Generating capacity prediction method based on photovoltaic power generation system and related device

Abstract

The invention provides a generating capacity prediction method based on a photovoltaic power generation system and a related device, and the method comprises the steps: firstly obtaining photovoltaic power generation system multi-source data including photovoltaic module level electrical data, photovoltaic panel topographic data, and photovoltaic panel shielding object data, and carrying out the local shielding recognition of a photovoltaic module according to the multi-source data, determining the corrected grey-scale map and the positioning data of the shielded battery piece, then combining the positioning data of the shielded battery piece with the multi-source data of the photovoltaic power generation system, calculating the generating capacity of the local shielded battery piece and obtaining a power station level electric quantity predicted value, and finally performing error evaluation on the power station level electric quantity predicted value. When the error threshold value is met, the final generating capacity prediction result is output, the photovoltaic panel topographic data and the photovoltaic panel shielding object data are introduced, and the precise positioning information of the shielded battery piece can be obtained in combination with local shielding recognition, so that the problem of output overestimation or underestimation caused by simplification or neglect of local shielding in an existing method is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation, in particular to a power generation prediction method based on a photovoltaic power generation system and related devices. BACKGROUND

[0002] With the rapid development of global new energy industry, photovoltaic power generation, as a clean and sustainable energy form, its installed capacity and application scenarios are constantly expanding, from large-scale ground photovoltaic power stations to distributed roof power stations, mountain power stations and other complex scenarios. As a core supporting technology for efficient operation of photovoltaic systems and grid dispatching, the accuracy of power generation prediction directly affects the new energy consumption rate and grid stability. At present, the prediction method based on neural network has become one of the mainstream technologies in the field of photovoltaic power generation prediction: such method usually takes the global irradiance, temperature, wind speed and other meteorological data provided by numerical weather prediction (NWP), combined with the historical output data of photovoltaic system, the nominal parameters of components as input, through training long short-term memory network (LSTM), convolutional neural network (CNN) and other models, learning the mapping relationship between meteorological conditions and power generation, and then realizing the power generation prediction in future period.

[0003] However, the existing photovoltaic power generation prediction method based on neural network still has defects in complex application scenarios, for example, photovoltaic panels often face dynamic or static shading such as tree shadow, temporary construction equipment, low-altitude cumulus cloud, flying birds and the like during operation, such shading only affects part of the cell or component, but due to the series-parallel circuit characteristics of photovoltaic components, it will pull down the current output of the entire series branch and overestimate or underestimate the output of part of the photovoltaic panel, ultimately leading to a large deviation between the prediction result of the model and the actual power generation, which is difficult to meet the demand of grid dispatching for high-precision prediction. SUMMARY

[0004] The embodiments of the present application provide a power generation prediction method based on a photovoltaic power generation system and related devices, which can introduce photovoltaic panel terrain data and photovoltaic panel shading data, and obtain accurate positioning information of shaded cells by combining local shading identification, so as to avoid the overestimation or underestimation problem caused by simplification or neglect of local shading in the existing method.

[0005] The first aspect of the embodiments of the present application provides a power generation prediction method based on a photovoltaic power generation system, the method comprising: obtaining photovoltaic power generation system multi-source data, wherein the photovoltaic power generation system multi-source data comprises photovoltaic component level electrical data, photovoltaic panel terrain data, and photovoltaic panel shading data; identifying local shading of the photovoltaic component according to the photovoltaic power generation system multi-source data, determining a corrected gray image and shaded cell positioning data; According to the shaded battery piece positioning data and the photovoltaic power generation system multi-source data, local shaded battery power generation is determined, and a power station level power prediction value is obtained; The power station level power prediction value is subjected to prediction result evaluation, and when the power station level power prediction value prediction value satisfies an error threshold, the power station level power prediction value is output as a final power prediction result.

[0006] In a possible implementation, the local shading of the photovoltaic module is identified according to the photovoltaic power generation system multi-source data, and a corrected gray map is determined, including: The photovoltaic module cell curve data and the cell irradiance proportion data are extracted from the photovoltaic module level electrical data; According to the photovoltaic panel terrain data and the photovoltaic module cell curve data, a micro-terrain feature comprehensive value of each battery piece is determined; According to the shading object type in the photovoltaic panel shading object data, a micro-terrain correction coefficient is determined; According to the cell irradiance proportion data, the micro-terrain feature comprehensive value, and the micro-terrain correction coefficient, a micro-terrain corrected battery piece gray value is determined; According to the shading object moving speed in the photovoltaic panel shading object data, a gray map update period is determined; According to the micro-terrain corrected battery piece gray value and the update period, a corrected gray map is determined in combination with the battery piece position.

[0007] In a possible implementation, the local shading of the photovoltaic module is identified according to the photovoltaic power generation system multi-source data, and shaded battery piece positioning data is determined, including: The photovoltaic module cell curve data is taken as a training sample, and the corrected gray map is taken as a training label, and a variational autoencoder neural network with a fusion attention mechanism is trained; According to the shading object type in the photovoltaic panel shading object data, an attention weight of the variational autoencoder neural network is optimized; Real-time collected photovoltaic module cell curve data is input into the variational autoencoder neural network, and a real-time corrected gray map is output; According to the real-time corrected gray map, the shaded battery piece positioning data is determined.

[0008] In a possible implementation, according to the shaded battery piece positioning data and the photovoltaic power generation system multi-source data, local shaded battery power generation is determined, and a power station level power prediction value is obtained, including: According to the photovoltaic module level electrical data and the shaded battery piece positioning data, a battery piece local shading factor and a battery piece mismatch factor are determined; According to the battery piece local shading factor, an actual irradiance intensity of each battery piece is determined; extracting cell effective area and cell working efficiency from the photovoltaic module level electrical data; determining single cell actual output according to the cell effective area, cell working efficiency and cell mismatch factor; summing up the single cell actual output to obtain module level output; summing up the module level output to obtain power station level power prediction value.

