Simulation method and device for full-year photovoltaic scene, electronic equipment and storage medium

Abstract

The invention relates to an annual photovoltaic scene simulation method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining historical climate data of a target region; inputting the historical climate data into a pre-constructed photovoltaic power generation model to obtain historical annual photovoltaic power generation scene data of the target region; and constructing a model training set by using the historical full-year photovoltaic power generation scene data, training a pre-constructed initial full-year photovoltaic scene simulation model by using the model training set to obtain an actual full-year photovoltaic scene simulation model, and calculating full-year random weather scene data of the target region by using the actual full-year photovoltaic scene simulation model. Therefore, the technical problems that an explicit model lacks flexibility of capturing randomness and diversity of a generated scene, uncertainty of photovoltaic data is difficult to accurately describe by using a mathematical formula, and an implicit model is difficult to completely reflect complexity and diversity of a real scene in related technologies are solved.

Description

Technical Field

[0001] The present application relates to the technical field of electrical digital data processing, and in particular to a simulation method, device, electronic device and storage medium for a year-round photovoltaic scene. Background Art

[0002] Photovoltaic power generation faces many uncertainties. For example, weather fluctuations lead to unstable photovoltaic power generation, which may cause fluctuations in power supply, thus having a significant impact on the planning, operation and reliability analysis of energy systems. Therefore, how to analyze the uncertainty of photovoltaic power generation is an issue that needs to be studied urgently.

[0003] Scenario analysis is the main method for uncertainty analysis of photovoltaic power generation. Through photovoltaic scenario analysis, decision makers can better understand the challenges and opportunities that photovoltaic power generation projects may face in different environments.

[0004] In terms of scene generation, in the relevant technologies, explicit methods based on probability statistical models have small model capacity and can only capture the global overall digital characteristics of the data. It is difficult to model the time series characteristics of the data and lacks the flexibility to capture the randomness and diversity of the generated scenes. Secondly, photovoltaic data is highly uncertain due to the influence of weather, and the explicit model is overly dependent on the probability distribution function set in advance, making it difficult to accurately describe the uncertainty of photovoltaic data using mathematical formulas. At the same time, it is difficult for explicit models to model the correlation between data variables, especially between variables with large differences in distribution patterns, which will reduce the accuracy of photovoltaic scene generation. In addition, photovoltaic data have different distribution patterns under different spatiotemporal backgrounds, and it is difficult for explicit models to generalize photovoltaic data in various scenarios.

[0005] As for implicit methods represented by deep learning, the implicit model has limitations caused by assumptions about the distribution of real data, lack of creativity and diversity in generated data, and reliance on known data. When generating scene data, it may not be able to fully reflect the complexity and diversity of real scenes.

[0006] In summary, in the relevant technologies, explicit models lack the flexibility to capture the randomness and diversity of generated scenarios, and it is difficult to accurately describe the uncertainty of photovoltaic data using mathematical formulas. Implicit models are difficult to fully reflect the complexity and diversity of real scenarios and need to be improved. Summary of the invention

[0007] The present application provides a simulation method, device, electronic device and storage medium for photovoltaic scenes throughout the year to solve the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scenes, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of real scenes.

[0008] The first aspect of the present application provides a method for simulating a year-round photovoltaic scene, which is applied to a model building stage, wherein the method comprises the following steps: acquiring historical climate data of a target area; inputting the historical climate data into a pre-constructed photovoltaic power generation model to obtain historical year-round photovoltaic power generation scene data of the target area; constructing a model training set using the historical year-round photovoltaic power generation scene data, and using the model training set to train a pre-constructed initial year-round photovoltaic scene simulation model to obtain an actual year-round photovoltaic scene simulation model, and using the actual year-round photovoltaic scene simulation model to calculate the year-round random weather scene data of the target area.

[0009] Optionally, in one embodiment of the present application, constructing a model training set using the historical full-year photovoltaic power generation scene data includes: sorting the historical full-year photovoltaic power generation scene data in a preset order to obtain a sorting result; and constructing the model training set based on the sorting result.

[0010] Optionally, in one embodiment of the present application, the use of the model training set to train a pre-constructed initial year-round photovoltaic scene simulation model includes: inputting any input data in the model training set into the initial year-round photovoltaic scene simulation model to obtain a simulation result; calculating the discrimination loss based on the simulation result and the corresponding real data in the historical year-round photovoltaic power generation scene data to obtain a calculation result; and using the calculation result to update the model parameters of the initial year-round photovoltaic scene simulation model until the updated initial year-round photovoltaic scene simulation model meets the preset convergence conditions to obtain the actual year-round photovoltaic scene simulation model.

