects.
Specifically, in embodiments of the present application, the processor 220 may be configured to:
acquiring a plurality of first historical satellite cloud pictures in a first preset time period in a first target area;
extracting first historical cloud cluster shielding features in a first historical satellite cloud picture;
according to the first historical cloud cover characteristics, predicting the cloud cover characteristics in a second preset time period;
acquiring first weather information in a second preset time period;
and inputting the cloud cover characteristics and the first atmospheric information into a preset model to obtain a photovoltaic output predicted value, wherein the preset model is obtained based on ConvTransformer model training.
It can be understood that the photovoltaic output prediction device provided in the embodiment of the present application can implement each process of the photovoltaic output prediction method in the above embodiment, and can achieve the same technical effect, so that repetition is avoided, and no further description is provided here.
The embodiment of the application also provides an electronic device, which may include:
the photovoltaic output predicting device according to the above embodiment.
It can be appreciated that the electronic device provided in the embodiments of the present application includes the photovoltaic output prediction apparatus according to the above embodiments, and may achieve the same technical effects, so that repetition is avoided and no further description is provided herein.
The embodiment of the application further provides a machine-readable storage medium, on which instructions are stored, the instructions are configured to enable a machine to execute the photovoltaic output prediction method according to the above embodiment, and the same technical effects can be achieved, so that repetition is avoided, and further description is omitted.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.
Claims (10)
1. A method of photovoltaic output prediction, the method comprising:
acquiring a plurality of first historical satellite cloud pictures in a first preset time period in a first target area;
extracting first historical cloud cluster shielding features in the first historical satellite cloud picture;
according to the first historical cloud cover characteristics, predicting cloud cover characteristics in a second preset time period;
acquiring first weather information in the second preset time period;
and inputting the cloud blocking characteristics and the first atmospheric information into a preset model to obtain the photovoltaic output predicted value, wherein the preset model is obtained based on ConvTransformer model training.
2. The method of claim 1, wherein the extracting the first historical cloud cover feature in the first historical satellite cloud map comprises:
image segmentation is carried out on the first historical satellite cloud image by using a gradient vector flow model, so that a plurality of historical cloud image in the first historical satellite cloud image are obtained;
and extracting features of the plurality of historical cloud cluster images by using a gray level co-occurrence matrix to obtain first historical cloud cluster shielding features in the first historical satellite cloud image.
3. The method of claim 1, further comprising, prior to said extracting a first historical cloud occlusion feature in the first historical satellite cloud map:
and preprocessing the first historical satellite cloud image.
4. The method of claim 3, wherein the preprocessing the first historical satellite cloud map comprises:
performing gray histogram equalization on the first historical satellite cloud picture to obtain an equalized first historical satellite cloud picture;
and carrying out smooth filtering processing on the equalized first historical satellite cloud image.
5. The method of claim 4, wherein smoothing the equalized first historical satellite cloud pattern comprises:
removing Gaussian noise in the equalized first historical satellite cloud picture by means of mean filtering;
and removing the spiced salt noise in the equalized first historical satellite cloud picture by using median filtering.
6. The method of claim 1, wherein the first historical cloud occlusion feature comprises a historical cloud thickness and a historical cloud texture.
7. The method of claim 1, wherein the training step of the predetermined model comprises:
acquiring a plurality of second historical satellite cloud pictures, second weather information and corresponding photovoltaic output historical values in a third preset time period in a second target area;
extracting second historical cloud cluster shielding features in the second historical satellite cloud picture;
and training according to the second weather information, the second historical cloud occlusion characteristics and the photovoltaic output historical value to obtain the preset model based on the ConvTransformer model.
8. A photovoltaic output predicting device, comprising:
a memory configured to store instructions; and
a processor configured to invoke the instructions from the memory and when executing the instructions is capable of implementing the photovoltaic output prediction method according to any of claims 1 to 7.
9. An electronic device, comprising:
the photovoltaic output predicting device of claim 8.