[0009] In a possible implementation, the determining of the cell local shading factor according to the photovoltaic module level electrical data and the shaded cell positioning data comprises: extracting reference irradiance and ground albedo data from the photovoltaic module level electrical data; determining initial cell local shading factor according to the reference irradiance and ground albedo data and the shaded cell positioning data; determining the update range of the initial cell local shading factor according to the shading object moving speed in the photovoltaic panel shading object data, to obtain the cell local shading factor.

[0010] In a possible implementation, the determining of the cell mismatch factor according to the photovoltaic module level electrical data and the shaded cell positioning data comprises: extracting series connection mode data of cells in the module from the photovoltaic module level electrical data; determining the serial branch number k in which the nth cell is located according to the series connection mode data; determining the total number of cells, the number of shaded cells and the cell shading degree data of each shaded cell in the serial branch with the serial branch number k according to the shaded cell positioning data; determining branch level shading influence coefficient according to the total number of cells, the number of shaded cells and the cell shading degree data; determining initial cell mismatch factor according to the branch level shading influence coefficient; correcting the cell mismatch factor according to the shading object type in the photovoltaic panel shading object data, to obtain the cell mismatch factor.

[0011] In a possible implementation, the prediction result evaluation of the power station level power prediction value comprises: performing prediction result evaluation of the power station level power prediction value by using root mean square error and mean absolute error; If the root mean square error is < 50 W and the average absolute error is < 8%, the power station level power prediction value is output as the final power generation prediction result.

[0012] The example provides a power generation prediction method based on a photovoltaic power generation system. First, multi-source data of the photovoltaic power generation system including photovoltaic component level electrical data, photovoltaic panel terrain data, and photovoltaic panel shelter data is obtained. Then, local shelter identification is performed on the photovoltaic component according to the multi-source data to determine a corrected grayscale map and sheltered cell positioning data. Subsequently, the power generation of the locally sheltered cells is calculated and a power station level power prediction value is obtained by combining the sheltered cell positioning data and the multi-source data of the photovoltaic power generation system. Finally, error evaluation is performed on the power station level power prediction value. When the error threshold is met, the final power generation prediction result is output. The introduction of the photovoltaic panel terrain data and the photovoltaic panel shelter data, combined with the local shelter identification to obtain accurate positioning information of the sheltered cells, can avoid the overestimation or underestimation of the output caused by the simplification or neglect of local shelter in existing methods. At the same time, the local irradiation differences of the photovoltaic panel under different terrain conditions can be captured, the prediction deviation caused by the inability of the global terrain data to reflect the micro differences can be reduced, and the demand of grid dispatching for high-precision prediction can be met, thereby improving the new energy consumption rate and ensuring the stability of the power grid operation.

[0013] The second aspect of the embodiment of the present application provides a power generation prediction device of a photovoltaic power generation system. The device comprises: A first acquisition unit is configured to acquire multi-source data of a photovoltaic power generation system. The multi-source data of the photovoltaic power generation system includes photovoltaic component level electrical data, photovoltaic panel terrain data, and photovoltaic panel shelter data. A first processing unit is configured to identify local shelter of a photovoltaic component according to the multi-source data of the photovoltaic power generation system to determine a corrected grayscale map and sheltered cell positioning data. A second processing unit is configured to determine the power generation of locally sheltered cells and obtain a power station level power prediction value according to the sheltered cell positioning data and the multi-source data of the photovoltaic power generation system. A third processing unit is configured to perform prediction result evaluation on the power station level power prediction value. When the power station level power prediction value meets the error threshold, the power station level power prediction value is output as the final power generation prediction result.

[0014] The third aspect of the embodiment of the present application provides a terminal comprising a processor, an input device, an output device, and a memory. The processor, the input device, the output device, and the memory are connected to each other. The memory is configured to store a computer program. The computer program comprises program instructions. The processor is configured to invoke the program instructions to execute the steps of the power generation prediction method based on a photovoltaic power generation system as described in the first aspect of the embodiment of the present application.

[0015] A fourth aspect of the embodiments of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program for electronic data exchange, and the computer program causes a computer to perform some or all of the steps described in the method for predicting power generation of a photovoltaic power generation system according to the first aspect of the embodiments of the present application.

[0016] A fifth aspect of the embodiments of the present application provides a computer program product, which includes a non-transitory computer readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute some or all of the steps described in the method for predicting power generation of a photovoltaic power generation system according to the first aspect of the embodiments of the present application. The computer program product can be a software installation package. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 A total flowchart of the method for predicting power generation of a photovoltaic power generation system is provided for the embodiments of the present application; Figure 2 A total structure diagram of the device for predicting power generation of a photovoltaic power generation system is provided for the embodiments of the present application; Figure 3 A structure diagram of a terminal is provided for the embodiments of the present application; Reference signs: First acquisition unit-1, first processing unit-2, second processing unit-3, third processing unit-4. DETAILED DESCRIPTION

[0019] The technical solutions of the embodiments of the present application will be described clearly and completely in the embodiments of the present application in combination with the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0020] The terms "first", "second", and the like in the description and in the claims of the present application and above-described drawings are used to distinguish different objects, rather than to describe a particular sequential order. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but can optionally further include steps or units not listed, or can optionally further include other steps or units inherent to such processes, methods, products, or devices.

[0021] Reference to "embodiments" in this application means that the particular feature, structure, or characteristic described in connection with an embodiment can be included in at least one embodiment of the application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor does it necessarily refer to a particular embodiment that is preferred over other embodiments. It is expressly understood that the embodiments described in this application can be combined with each other in various combinations.