[0011] Optionally, in one embodiment of the present application, the historical climate data includes at least one of historical temperature data with hourly accuracy throughout the year, direct radiation data with hourly accuracy throughout the year, and scattered radiation data with hourly accuracy throughout the year.

[0012] The second aspect of the present application provides a method for simulating a year-round photovoltaic scene, which is applied to the model use stage, wherein the method includes the following steps: obtaining current climate data of the target area; inputting the current climate data into a pre-built year-round photovoltaic scene simulation model to obtain the year-round random weather scene data of the target area, wherein the year-round photovoltaic scene simulation model is trained by the historical year-round photovoltaic power generation scene data of the target area.

[0013] The third aspect of the present application provides a simulation device for a year-round photovoltaic scene, which is applied to a model building stage, wherein the device includes: an acquisition module for acquiring historical climate data of a target area; a calculation module for inputting the historical climate data into a pre-built photovoltaic power generation model to obtain historical year-round photovoltaic power generation scene data of the target area; a training module for constructing a model training set using the historical year-round photovoltaic power generation scene data, and using the model training set to train a pre-built initial year-round photovoltaic scene simulation model to obtain an actual year-round photovoltaic scene simulation model, and using the actual year-round photovoltaic scene simulation model to calculate the target area's year-round random weather scene data.

[0014] Optionally, in one embodiment of the present application, the calculation module includes: a sorting unit, used to sort the historical annual photovoltaic power generation scene data in a preset order to obtain a sorting result; and a construction unit, used to construct the model training set based on the sorting result.

[0015] Optionally, in one embodiment of the present application, the training module includes: a first calculation unit, used to input any input data in the model training set into the initial full-year photovoltaic scene simulation model to obtain a simulation result; a second calculation unit, used to calculate the discrimination loss based on the simulation result and the corresponding real data in the historical full-year photovoltaic power generation scene data to obtain a calculation result; an updating unit, used to update the model parameters of the initial full-year photovoltaic scene simulation model using the calculation result until the updated initial full-year photovoltaic scene simulation model meets the preset convergence condition to obtain the actual full-year photovoltaic scene simulation model.

[0016] Optionally, in one embodiment of the present application, the historical climate data includes at least one of historical temperature data with hourly accuracy throughout the year, direct radiation data with hourly accuracy throughout the year, and scattered radiation data with hourly accuracy throughout the year.

[0017] The fourth aspect of the present application provides a simulation device for a year-round photovoltaic scene, which is applied to the model use stage, wherein the device includes: an acquisition module for acquiring current climate data of a target area; a simulation module for inputting the current climate data into a pre-built year-round photovoltaic scene simulation model to obtain the year-round random weather scene data of the target area, wherein the year-round photovoltaic scene simulation model is trained by historical year-round photovoltaic power generation scene data of the target area.

[0018] The fifth aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for simulating the photovoltaic scene throughout the year as described in the above embodiment.

[0019] A sixth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the simulation method of the year-round photovoltaic scene as described in the above embodiment.

[0020] The seventh aspect of the present application provides a computer program product, including a computer program, which, when executed, is used to implement the above-mentioned method for simulating the photovoltaic scene throughout the year.

[0021] The embodiment of the present application can use the historical climate data of the target area, obtain the historical annual photovoltaic power generation scene data of the target area through model calculation, and then use the historical annual photovoltaic power generation scene data to build a model training set, realize the model training of the annual photovoltaic scene simulation model, so as to obtain the annual photovoltaic scene simulation model that can be used for scene generation, and improve the accuracy of the simulation by generating adversarial networks through style transfer. Thus, the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0022] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0024] Figure 1 A flowchart of a method for simulating a photovoltaic scene throughout the year provided according to an embodiment of the present application;

[0025] Figure 2It is a schematic diagram of the principle of a method for simulating a photovoltaic scene throughout the year according to an embodiment of the present application;

[0026] Figure 3 A schematic diagram of the structure of a simulation device for a year-round photovoltaic scene provided according to an embodiment of the present application;

[0027] Figure 4 A flowchart of another method for simulating a year-round photovoltaic scene according to an embodiment of the present application;