10. A machine-readable storage medium having stored thereon instructions for causing a machine to perform the photovoltaic output prediction method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311724642.1A CN117691592A (en) | 2023-12-14 | 2023-12-14 | Photovoltaic output prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311724642.1A CN117691592A (en) | 2023-12-14 | 2023-12-14 | Photovoltaic output prediction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117691592A true CN117691592A (en) | 2024-03-12 |
Family
ID=90134915
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311724642.1A Pending CN117691592A (en) | 2023-12-14 | 2023-12-14 | Photovoltaic output prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117691592A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111815038A (en) * | 2020-06-24 | 2020-10-23 | 山东大学 | Photovoltaic ultra-short term prediction method and system |
| CN112926789A (en) * | 2021-03-17 | 2021-06-08 | 阳光电源股份有限公司 | Satellite cloud picture prediction method, prediction device and readable storage medium |
| CN115222089A (en) * | 2022-05-30 | 2022-10-21 | 西南交通大学 | Road traffic jam prediction method, device, equipment and readable storage medium |
| US20220373984A1 (en) * | 2021-05-19 | 2022-11-24 | Shandong University | Hybrid photovoltaic power prediction method and system based on multi-source data fusion |
| 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 |
| CN115564100A (en) * | 2022-09-22 | 2023-01-03 | 济南大学 | Photovoltaic power prediction method, system and equipment |
| CN115660887A (en) * | 2022-09-27 | 2023-01-31 | 国网四川省电力公司成都供电公司 | Photovoltaic output prediction method and system based on limited weather forecast information |
| CN115689021A (en) * | 2022-11-01 | 2023-02-03 | 湖南防灾科技有限公司 | Method, device, storage medium and processor for predicting power of photovoltaic power station |
| CN116191680A (en) * | 2023-04-25 | 2023-05-30 | 丰德信电力发展股份有限公司 | Monitoring management system applied to photovoltaic power generation |
| CN116914730A (en) * | 2023-07-04 | 2023-10-20 | 中国华能集团清洁能源技术研究院有限公司 | Day-ahead photovoltaic power prediction method and device based on wide-area satellite cloud image data |
-
2023
- 2023-12-14 CN CN202311724642.1A patent/CN117691592A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111815038A (en) * | 2020-06-24 | 2020-10-23 | 山东大学 | Photovoltaic ultra-short term prediction method and system |
| CN112926789A (en) * | 2021-03-17 | 2021-06-08 | 阳光电源股份有限公司 | Satellite cloud picture prediction method, prediction device and readable storage medium |
| US20220373984A1 (en) * | 2021-05-19 | 2022-11-24 | Shandong University | Hybrid photovoltaic power prediction method and system based on multi-source data fusion |
| CN115222089A (en) * | 2022-05-30 | 2022-10-21 | 西南交通大学 | Road traffic jam prediction method, device, equipment and readable storage medium |
| 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 |
| CN115564100A (en) * | 2022-09-22 | 2023-01-03 | 济南大学 | Photovoltaic power prediction method, system and equipment |
| CN115660887A (en) * | 2022-09-27 | 2023-01-31 | 国网四川省电力公司成都供电公司 | Photovoltaic output prediction method and system based on limited weather forecast information |
| CN115689021A (en) * | 2022-11-01 | 2023-02-03 | 湖南防灾科技有限公司 | Method, device, storage medium and processor for predicting power of photovoltaic power station |
| CN116191680A (en) * | 2023-04-25 | 2023-05-30 | 丰德信电力发展股份有限公司 | Monitoring management system applied to photovoltaic power generation |
| CN116914730A (en) * | 2023-07-04 | 2023-10-20 | 中国华能集团清洁能源技术研究院有限公司 | Day-ahead photovoltaic power prediction method and device based on wide-area satellite cloud image data |
Non-Patent Citations (2)
| Title |
|---|
| 刘伟: "高分辨率SAR图像海洋目标识别", 31 May 2022, 国防工业出版社, pages: 12 - 14 * |
| 张秀彬: "发明解析论", 30 June 2014, 上海交通大学出版社, pages: 323 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Huang et al. | Cloud motion estimation for short term solar irradiation prediction | |
| CN110222787B (en) | Multi-scale target detection method and device, computer equipment and storage medium | |
| CN108921799B (en) | Remote sensing image thin cloud removing method based on multi-scale collaborative learning convolutional neural network | |
| CN111428748B (en) | HOG feature and SVM-based infrared image insulator identification detection method | |
| CN111460968B (en) | Unmanned aerial vehicle identification and tracking method and device based on video | |
| US20170031056A1 (en) | Solar Energy Forecasting | |
| CN111523414A (en) | Face recognition method and device, computer equipment and storage medium | |
| CN111462206A (en) | Monocular structure light depth imaging method based on convolutional neural network | |
| CN109949209B (en) | Rope detection and removal method based on deep learning | |
| CN110136162B (en) | Unmanned aerial vehicle visual angle remote sensing target tracking method and device | |
| CN112989995B (en) | Text detection method and device and electronic equipment | |
| CN113761999A (en) | Target detection method and device, electronic equipment and storage medium | |
| Liu et al. | Surface defect detection of solar cells based on feature pyramid network and GA-faster-RCNN | |
| Feng et al. | A novel saliency detection method for wild animal monitoring images with WMSN | |
| CN117576724A (en) | Unmanned plane bird detection method, system, equipment and medium | |
| CN117409083B (en) | Cable terminal identification method and device based on infrared image and improved YOLOV5 | |
| CN113807185B (en) | Data processing method and device | |
| CN111915558A (en) | Pin state detection method for high-voltage transmission line | |
| CN113902792A (en) | Building height detection method and system based on improved RetinaNet network and electronic equipment | |
| CN116740375A (en) | Image feature extraction method, system and medium | |
| CN117691592A (en) | Photovoltaic output prediction method and device, electronic equipment and storage medium | |
| AU2020272936B2 (en) | Methods and systems for crack detection using a fully convolutional network | |
| CN113496159B (en) | Multi-scale convolution and dynamic weight cost function smoke target segmentation method | |
| Li et al. | RFID tag group recognition based on motion blur estimation and YOLOv2 improved by Gaussian algorithm | |
| CN117292120B (en) | Light-weight visible light insulator target detection method and system |
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 |