[0022] In order to better understand the photovoltaic power generation system-based power generation amount prediction method provided by the embodiments of the present application, the scene of applying the photovoltaic power generation system-based power generation amount prediction method will be briefly introduced first. In the prediction method used in the photovoltaic power generation amount prediction, local shading is a key factor leading to a sharp increase in prediction error. Photovoltaic panels often face dynamic or static shading such as tree shadow, temporary construction equipment, low-altitude cumulus cloud, and flying birds during operation. Such shading only affects part of the cell or component, but it can cause the shaded cell to pull down the current output of the entire series branch due to the series-parallel circuit characteristics of the photovoltaic component. Existing prediction methods either do not specifically identify the shading condition and directly attribute the output drop caused by shading to changes in meteorological conditions, or only correct it through a simple shading ratio coefficient without distinguishing the dynamic influence of shading object type and moving speed on the shading range, and without accurately positioning the specific cell position that is shaded. Ultimately, the model cannot accurately quantify the output loss caused by shading, the prediction result deviates greatly from the actual power generation amount, especially in cloudy and complex terrain scenes, the error is large, and it is difficult to meet the demand of grid dispatching for high-precision prediction.

[0023] The photovoltaic power generation system-based power generation amount prediction method is applied to a photovoltaic power generation system-based power generation amount prediction device, Figure 1 A general flowchart of a photovoltaic power generation system-based power generation amount prediction method is shown. As shown in Figure 1 It includes: S1, acquiring photovoltaic power generation system multi-source data, wherein the photovoltaic power generation system multi-source data includes photovoltaic component-level electrical data, photovoltaic panel terrain data, and photovoltaic panel shading object data.

[0024] The photovoltaic module level electrical data includes photovoltaic module cell curve data and cell irradiance ratio data. The photovoltaic module cell curve data can be obtained by deploying a module level IV tester at the junction box of the photovoltaic module. The cell irradiance ratio data can be collected by a micro irradiance sensor.

[0025] The photovoltaic panel terrain data includes slope data, orientation data, altitude data and ground albedo type data. The slope data and the orientation data can be obtained by aerial photography of a UAV combined with three-dimensional modeling. Specifically, a digital elevation model of the power station area is generated by scanning the area with a laser radar carried by a UAV, and then the slope and orientation corresponding to each photovoltaic panel are extracted by GIS software. The altitude data can be extracted from a high-precision GIS map and calibrated in combination with field GPS positioning. The ground albedo type data (such as grassland, cement ground, snow ground, etc.) can be obtained by identifying the pixel classification algorithm of the visible light aerial image of the UAV.

[0026] The photovoltaic panel shelter data includes shelter type data and shelter moving speed data. The shelter type data is used to distinguish between fixed shelters (such as surrounding buildings and long-term trees) and dynamic shelters (such as low-altitude cumulus clouds and flying birds), which can be obtained by deploying a high-definition intelligent camera around the photovoltaic array and using a pre-trained target detection model for real-time identification. The shelter moving speed data is for dynamic shelters, which can be obtained by measuring with a millimeter wave radar matched with the camera. The moving speed of the fixed shelter is set to 0.

[0027] S2, identifying local shading of the photovoltaic module according to the multi-source data of the photovoltaic power generation system, and determining the corrected gray scale image and the positioning data of the shaded cell.

[0028] Specifically, step S2 includes the following steps: S201, extracting photovoltaic module cell curve data and cell irradiance ratio data from the photovoltaic module level electrical data.

[0029] S202, determining the micro-terrain feature comprehensive value of each cell according to the photovoltaic panel terrain data and the photovoltaic module cell curve data.

[0030] The calculation formula of the micro-terrain feature comprehensive value is as follows: wherein, is the micro-terrain feature comprehensive value of the nth battery piece, is the slope weight coefficient of the battery piece, is the orientation weight coefficient of the battery piece, is the weight coefficient of the surface albedo, is the slope factor of the nth battery piece, is the orientation factor of the nth battery piece, is the surface albedo factor of the nth battery piece.

[0031] S203, determining a micro-terrain correction coefficient according to the shelter type in the photovoltaic panel shelter data.

[0032] wherein, when the "shelter type" in the photovoltaic panel shelter data obtained in step S1 is a fixed shelter such as a building or a long-term tree, the micro-terrain correction coefficient takes 0.1, when the shelter type is a dynamic shelter takes 0.05.

[0033] S204, determining a micro-terrain corrected battery piece gray value according to the battery irradiance proportion data, the micro-terrain feature comprehensive value and the micro-terrain correction coefficient.

[0034] wherein, the calculation formula of the micro-terrain corrected battery piece gray value is as follows: wherein, is the micro-terrain corrected gray value of the nth battery piece, and 255 is the maximum value of the gray value, is the irradiance proportion data of the nth battery piece, is the micro-terrain correction coefficient, is the micro-terrain feature comprehensive value of the nth battery piece.

[0035] S205, determining a gray map update period according to the shelter moving speed in the photovoltaic panel shelter data.

[0036] wherein, when the shelter moving speed in the photovoltaic panel shelter data obtained in step S1 is >5km / h, the gray map update period is set to 10 seconds; when the moving speed is ≤5km / h and >0, the update period is set to 60 seconds; when the moving speed =0, the update period is set to 300 seconds, the new period starts timing from the shelter speed data update, and the next gray map generation is triggered after the period.

[0037] S206, determining a corrected gray map according to the micro-terrain corrected battery piece gray value and the update period in combination with the battery piece position.

[0038] In which, firstly, the coordinate position of each cell in the gray image is determined according to the physical structure of the photovoltaic module; then the micro-topography corrected cell gray value Gn calculated in step S204 is filled in the corresponding coordinate; finally, the gray image data is refreshed regularly according to the update period determined in step S205.

[0039] In this example, the size of the corrected gray image is consistent with the arrangement of the photovoltaic module cells (such as 60 cells corresponding to a 6x10 pixel image), and the gray value distribution intuitively reflects the shading degree of each cell - the dark area (low gray value) represents severe shading, and the light area (high gray value) represents slight or no shading, providing clear image input for subsequent shading cell positioning, to generate a corrected gray image in combination with the foregoing results, realizing the visualization and quantification of the shading state.