[0028] Figure 5 It is a structural schematic diagram of another year-round photovoltaic scene simulation device provided according to an embodiment of the present application;

[0029] Figure 6 It is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0031] The following describes the simulation method, device, electronic device and storage medium of the photovoltaic scene of the embodiment of the present application with reference to the accompanying drawings. In view of the technical problems that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene in the related technology mentioned in the above background technology, and it is difficult to accurately describe the uncertainty of photovoltaic data using mathematical formulas, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene, the present application provides a simulation method for photovoltaic scenes throughout the year, in which the historical climate data of the target area can be used to obtain the historical photovoltaic power generation scene data of the target area throughout the year through model calculation, and then the historical photovoltaic power generation scene data throughout the year is used to construct a model training set to achieve model training of the photovoltaic scene simulation model throughout the year, so as to obtain a photovoltaic scene simulation model throughout the year that can be used for scene generation, and the simulation accuracy is improved by generating an adversarial network through style transfer. Thus, the technical problems that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene in the related technology, and it is difficult to accurately describe the uncertainty of photovoltaic data using mathematical formulas, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0032] Specifically, Figure 1 A schematic flow chart of a method for simulating a year-round photovoltaic scene provided in an embodiment of the present application.

[0033] like Figure 1As shown, the simulation method of the photovoltaic scene throughout the year is applied to the model building stage, wherein the method includes the following steps:

[0034] In step S101, historical climate data of a target area is obtained.

[0035] In the actual implementation process, the embodiment of the present application can collect historical climate data of the target area, wherein the historical climate data can be accurate to the historical full-year hourly accuracy to ensure the subsequent model accuracy.

[0036] Optionally, in one embodiment of the present application, the historical climate data includes at least one of historical temperature data with hourly accuracy throughout the year, direct radiation data with hourly accuracy throughout the year, and scattered radiation data with hourly accuracy throughout the year.

[0037] Among them, historical climate data may include historical year-round hourly accuracy temperature, year-round hourly accuracy direct radiation, and year-round hourly accuracy scattered radiation of the target area.

[0038] In step S102, historical climate data is input into a pre-built photovoltaic power generation model to obtain historical photovoltaic power generation scenario data for the entire year in the target area.

[0039] Furthermore, the embodiments of the present application can obtain historical annual photovoltaic power generation scenario data of the target area by inputting historical climate data into the photovoltaic power generation model.

[0040] Among them, the construction of the photovoltaic power generation model can be as follows:

[0041] The total amount of solar radiation received by the photovoltaic panels is

[0042] I PV =I direct,PV +I diffuse,PV +I reflect,PV ,

[0043] Among them, I PV is the total solar radiation received by the photovoltaic panel, I direct,PV is the amount of direct radiation received by the photovoltaic panel, I diffuse,PV is the amount of scattered radiation received by the photovoltaic panel, I reflect,PV is the amount of ground reflected radiation received by the photovoltaic panels.

[0044] The amount of direct radiation received by the photovoltaic panel can be calculated as

[0045]

[0046] Where γ is the angle between the direct radiation and the photovoltaic panel, h is the solar elevation angle, and I direct is direct radiation on a horizontal surface.

[0047] h can be calculated as

[0048]

[0049] in, is the latitude of the PV station, δ is the solar declination, ω is the hour angle, T is the day number, LAT is the local apparent solar time, t is the current second, ψ is the longitude of the PV station, ψ 0 is the standard time accuracy, which can be the longitude of the standard time of the country where the target area is located, such as 116.407°, and E is the time equation in minutes.

[0050] γ can be calculated as

[0051] γ=arccos(sinh·cosβ+cosh·sinβ·cos(A s -A PV )),

[0052] Where β is the tilt angle of the photovoltaic panel, such as 26.5°, A s is the solar azimuth position; A PV is the azimuth of the photovoltaic panel. Due south can be defined as A PV of 0°.

[0053] The amount of scattered radiation received by the photovoltaic panel can be calculated as

[0054]

[0055] Among them, I diffuse is the diffuse radiation on a horizontal surface.

[0056] The amount of ground reflected radiation received by the photovoltaic panel can be calculated as

[0057]

[0058] I H =I direct +I diffuse ,

[0059] Among them, I H is the total radiation on the horizontal surface, and ρ is the ground albedo, which can be set to 0.10 in the embodiment of the present application.

[0060] Through the above steps, the embodiment of the present application can convert the weather data of the target area into the amount of solar radiation received by the photovoltaic panels in the target area, and further calculate the photovoltaic power generation power according to the actual situation of the photovoltaic power station in the target area.

[0061] In step S103, a model training set is constructed using historical full-year photovoltaic power generation scene data, and a pre-constructed initial full-year photovoltaic scene simulation model is trained using the model training set to obtain an actual full-year photovoltaic scene simulation model, and the actual full-year photovoltaic scene simulation model is used to calculate the full-year random weather scene data of the target area.

[0062] As a possible implementation method, the embodiment of the present application can construct a model training set based on historical photovoltaic power generation scene data throughout the year to introduce the idea of ​​style transfer in the generation network, use adaptive instance normalization to adjust the statistical properties of the generated features, and achieve improved diversity of generated samples.

[0063] It is understandable that the Generative Adversarial Network is a deep learning framework consisting of two neural networks: the generator (G) and the discriminator (D), which compete with each other during the training process. For the minimax game, we look at it separately, that is, given the generator, the value function V(D, G) is maximized, and given the discriminator, V(D, G) is minimized. Through continuous confrontation and learning, the generator can generate realistic data samples P gen (x), while the discriminator strives to distinguish the real data P ori (x) and generate data P gen (x). The game process can be shown as follows:

[0064]

[0065] Since photovoltaic scene data is time-series data, related technologies lack the ability to handle time dependency. Therefore, convolutional neural networks used to generate adversarial networks are difficult to capture time-series features.

[0066] In contrast, the LSTM network is a neural network architecture that is more suitable for processing time series data. The LSTM network has a cyclic structure that can maintain a hidden state inside the network to store information from past moments, and update the hidden state at new moments to include new information. This structure enables the LSTM network to process time series data of indefinite length and capture the temporal dependencies in time series data. The bidirectional LSTM network is a deep learning model that is a variant of the LSTM network. Unlike the LSTM network in related technologies, the bidirectional LSTM network considers not only the information from past moments but also the information from future moments when processing sequence data, which can better capture the contextual information in the sequence data.

[0067]

[0068] Among them, LSTM forwardRepresents a positive long short-term memory network unit, LSTM backward Represents a reverse long short-term memory network unit. t is the input of the photovoltaic scene sequence at time t, and They are the hidden layer states of the forward and reverse long short-term memory networks at time t, respectively.

[0069] The final output can be concatenated through the forward and reverse hidden states.

[0070]

[0071] Style transfer is an image processing technology that can combine the content of one image with the style of another image. Combining style transfer with a generative adversarial network can enhance the diversity of generated images, so that the generated images have different styles and contents. Inspired by this, the embodiment of the present application can introduce the idea of ​​style transfer into photovoltaic scene generation to ensure the diversity of generated scene data. Adaptive instance normalization is a commonly used style transfer method. It is a technology for normalizing an image, which can adjust the statistical characteristics (such as mean and variance) of an input image to the statistical characteristics of a reference image, thereby realizing the style transfer of the reference image to the input image.

[0072]

[0073] Where x and y are the content image and the encoded reference image, respectively, and σ and μ are the mean and standard deviation, respectively.

[0074] The embodiment of the present application can use adaptive instance normalization in the generator network to transfer the style of the generated features. By introducing the adaptive instance normalization layer in the generator network, the diversity of the generated images can be enhanced, and different reference images will cause the generated images to have different styles.

[0075] Specifically, the generator of the generative adversarial network model based on the style transfer strategy proposed in the embodiment of the present application includes two sub-networks, namely a mapping network and a synthesis network. The generator first maps the real photovoltaic scene and weather data labels to a vector in a latent space through a mapping network. This latent vector encodes the basic features of the input image and weather data. Using this latent vector, the data stream of each block in the synthesis network is processed through the adaptive instance normalization operation mentioned above to achieve the purpose of style mixing, so that the generated photovoltaic scene has greater diversity.

[0076] Specifically, the mapping network consists of several layers of fully connected layers, which have achieved the decoupling of the input photovoltaic scene and weather data features. The obtained latent vector is propagated to each block in the synthetic network. In each block, the data in the input block is first modulated using the latent variables, then the temporal characteristics of the photovoltaic scene are learned through two layers of stacked bidirectional long short-term memory network layers, and finally the format of the block output data is adjusted through a fully connected layer.