[0040] S207, taking the photovoltaic module cell curve data as a training sample and the corrected gray image as a training label, training a variational autoencoder neural network fused with an attention mechanism.

[0041] In which, a variational autoencoder (VAE) fused with an attention mechanism can be trained to realize the mapping from cell curve data to shading gray image. Specifically, the photovoltaic module cell curve data extracted in step S201 is taken as a training sample (input layer), each curve contains 200 voltage-current sampling points, and the corrected gray image generated in step S206 is taken as a training label (output layer, size consistent with the arrangement of the cells). The network structure adopts an "encoder-decoder" architecture, the encoder compresses the cell curve data into a latent vector through a 3-layer convolutional neural network (CNN), and the decoder reconstructs the latent vector into a gray image through a 3-layer deconvolution network; at the same time, an attention mechanism module is embedded between the encoder and the decoder, which can strengthen the feature extraction of the shading sensitive area by calculating the correlation between the cell curve features and the gray image area. During the training process, the Adam optimizer can be used to take the weighted sum of the reconstruction loss and the KL divergence loss as the total loss function, and the training period is set to 50 epochs until the validation set loss tends to be stable.

[0042] S208, according to the shading object type in the photovoltaic panel shading object data, optimizing the attention weight of the variational autoencoder neural network.

[0043] When the type of the occlusion obtained in step S1 is a fixed occlusion (such as a building or a tree), the recognition ability of the fixed contour occlusion is strengthened by increasing the attention weight corresponding to the edge feature; when the type of the occlusion is a dynamic occlusion (such as a cumulus cloud or a flying bird), the capture ability of the dynamic change occlusion is enhanced by increasing the attention weight corresponding to the gray scale change rate feature. The weight optimization is realized through an “occlusion type-weight mapping table”, and the weight parameters in the mapping table are calibrated offline based on 100,000 sets of historical data of different occlusion scenes, so as to ensure that the weight adjustment for a certain type of occlusion can improve the recognition accuracy.

[0044] S209, inputting the real-time collected photovoltaic module cell curve data into the variational autoencoder neural network to output a real-time corrected gray scale image.

[0045] The photovoltaic module cell curve data can be collected in real time by a module-level IV tester (the collection frequency is consistent with the gray scale image update period of step S205, such as 10 seconds / time), and after the data is collected, the data is standardized (the current value is scaled to the range of 0-1), and then input into the variational autoencoder neural network trained in step S207 and optimized in step S208; the network calculates by forward propagation to output a real-time corrected gray scale image corresponding to the current time (the size is consistent with step S206, such as 6x10 pixels), which can reflect the occlusion state of each cell in real time (such as when a bird flies over a certain area, the gray scale value of the corresponding pixel decreases instantaneously). In order to ensure real-time performance, GPU acceleration is used in the network inference process, and the generation time of a single gray scale image is controlled to be within 50 ms, meeting the update period requirement.

[0046] S2010, determining the occluded cell positioning data according to the real-time corrected gray scale image.

[0047] The pixel coordinates of the occluded area are converted into the corresponding cell number according to the “cell-pixel coordinate” mapping relationship established in step S206 (for example, in a 6x10 pixel image, the pixel in the second row and the third column corresponds to the 13th cell); and finally, the occluded cell positioning data is output in the form of a list containing all the occluded cell numbers (such as [5, 6, 13, 14]), and the occlusion time of each cell is recorded, which provides a timestamp basis for the dynamic occlusion treatment in the subsequent power generation calculation. For the fuzzy area with a gray scale value near the threshold, the change trend of the continuous 3 frames of gray scale images is combined to assist in the judgment: if the gray scale values of the continuous 3 frames are all less than the threshold, it is determined that the area is occluded; if only one frame is less than the threshold, it is determined that the area is noise and is excluded from the positioning result, so as to reduce the misjudgment.

[0048] S3, determining local shading cell power generation according to the shaded cell positioning data and the multi-source data of the photovoltaic power generation system, to obtain a power station level power prediction value.

[0049] Specifically, the step S3 comprises the following steps: S301, determining a cell local shading factor and a cell mismatch factor according to the photovoltaic module level electrical data and the shaded cell positioning data.

[0050] The cell local shading factor can be determined by comparing the reference irradiance intensity with the actual received irradiance of the shaded cell, and is dynamically updated in combination with the moving speed of the shading object, to reflect the real-time shading degree. The cell mismatch factor can be determined by the distribution characteristics (such as the number of shaded cells and the shading degree) of the shaded cells in the series branch, and is corrected according to the type of the shading object, to reflect the output loss caused by the barrel effect in the series circuit.

[0051] S302, determining the actual irradiance intensity of each cell according to the cell local shading factor.

[0052] The calculation formula of the actual irradiance intensity of each cell is as follows: Wherein, is the actual irradiance intensity of the nth cell, is the reference irradiance intensity, is the local shading factor of the nth cell, is the ground reflection gain factor.

[0053] S303, extracting the cell effective area and the cell working efficiency from the photovoltaic module level electrical data.

[0054] S304, determining the actual output of a single cell according to the cell effective area, the cell working efficiency and the cell mismatch factor.

[0055] The calculation formula of the actual output of a single cell is as follows: Wherein, is the actual output of the nth cell, is the effective area of the nth cell, is the actual irradiance intensity determined in step S302, is the cell working efficiency, is the cell mismatch factor.

[0056] S305, summing up the actual output of a single cell to obtain the module level output.

[0057] wherein the formula of the component-level power output is as follows: wherein, is the component-level power output, is the total number of battery pieces of a single component, is the actual power output of the nth battery piece.

[0058] S306, sum the component-level power outputs to obtain a power station-level power prediction value.