[0077] The discriminator is composed of a long short-term memory network, and s-spectral normalization is applied to the kernel of the long short-term memory network, which can improve the stability of the model, prevent mode collapse, and improve the quality of generated samples to a certain extent.

[0078] On this basis, the embodiment of the present application can use high-quality samples, namely the model training set, to perform model training to continuously optimize the model parameters, obtain the actual photovoltaic scene simulation model throughout the year, and then realize the generation of photovoltaic scenes in the target area.

[0079] Optionally, in one embodiment of the present application, a model training set is constructed using historical photovoltaic power generation scene data throughout the year, including: sorting the historical photovoltaic power generation scene data throughout the year in a preset order to obtain a sorting result; and constructing a model training set based on the sorting result.

[0080] It is understandable that in the related art, deep learning models have limitations in processing the temporal nature of photovoltaic data, especially convolutional neural network models cannot effectively capture long-term dependencies. Photovoltaic scene data usually has strong temporal correlation, which leads to the lack of continuity and reasonable temporal changes in the scenes generated by the models in the related art.

[0081] Moreover, the generative model faces difficulties in generating diverse scenarios, and the generated results tend to lack innovation and diversity, and it is difficult to capture extreme or rare photovoltaic scenarios. This makes the model inaccurate in simulating extreme weather conditions and unable to fully reflect the complexity of real scenarios.

[0082] Therefore, in some embodiments, the construction of the model training set can be achieved by randomly sorting the historical photovoltaic power generation scene data throughout the year.

[0083] For example, the embodiment of the present application can obtain a set of input variables of the model according to the target purpose, and use it to obtain a set of input variables of the annual photovoltaic scene simulation model according to the principle of random shuffling of the input variable candidate set, so as to combine the input variable set and the corresponding output variables in the historical annual photovoltaic power generation scene data, that is, the real photovoltaic scene, to construct a model training set. Among them, the target purpose can be the input purpose of the data, the verification purpose of the data, etc.

[0084] That is to say, the embodiment of the present application can obtain a candidate set of input variables of the model according to the target purpose, so as to obtain each historical hourly precision photovoltaic data set respectively, and add them to the model input variable set, randomly sort the elements in the candidate input variables, eliminate the correlation between the elements, and the candidate input variables are excluded from the candidate set of input variables of the prediction model, that is, the time series of the candidate set of input variables is randomly sorted by a random shuffling algorithm to obtain the candidate input variables after each shuffling, and add them to the selected set of input variables of the random photovoltaic scene simulation model throughout the year.

[0085] Optionally, in one embodiment of the present application, a pre-constructed initial year-round photovoltaic scene simulation model is trained using a model training set, including: inputting any input data in the model training set into the initial year-round photovoltaic scene simulation model to obtain a simulation result; calculating the discrimination loss based on the simulation result and the corresponding real data in the historical year-round photovoltaic power generation scene data to obtain a calculation result; and using the calculation result to update the model parameters of the initial year-round photovoltaic scene simulation model until the updated initial year-round photovoltaic scene simulation model meets the preset convergence conditions to obtain the actual year-round photovoltaic scene simulation model.

[0086] In other embodiments, the discriminant network can be used to calculate the discriminant loss of the simulation results obtained from the input data of the model training set and the corresponding real results in the model training set, and use it as the optimization target to update the model parameters. The idea of ​​style transfer is introduced into the generative network, and the statistical properties of the generated features are adjusted by adaptive instance normalization to achieve the improvement of the diversity of generated samples. The generative network continuously optimizes the parameters through the loss signal fed back by the discriminant network to improve the authenticity and diversity of the generated data; at the same time, the discriminant network enhances the discriminant ability of the generated samples by learning to distinguish between real data and generated data. Through the collaborative optimization of the generative network and the discriminant network, the learning module realizes the efficient training of the generative adversarial network based on style transfer, and generates random photovoltaic scene data throughout the year with rich features and highly consistent with the actual scene.

[0087] Combination Figure 2 As shown, the working principle of the simulation method of the photovoltaic scene throughout the year of the embodiment of the present application is described in detail with an embodiment.

[0088] like Figure 2 As shown, the embodiment of the present application may include the following steps:

[0089] Step S201: Collect historical climate data of the target area, including local historical annual hourly precision temperature, annual hourly precision direct radiation, and annual hourly precision diffuse radiation.

[0090] Step S202: inputting historical climate data into the photovoltaic power generation model to obtain historical photovoltaic power generation scene data for the whole year.

[0091] Step S203: obtaining a set of input variables of the model according to the target purpose, and using the set of input variables of the whole year photovoltaic scene simulation model to obtain a set of input variables according to the random shuffling principle of the input variable set to be selected, so as to construct a model training set.

[0092] Step S204: training the input variable set and output variable data of the model training set through a network model based on the style transfer generative adversarial network to obtain a full-year photovoltaic scene simulation model.

[0093] Step S205: obtaining the annual random photovoltaic scene data through the annual photovoltaic scene simulation model.

[0094] It should be noted that the embodiment of the present application can also generate diversified data by inputting random Gaussian noise training generator during the model building stage, and generate new photovoltaic scenarios during the model use stage.

[0095] Specifically, random Gaussian noise is generated by randomly extracting values ​​from a standard normal distribution, with zero mean and unit variance. These noises are used as input to the generator together with current climate data, and the style transfer strategy helps the generator learn features and convert the noise into simulated photovoltaic scene data throughout the year. The input of noise can provide randomness and diversity to the generator, avoiding the generation of similar photovoltaic scene data, thereby enhancing the richness and authenticity of the simulation results.

[0096] In summary, the embodiment of the present application can generate photovoltaic power generation scene data throughout the year by combining historical annual hourly precision temperature, direct radiation and scattered radiation data with the photovoltaic power generation model, and use the obtained photovoltaic power generation scene data as the input variable of the annual random photovoltaic scene simulation model, thereby obtaining high-precision photovoltaic power generation scene data throughout the year. This provides reliable scene data support for the system planning, optimal scheduling and energy management of photovoltaic power stations.

[0097] According to the simulation method of the photovoltaic scene throughout the year proposed in the embodiment of the present application, the historical climate data of the target area can be used to obtain the historical photovoltaic power generation scene data of the target area throughout the year through model calculation, and then the historical photovoltaic power generation scene data throughout the year is used to build a model training set to achieve the model training of the photovoltaic scene simulation model throughout the year, so as to obtain the photovoltaic scene simulation model throughout the year that can be used for scene generation, and improve the accuracy of the simulation by generating an adversarial network through style transfer. Thus, the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0098] Next, a simulation device for a year-round photovoltaic scene proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0099] Figure 3 It is a block diagram of a simulation device for a year-round photovoltaic scene according to an embodiment of the present application.

[0100] like Figure 3 As shown, the simulation device 10 for the photovoltaic scene throughout the year is applied to the model building stage, wherein the device 10 includes: an acquisition module 101 , a calculation module 102 and a training module 103 .

[0101] Specifically, the acquisition module 101 is used to acquire historical climate data of the target area.

[0102] The calculation module 102 is used to input the historical climate data into the pre-built photovoltaic power generation model to obtain the historical photovoltaic power generation scene data of the target area throughout the year.

[0103] The training module 103 is used to construct a model training set using historical full-year photovoltaic power generation scene data, and use the model training set to train a pre-constructed initial full-year photovoltaic scene simulation model to obtain an actual full-year photovoltaic scene simulation model, and use the actual full-year photovoltaic scene simulation model to calculate the full-year random weather scene data of the target area.

[0104] Optionally, in one embodiment of the present application, the calculation module 102 includes: a sorting unit and a construction unit.

[0105] Among them, the sorting unit is used to sort the historical full-year photovoltaic power generation scene data according to a preset order to obtain a sorting result.

[0106] The construction unit is used to construct a model training set based on the sorting results.

[0107] Optionally, in one embodiment of the present application, the training module 103 includes: a first calculation unit, a second calculation unit and an update unit.

[0108] Among them, the first calculation unit is used to input any input data in the model training set into the initial full-year photovoltaic scene simulation model to obtain a simulation result.

[0109] The second calculation unit is used to calculate the loss based on the simulation result and the corresponding real data in the historical full-year photovoltaic power generation scene data to obtain the calculation result.

[0110] The updating unit is used to update the model parameters of the initial full-year photovoltaic scene simulation model using the calculation results until the updated initial full-year photovoltaic scene simulation model meets the preset convergence conditions to obtain the actual full-year photovoltaic scene simulation model.