[0059] wherein the power station-level power prediction value is obtained by accumulating the power outputs of all components, realizing the power aggregation from the battery piece to the overall power station, first counting the total number of photovoltaic components in the power station, summing the component-level power output calculated in step S305 to obtain the total power of the power station, and then multiplying the prediction duration to obtain the power prediction value, and the formula of the power station-level power prediction value is as follows: wherein, is the power station-level power prediction value, is the total number of components in the power station, is the power output of the mth component, is the prediction duration.

[0060] In one possible implementation, the battery piece local shading factor is determined according to the photovoltaic component-level electrical data and the shaded battery piece positioning data, comprising: S3011, extracting reference irradiance and ground albedo data from the photovoltaic component-level electrical data.

[0061] S3012, determining an initial battery piece local shading factor according to the reference irradiance, ground albedo data and shaded battery piece positioning data.

[0062] wherein the initial battery piece local shading factor is used to preliminarily quantify the shading degree of a single battery piece, and the calculation can be combined with cross-validation of multiple source data, and the formula of the initial battery piece local shading factor is as follows: wherein, is the initial local shading factor of the nth battery piece, is the calculated irradiance of the nth battery piece, is the reference irradiance, is the ground reflection gain factor of the nth battery piece.

[0063] S3013, determining an update range of the initial cell local shading factor according to the shading object moving speed in the shading object data of the photovoltaic panel, to obtain a cell local shading factor.

[0064] wherein the update range of the initial factor can be determined The update range is related to the shading object moving speed in the shading object data of the photovoltaic panel obtained in step S1, and a speed grading rule can be used, when the moving speed V > 10 km / h, the update range is 0.3, when 5 km / h < V ≤ 10 km / h, the update range is 0.15; when 0 km / h < V ≤ 5 km / h, the update range is 0.05; and when V = 0, the update range is 0.

[0065] Further, the junction 14 combines the update range to calculate the final local shading factor, and the calculation formula of the cell local shading factor is as follows: wherein, is the final local shading factor of the nth cell, is the initial local shading factor, is the update range, is a speed direction coefficient, which can be 1 when the shading object moves towards the cell and -1 when the shading object moves away from the cell, and is determined by the moving direction of the shading object monitored by the millimeter wave radar.

[0066] In a possible implementation, the cell mismatch factor is determined according to the photovoltaic module level electrical data and the shaded cell positioning data, comprising: S3014, extracting the series connection mode data of the cells in the module from the photovoltaic module level electrical data.

[0067] S3015, determining the series connection branch number k in which the nth cell is located according to the series connection mode data.

[0068] S3016, determining the total number of cells, the number of shaded cells and the cell shading degree data of each shaded cell in the series connection branch with the series connection branch number k according to the shaded cell positioning data.

[0069] S3017, determining the branch level shading influence coefficient according to the total number of cells, the number of shaded cells and the cell shading degree data.

[0070] wherein the calculation formula of the branch level shading influence coefficient is as follows: wherein, is the shading influence coefficient of the kth series connection branch, The sum of the shading degrees of all shaded cell pieces in the branch, The total number of cell pieces in the branch.

[0071] S3018, determining an initial cell piece mismatch factor according to the branch-level shading influence coefficient.

[0072] The calculation formula of the initial cell piece mismatch factor is as follows: Wherein, The initial mismatch factor of the nth cell piece, The mismatch amplification coefficient is 2.5 for polycrystalline silicon cell pieces and 2.0 for monocrystalline silicon cell pieces, because the current consistency of polycrystalline silicon cell pieces is more sensitive to shading, the amplification coefficient is larger, The shading influence coefficient of the branch k where the nth cell piece is located.

[0073] S3019, correcting the cell piece mismatch factor according to the shading object type in the photovoltaic panel shading object data to obtain a cell piece mismatch factor.

[0074] Wherein, the shading object type in the photovoltaic panel shading object data obtained in step S1 can be introduced to make the factor match the mismatch characteristics of different shading scenarios. The calculation formula of the cell piece mismatch factor is as follows: Wherein, The final mismatch factor of the nth cell piece, The initial mismatch factor, The shading object type coefficient (set according to the shading object type: 0.8 for fixed shading objects such as buildings and long-term trees, because the shadow position of fixed shading is stable, the current fluctuation of cell pieces in the branch is small, and the mismatch influence is relatively weak; 1.2 for dynamic shading objects such as cumulus clouds and flying birds, because dynamic shading causes frequent changes in the current of cell pieces in the branch, which is prone to transient mismatch and needs to be strengthened) S4, evaluating the prediction result of the power station-level power prediction value, when the power station-level power prediction value prediction value meets the error threshold, outputting the power station-level power prediction value as the final power generation prediction result.

[0075] Specifically, the root mean square error and the mean absolute error can be used to evaluate the prediction results of the power station level power prediction value, the power station level power prediction value can be obtained synchronously, and then can be stored according to the “prediction time-prediction duration” classification; and the actual power generation data of the power station in the same period can be obtained, which can be directly extracted from the grid-connected inverter or the metering meter of the photovoltaic power station. When extracting, it is necessary to ensure that the “time dimension is completely matched with the prediction value. If the prediction value is the power of 1 hour period (such as 10:00-11:00), the actual power generation also needs to take the cumulative measurement value of the same period, and the abnormal value caused by the failure of the metering device is excluded to ensure the accuracy of the evaluation data.

[0076] The root mean square error (RMSE) calculation formula is as follows: Wherein, the root mean square error is RMSE, the actual power generation of the power station in the jth prediction period, the power station level power prediction value in the jth prediction period, the number of evaluation prediction samples (according to actual demand, such as 24 times per hour for a day, then N=24, or 168 times for a week, then N=168) The mean absolute error (MAE) calculation formula is as follows: Wherein, the mean absolute error is MAE, the absolute difference between the jth prediction period actual value and the prediction value, the actual power generation of the power station in the jth prediction period, the power station level power prediction value in the jth prediction period, the number of evaluation prediction samples.