[0111] Optionally, in one embodiment of the present application, the historical climate data includes at least one of historical temperature data with hourly accuracy throughout the year, direct radiation data with hourly accuracy throughout the year, and scattered radiation data with hourly accuracy throughout the year.

[0112] It should be noted that the aforementioned explanation of the embodiment of the simulation method for the year-round photovoltaic scene is also applicable to the simulation device for the year-round photovoltaic scene of this embodiment, and will not be repeated here.

[0113] According to the simulation device of the photovoltaic scene throughout the year proposed in the embodiment of the present application, the historical climate data of the target area can be used to obtain the historical photovoltaic power generation scene data of the target area throughout the year through model calculation, and then the historical photovoltaic power generation scene data throughout the year is used to build a model training set to achieve the model training of the photovoltaic scene simulation model throughout the year, so as to obtain the photovoltaic scene simulation model throughout the year that can be used for scene generation, and improve the accuracy of the simulation by generating adversarial networks through style transfer. Thus, the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0114] The above is an explanation of the embodiment of the present application at the model building stage. The following is an explanation of the embodiment of the present application at the model use stage.

[0115] Specifically, Figure 4 A schematic flow chart of a method for simulating a year-round photovoltaic scene provided in an embodiment of the present application.

[0116] like Figure 4 As shown, the simulation method of the photovoltaic scene throughout the year is applied to the model use stage, wherein the method comprises the following steps:

[0117] In step S401, current climate data of the target area is obtained.

[0118] In the actual implementation process, the embodiment of the present application can obtain the current climate data of the target area, that is, the area where the photovoltaic scene simulation needs to be performed, such as the current temperature, radiation, and scattered radiation.

[0119] In step S402, the current climate data is input into a pre-built annual photovoltaic scene simulation model to obtain annual random weather scene data of the target area, wherein the annual photovoltaic scene simulation model is trained by historical annual photovoltaic power generation scene data of the target area.

[0120] Furthermore, the embodiment of the present application can generate photovoltaic scenes using a pre-built year-round photovoltaic scene simulation model, that is, obtaining year-round random photovoltaic power generation scene data of the target area based on random Gaussian noise of current climate data to achieve high-precision scene simulation.

[0121] According to the simulation method of the photovoltaic scene throughout the year proposed in the embodiment of the present application, the historical climate data of the target area can be used to obtain the historical photovoltaic power generation scene data of the target area throughout the year through model calculation, and then the historical photovoltaic power generation scene data throughout the year is used to build a model training set to achieve the model training of the photovoltaic scene simulation model throughout the year, so as to obtain the photovoltaic scene simulation model throughout the year that can be used for scene generation, and improve the accuracy of the simulation by generating an adversarial network through style transfer. Thus, the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0122] Next, a simulation device for a year-round photovoltaic scene proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0123] Figure 5 It is a block diagram of a simulation device for a year-round photovoltaic scene according to an embodiment of the present application.

[0124] like Figure 5 As shown, the simulation device 20 for the photovoltaic scene throughout the year is applied in the model use stage, wherein the device 20 includes: an acquisition module 201 and a simulation module 202 .

[0125] Specifically, the acquisition module 201 is used to acquire the current climate data of the target area.

[0126] The simulation module 202 is used to input the current climate data into a pre-built full-year photovoltaic scene simulation model to obtain full-year random weather scene data of the target area, wherein the full-year photovoltaic scene simulation model is trained by the historical full-year photovoltaic power generation scene data of the target area.

[0127] It should be noted that the aforementioned explanation of the embodiment of the simulation method for the year-round photovoltaic scene is also applicable to the simulation device for the year-round photovoltaic scene of this embodiment, and will not be repeated here.

[0128] According to the simulation device of the photovoltaic scene throughout the year proposed in the embodiment of the present application, the historical climate data of the target area can be used to obtain the historical photovoltaic power generation scene data of the target area throughout the year through model calculation, and then the historical photovoltaic power generation scene data throughout the year is used to build a model training set to achieve the model training of the photovoltaic scene simulation model throughout the year, so as to obtain the photovoltaic scene simulation model throughout the year that can be used for scene generation, and improve the accuracy of the simulation by generating adversarial networks through style transfer. Thus, the technical problems in the related art that the explicit model lacks the flexibility to capture the randomness and diversity of the generated scene, and it is difficult to use mathematical formulas to accurately describe the uncertainty of photovoltaic data, and the implicit model is difficult to fully reflect the complexity and diversity of the real scene are solved.

[0129] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0130] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0131] When the processor 602 executes the program, the simulation method of the photovoltaic scene throughout the year provided in the above embodiment is implemented.

[0132] Furthermore, the electronic device further comprises:

[0133] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0134] The memory 601 is used to store computer programs that can be executed on the processor 602 .

[0135] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0136] If the memory 601, the processor 602 and the communication interface 603 are implemented independently, the communication interface 603, the memory 601 and the processor 602 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0137] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0138] The processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0139] This embodiment also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned method for simulating the photovoltaic scene throughout the year is implemented.

[0140] The embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the simulation method of the photovoltaic scene throughout the year provided by the embodiment of the present invention.

[0141] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0142] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0143] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0144] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0145] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0146] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0147] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0148] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for simulating photovoltaic scenes throughout the year, characterized in that: Applied to the model building stage, wherein the method comprises the following steps: Obtain historical climate data for the target area; Inputting the historical climate data into a pre-built photovoltaic power generation model to obtain historical full-year photovoltaic power generation scenario data for the target area; A model training set is constructed using the historical full-year photovoltaic power generation scene data, and a pre-constructed initial full-year photovoltaic scene simulation model is trained using the model training set to obtain an actual full-year photovoltaic scene simulation model, and the actual full-year photovoltaic scene simulation model is used to calculate the full-year random weather scene data of the target area.

2. The method according to claim 1, characterized in that The method of using the historical annual photovoltaic power generation scene data to construct a model training set includes: Sorting the historical annual photovoltaic power generation scene data according to a preset order to obtain a sorting result; The model training set is constructed based on the sorting results.

3. The method according to claim 1, characterized in that The method of using the model training set to train a pre-built initial full-year photovoltaic scene simulation model includes: Inputting any input data in the model training set into the initial full-year photovoltaic scenario simulation model to obtain a simulation result; Calculating the loss based on the simulation result and the corresponding real data in the historical full-year photovoltaic power generation scenario data to obtain a calculation result; The calculation results are used to update the model parameters of the initial full-year photovoltaic scene simulation model until the updated initial full-year photovoltaic scene simulation model meets the preset convergence conditions, thereby obtaining the actual full-year photovoltaic scene simulation model.

4. The method according to claim 1, characterized in that: The historical climate data includes at least one of historical temperature data with hourly accuracy throughout the year, direct radiation data with hourly accuracy throughout the year, and scattered radiation data with hourly accuracy throughout the year.

5. A method for simulating photovoltaic scenes throughout the year, characterized in that: The simulation method of the photovoltaic scene throughout the year as described in any one of claims 1 to 4 is applied to the model use stage, wherein the method comprises the following steps: Obtain current climate data for the target area; The current climate data is input into a pre-built full-year photovoltaic scene simulation model to obtain full-year random weather scene data for the target area, wherein the full-year photovoltaic scene simulation model is trained by historical full-year photovoltaic power generation scene data for the target area.

6. A device for simulating photovoltaic scenes throughout the year, characterized in that: Applied to the model building stage, wherein the device comprises: An acquisition module is used to obtain historical climate data of the target area; A calculation module, used for inputting the historical climate data into a pre-built photovoltaic power generation model to obtain the historical full-year photovoltaic power generation scenario data of the target area; A training module is used to construct a model training set using the historical full-year photovoltaic power generation scene data, and use the model training set to train a pre-constructed initial full-year photovoltaic scene simulation model to obtain an actual full-year photovoltaic scene simulation model, and use the actual full-year photovoltaic scene simulation model to calculate the full-year random weather scene data of the target area.

7. A device for simulating photovoltaic scenes throughout the year, characterized in that: Applied to the model use stage, wherein the device comprises: The acquisition module is used to obtain the current climate data of the target area; A simulation module is used to input the current climate data into a pre-built full-year photovoltaic scene simulation model to obtain the full-year random weather scene data of the target area, wherein the full-year photovoltaic scene simulation model is trained by the historical full-year photovoltaic power generation scene data of the target area.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for simulating a year-round photovoltaic scene as described in any one of claims 1-4 or 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method for simulating a photovoltaic scene throughout the year as described in any one of claims 1-4 or 5.

10. A computer program product, comprising a computer program, characterized in that The computer program is executed to implement the method for simulating a photovoltaic scene throughout the year as claimed in any one of claims 1 to 4 or 5.

Images