[0077] Further, a double error threshold is set, if the root mean square error < 50W and the mean absolute error < 8%, the power station level power prediction value is output as the final power prediction result, if any threshold is not met (such as RMSE = 60W or MAE = 9%), the error feedback optimization mechanism is triggered, and the updating range of the local shading factor of the battery piece or the correction coefficient of the battery piece mismatch factor is returned to step S3 for readjustment, such as increasing the mismatch factor correction coefficient of the dynamic shading object to 1.3, or returning to step S2 to optimize the attention weight of the variational autoencoder, such as strengthening the feature extraction of the low gray value area, re-executing the calculation of steps S2-S3, and then entering step S4 for evaluation again, until the error meets the threshold requirement; if the error still does not meet the threshold after 3 times of optimization, a warning signal is triggered to prompt the operation and maintenance personnel to check whether the shading identification sensor (such as a camera, a millimeter wave radar) or the metering equipment is abnormal to ensure the stable operation of the system.

[0078] The present example provides a power generation prediction method based on a photovoltaic power generation system. First, multi-source data of the photovoltaic power generation system including photovoltaic module level electrical data, photovoltaic panel terrain data, and photovoltaic panel shading data is obtained. Then, local shading of the photovoltaic module is identified according to the multi-source data to determine a corrected gray map and shading battery piece positioning data. Subsequently, the power generation of the locally shaded battery is calculated and a power station level power prediction value is obtained by combining the shading battery piece positioning data and the multi-source data of the photovoltaic power generation system. Finally, error evaluation is performed on the power station level power prediction value, and the final power prediction result is output when the error threshold is met. The introduction of photovoltaic panel terrain data and photovoltaic panel shading data, combined with local shading identification to obtain accurate positioning information of the shaded battery piece, can avoid the overestimation or underestimation of output caused by simplification or neglect of local shading in existing methods. At the same time, it can capture the local irradiation differences of photovoltaic panels under different terrain conditions, reduce the prediction deviation caused by the inability of global terrain data to reflect micro differences, and meet the demand of grid dispatching for high-precision prediction, thereby improving new energy consumption rate and ensuring grid operation stability.

[0079] Consistent with the above, please refer to Figure 2 , Figure 2 A structure diagram of a power generation prediction device of a photovoltaic power generation system is provided for the embodiments of the present application. As shown in Figure 2 , the device comprises: A first acquisition unit 1 is configured to acquire multi-source data of a photovoltaic power generation system, wherein the multi-source data of the photovoltaic power generation system includes photovoltaic module level electrical data, photovoltaic panel terrain data, and photovoltaic panel shading data. A first processing unit 2 is configured to identify local shading of a photovoltaic module according to the multi-source data of the photovoltaic power generation system, and determine a corrected gray map and shading battery piece positioning data. The second processing unit 3 is configured to determine local shading battery power generation according to the shaded battery positioning data and the photovoltaic power generation system multi-source data, and obtain a power station level power prediction value. The third processing unit 4 is configured to perform prediction result evaluation on the power station level power prediction value, and output the power station level power prediction value as a final power prediction result when the power station level power prediction value meets an error threshold.

[0080] In a possible implementation, in the aspect of identifying local shading of the photovoltaic module according to the photovoltaic power generation system multi-source data and determining the corrected gray map, the first processing unit 2 is configured to: extract photovoltaic module cell curve data and cell irradiance proportion data from the photovoltaic module level electrical data; determine a micro-terrain feature comprehensive value of each cell according to the photovoltaic panel terrain data and the photovoltaic module cell curve data; determine a micro-terrain correction coefficient according to a shading object type in the photovoltaic panel shading object data; determine a micro-terrain corrected cell gray value according to the cell irradiance proportion data, the micro-terrain feature comprehensive value, and the micro-terrain correction coefficient; determine a gray map update period according to a shading object moving speed in the photovoltaic panel shading object data; determine a corrected gray map by combining the micro-terrain corrected cell gray value and the update period with the cell positioning.

[0081] In a possible implementation, in the aspect of identifying local shading of the photovoltaic module according to the photovoltaic power generation system multi-source data and determining the shaded battery positioning data, the first processing unit 2 is configured to: train a variational autoencoder neural network with a fusion attention mechanism by taking the photovoltaic module cell curve data as a training sample and taking the corrected gray map as a training label; optimize an attention weight of the variational autoencoder neural network according to a shading object type in the photovoltaic panel shading object data; input real-time collected photovoltaic module cell curve data into the variational autoencoder neural network, and output a real-time corrected gray map; determine the shaded battery positioning data according to the real-time corrected gray map.

[0082] In a possible implementation, in the aspect of determining local shading battery power generation according to the shaded battery positioning data and the photovoltaic power generation system multi-source data, and obtaining a power station level power prediction value, the second processing unit 3 is configured to: determining a local shading factor of the cell and a cell mismatch factor according to the photovoltaic module level electrical data and the shaded cell location data; determining an actual irradiance of each cell according to the local shading factor of the cell; extracting an effective area of the cell and a cell working efficiency from the photovoltaic module level electrical data; determining a single cell actual output according to the effective area of the cell, the cell working efficiency and the cell mismatch factor; summing the single cell actual output to obtain a module level output; summing the module level output to obtain a power station level power prediction value.

[0083] In a possible implementation, in the aspect of determining the local shading factor of the cell according to the photovoltaic module level electrical data and the shaded cell location data, the second processing unit 3 is configured to: extracting reference irradiance and ground albedo data from the photovoltaic module level electrical data; determining an initial local shading factor of the cell according to the reference irradiance, the ground albedo data and the shaded cell location data; determining an update range of the initial local shading factor of the cell according to a moving speed of the shading object in the photovoltaic panel shading object data, to obtain the local shading factor of the cell.

[0084] In a possible implementation, in the aspect of determining the cell mismatch factor according to the photovoltaic module level electrical data and the shaded cell location data, the second processing unit 3 is configured to: extracting series connection mode data of the cells in the module from the photovoltaic module level electrical data; determining a serial connection branch number k in which the nth cell is located according to the series connection mode data; determining a total number of cells, a number of shaded cells and a cell shading degree data of each shaded cell in the serial connection branch with the serial connection branch number k according to the shaded cell location data; determining a branch level shading influence coefficient according to the total number of cells, the number of shaded cells and the cell shading degree data; determining an initial cell mismatch factor according to the branch level shading influence coefficient; correcting the cell mismatch factor according to a shading object type in the photovoltaic panel shading object data, to obtain the cell mismatch factor.

[0085] In a possible implementation, in the aspect of performing prediction result evaluation on the power station level power prediction value, when the power station level power prediction value prediction value meets the error threshold, outputting the power station level power prediction value as the final power generation prediction result, the third processing unit 4 is configured to: performing prediction result evaluation on the power station level power prediction value by using root mean square error and mean absolute error; if the root mean square error is less than 50 W and the mean absolute error is less than 8%, outputting the power station level power prediction value as the final power generation prediction result.

[0086] Consistent with the above embodiment, please refer to Figure 3 , Figure 3 A terminal structure schematic diagram provided by the embodiment of the present application is shown in the figure, including a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory are connected with each other, wherein the memory is used for storing a computer program, the computer program includes program instructions, the processor is configured to call the program instructions, the above program includes instructions for performing the following steps; acquiring photovoltaic power generation system multi-source data, wherein the photovoltaic power generation system multi-source data includes photovoltaic component level electrical data, photovoltaic panel terrain data, photovoltaic panel shelter data; According to the photovoltaic power generation system multi-source data, the local shielding of the photovoltaic component is identified, and the corrected gray map and the shielding cell positioning data are determined; According to the shielding cell positioning data and the photovoltaic power generation system multi-source data, the local shielding cell power generation is determined, and the power station level power prediction value is obtained. performing prediction result evaluation on the power station level power prediction value, when the power station level power prediction value prediction value meets the error threshold, outputting the power station level power prediction value as the final power generation prediction result.

[0087] In the example, by first acquiring the multi-source data of the photovoltaic power generation system including the photovoltaic component level electrical data, the photovoltaic panel terrain data, and the photovoltaic panel shelter data, then performing local shelter identification on the photovoltaic components according to the multi-source data, determining the corrected gray map and the sheltered cell positioning data, subsequently combining the sheltered cell positioning data and the multi-source data of the photovoltaic power generation system, calculating the power generation of the locally sheltered cells and obtaining the power station level power prediction value, and finally performing error evaluation on the power station level power prediction value and outputting the final power generation prediction result when the error threshold is met, the local shelter identification is combined with the introduction of the photovoltaic panel terrain data and the photovoltaic panel shelter data to obtain the accurate positioning information of the sheltered cells, so as to avoid the overestimation or underestimation problem caused by simplification or neglect of local shelter in the existing method, and at the same time, the local irradiation difference of the photovoltaic panel under different terrain conditions can be captured, the prediction deviation caused by the fact that the global terrain data cannot reflect the micro differences is reduced, and the demand of the power grid dispatching for high-precision prediction is met, and the new energy consumption rate is improved and the power grid operation stability is ensured.

[0088] The above describes the scheme of the embodiments of the present application mainly from the perspective of the execution process of the method. It can be understood that the terminal includes hardware structures and / or software modules corresponding to the execution of each function in order to implement the above functions. Those skilled in the art should easily realize that, in combination with the unit and algorithm steps of each example described in the embodiments provided herein, the present application can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0089] The embodiments of the present application can divide the functional units of the terminal according to the above method examples, for example, each functional unit can be divided according to each function, or two or more functions can be integrated in one processing unit. The integrated unit can be realized in the form of hardware or software functional unit. It should be noted that the division of units in the embodiments of the present application is illustrative, and is only a logical function division. There can be another division method when actually implemented.

[0090] The embodiments of the present application also provide a computer storage medium, wherein the computer storage medium stores a computer program for electronic data exchange, and the computer program causes the computer to execute part or all steps of any one of the photovoltaic power generation system based power generation prediction methods described in the above method embodiments.

[0091] The embodiment of the present application further provides a computer program product, which comprises a non-transitory computer-readable storage medium storing a computer program. The computer program causes a computer to execute some or all of the steps of any of the photovoltaic power generation system-based power generation amount prediction methods described in the above method embodiments.

[0092] It should be noted that, for the above-mentioned method embodiments, in order to simply describe, they are all described as a combination of a series of actions, but those skilled in the art should know that the present application is not limited to the action sequence described, because according to the present application, some steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0093] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0094] In several embodiments provided in the present application, it should be understood that the disclosed apparatus can be implemented by other means. For example, the apparatus embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, apparatus or unit, and can be electrical or other forms.

[0095] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0096] In addition, each functional unit in each embodiment of the application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software program module.

[0097] If the integrated unit is realized in the form of a software program module and sold or used as an independent product, it can be stored in a computer readable memory. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a memory and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned memory includes: a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0098] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer readable memory, which can include a flash disk, a read-only memory, a random access memory, a magnetic disk or an optical disk, etc.

[0099] The embodiments of the present application are described in detail above, and the principles and implementation manners of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea; meanwhile, for those of ordinary skill in the art, according to the idea of the present application, the specific implementation manners and application ranges will be changed, and the above description of the embodiments should not be understood as a limitation of the present application.

Claims

1. A method for predicting power generation based on a photovoltaic power generation system, characterized in that, include: Acquire multi-source data of the photovoltaic power generation system, wherein the multi-source data of the photovoltaic power generation system includes photovoltaic module-level electrical data, photovoltaic panel topographic data, and photovoltaic panel shading data; Based on the multi-source data of the photovoltaic power generation system, the photovoltaic modules are partially shaded to identify the corrected grayscale image and the location data of the shaded cells. Based on the location data of the shaded solar cells and the multi-source data of the photovoltaic power generation system, the power generation of the partially shaded cells is determined, and the power station-level power prediction value is obtained. The predicted power generation value at the power plant level is evaluated. If the predicted power generation value at the power plant level meets the error threshold, the predicted power generation value at the power plant level is output as the final power generation prediction result.

2. The power generation prediction method based on a photovoltaic power generation system according to claim 1, characterized in that, The step of identifying partial shading of photovoltaic modules based on multi-source data from the photovoltaic power generation system and determining the corrected grayscale image includes: Extract photovoltaic module cell curve data and cell irradiance ratio data from the photovoltaic module-level electrical data; Based on the photovoltaic panel topography data and photovoltaic module cell curve data, determine the comprehensive value of the micro-topography features for each cell; The micro-topography correction coefficient is determined based on the type of shading object in the photovoltaic panel shading object data; The grayscale value of the micro-topography correction cell is determined based on the battery irradiance ratio data, the comprehensive value of micro-topography features, and the micro-topography correction coefficient. The grayscale image update cycle is determined based on the moving speed of the obstruction in the photovoltaic panel obstruction data; The grayscale value of the battery cell is corrected based on the micro-topography and the update cycle, and the corrected grayscale image is determined in combination with the location of the battery cell.

3. The power generation prediction method based on a photovoltaic power generation system according to claim 2, characterized in that, The step of identifying partial shading of photovoltaic modules based on multi-source data of the photovoltaic power generation system and determining the location data of the shaded cells includes: The photovoltaic module cell curve data is used as training samples, and the corrected grayscale image is used as training labels to train a variational autoencoder neural network with an attention mechanism. Based on the type of shading object in the photovoltaic panel shading object data, optimize the attention weights of the variational autoencoder neural network; The real-time collected photovoltaic module cell curve data is input into the variational autoencoder neural network, and the output is a real-time corrected grayscale image. Based on the real-time corrected grayscale image, the location data of the obscured battery cell is determined.

4. The power generation prediction method based on a photovoltaic power generation system according to claim 1, characterized in that, Based on the location data of the shaded solar cells and multi-source data of the photovoltaic power generation system, the power generation of the partially shaded cells is determined, and the power station-level power prediction value is obtained, including: Based on the photovoltaic module-level electrical data and the location data of the shaded solar cells, the local shading factor and the mismatch factor of the solar cells are determined. The actual irradiance of each solar cell is determined based on the local shading factor of the solar cell. The effective area and working efficiency of the solar cells are extracted from the photovoltaic module-level electrical data. The actual output of a single solar cell is determined based on the effective area of ​​the solar cell, the working efficiency of the solar cell, and the mismatch factor of the solar cell. The module-level output is obtained by summing the actual output of each individual battery cell. The summation of the component-level outputs yields the predicted power generation value at the power station level.

5. The power generation prediction method based on a photovoltaic power generation system according to claim 4, characterized in that, The step of determining the local shading factor of the solar cell based on the photovoltaic module-level electrical data and the location data of the shaded solar cell includes: Extract baseline irradiance and surface albedo data from the photovoltaic module-level electrical data; Based on the reference irradiance, surface albedo data, and the location data of the obscured solar cells, the initial local shading factor of the solar cells is determined. Based on the moving speed of the obstruction in the photovoltaic panel obstruction data, the update range of the initial cell local shading factor is determined, and the cell local shading factor is obtained.

6. The power generation prediction method based on a photovoltaic power generation system according to claim 4, characterized in that, The step of determining the cell mismatch factor based on the photovoltaic module-level electrical data and the location data of the shaded cells includes: Extract the series connection configuration data of the solar cells within the photovoltaic module from the photovoltaic module-level electrical data; Based on the series connection data, determine the series branch number k where the nth battery cell is located; Based on the location data of the obscured battery cells, determine the total number of battery cells, the number of obscured battery cells, and the degree of obstruction of each obscured battery cell in the series branch number k. Based on the total number of solar cells, the number of solar cells that are blocked, and the degree of solar cell blocking, determine the branch-level blocking impact coefficient; The initial cell mismatch factor is determined based on the branch-level shading influence coefficient. The cell mismatch factor is obtained by correcting the cell mismatch factor based on the type of shading object in the photovoltaic panel shading object data.

7. The power generation prediction method based on a photovoltaic power generation system according to claim 1, characterized in that, The evaluation of the predicted power generation value at the power plant level, wherein if the predicted power generation value at the power plant level meets the error threshold, the predicted power generation value at the power plant level is output as the final power generation prediction result, includes: The root mean square error and mean absolute error are used to evaluate the prediction results of the power station-level power prediction values. If the root mean square error is less than 50W and the mean absolute error is less than 8%, then the power plant-level power prediction value is output as the final power generation prediction result.

8. A power generation prediction device for a photovoltaic power generation system, characterized in that, include: The first acquisition unit is used to acquire multi-source data of the photovoltaic power generation system, wherein the multi-source data of the photovoltaic power generation system includes photovoltaic module-level electrical data, photovoltaic panel topographic data, and photovoltaic panel shading data; The first processing unit is used to identify partial shading of photovoltaic modules based on multi-source data of the photovoltaic power generation system, and to determine the corrected grayscale image and the location data of the shaded battery cells. The second processing unit is used to determine the power generation of the partially shaded battery based on the location data of the shaded battery cell and the multi-source data of the photovoltaic power generation system, and to obtain the power station-level power prediction value. The third processing unit is used to evaluate the prediction results of the power plant-level power prediction value. When the predicted value of the power plant-level power prediction value meets the error threshold, the power plant-level power prediction value is output as the final power generation prediction result.

9. A terminal, characterized in that, The system includes a processor, an input device, an output device, and a memory, which are interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute the power generation prediction method based on a photovoltaic power generation system as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the power generation prediction method based on a photovoltaic power generation system as described in any one of claims 1-7.

Images