Generated power prediction method based on dust deposition state of photovoltaic panel

Abstract

The invention relates to the technical field of photovoltaic panels, and discloses a photovoltaic panel dust deposition state-based power generation power prediction method, which comprises the steps of obtaining an infrared radiation image and spectral data of a photovoltaic panel, and performing primary processing on the infrared radiation image to obtain a calibration temperature matrix and a pre-processing image set; after the visible light or the ultraviolet light is converted to generate electric energy, actually measured power generation efficiency is collected for error calculation, a dynamic calibration process is triggered based on errors, and a calibrated power generation power prediction value is obtained; according to the invention, through coupling radiation attenuation, temperature loss, wind speed loss and other multi-physical field effects, the infrared spectroscopic analysis is utilized to identify dust components, so that the accuracy and operability of a physical mechanism are enhanced; besides, in combination with multi-source data such as infrared data, meteorological data and spectrum data, extreme conditions such as high humidity, strong wind speed and multi-type dust deposition can be effectively handled, the risk of single data source deviation is reduced, and high robustness and generalization ability are achieved.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic panel technology, and more specifically, to a method for predicting power generation based on the dust accumulation status of photovoltaic panels. Background Technology

[0002] A photovoltaic (PV) panel is a device that converts solar energy into electrical energy using the photoelectric effect. It is mainly composed of multiple photovoltaic cell units, typically using silicon as the semiconductor material. Silicon absorbs infrared, visible, or ultraviolet light from solar radiation. When these rays shine on the surface of the PV panel, electrons in the photovoltaic cells are excited, generating an electric current, thus converting solar radiation energy into direct current. The working principle of a PV panel depends on the response of the semiconductor material to different wavelengths of light. Infrared light mainly contributes to the thermal effect, while visible and ultraviolet light directly participate in the conversion of electrical energy.

[0003] Existing prediction methods rely on machine learning models trained on historical data or fixed-parameter physical models, which have certain shortcomings. First, the models have poor interpretability, often employing black-box models that lack in-depth explanations of the physical mechanisms and make it difficult to understand their internal working principles. Second, their error control capabilities are weak, especially when there are significant changes in the dust accumulation state, resulting in large fluctuations in prediction errors. Furthermore, they lack dynamic calibration mechanisms, making it impossible to adjust and adapt to changes in the actual environment in a timely manner. Finally, the coupling degree of the physical mechanisms is weak, failing to fully consider the coupling effects of multiple physical fields of photovoltaic panels, such as the interaction between dust accumulation and factors like climate change and light intensity, which limits the accuracy of the model's predictions. Summary of the Invention

[0004] To overcome the above-mentioned deficiencies of the prior art, embodiments of the present invention provide a method for predicting power generation based on the dust accumulation status of photovoltaic panels.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for predicting power generation based on the dust accumulation status of photovoltaic panels, the method comprising: Infrared radiation images and spectral data of photovoltaic panels are acquired, and the infrared radiation images are preliminarily processed to obtain a calibration temperature matrix and a preprocessed image set. The reference temperature is obtained, and the dust accumulation area of ​​the photovoltaic panel is calculated by combining the calibration temperature matrix and the proportion of dust accumulation pixels is obtained. The thickness of the dust accumulation area is inverted to obtain the dust accumulation thickness. The dust composition is identified and correlated with the power generation loss coefficient. The comprehensive score of the dust accumulation area is calculated by the dynamic weighted average method. Calculate the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area, and derive the total power generation efficiency and the total efficiency distribution map based on these efficiency values. After visible light or ultraviolet light is converted into electrical energy, the measured power generation efficiency is collected to calculate the error. Based on the error, a dynamic calibration process is triggered to obtain the calibrated predicted power generation value.

[0006] Furthermore, the step of calculating the dust accumulation area of ​​the photovoltaic panel and obtaining the percentage of dust-accumulated pixels by combining the calibration temperature matrix includes: Set the temperature threshold to T, traverse each pixel in the calibration temperature matrix, and calculate the deviation between its temperature and the reference temperature. If the deviation is ≥ T, it is marked as a potential dust accumulation area, a binary preliminary mask is generated, the edge detection algorithm is used to extract the edge of the preliminary mask to obtain the continuous dust accumulation area boundary, and then the boundary is filled by morphological closing operation to form a complete dust accumulation area. The number of pixels within the dust accumulation area is counted to obtain the dust accumulation pixel ratio.

[0007] Furthermore, the thickness inversion of the ash accumulation area is performed using the following formula: ; In the formula: As the reference temperature, For ambient temperature, Here, k represents the pixel temperature in the dust accumulation area, and k is the thermal conductivity of the dust. It is an infrared wavelength.

[0008] Furthermore, the identification of dust components and their correlation with the power generation loss coefficient includes: The infrared radiation image is aligned with the spectral data, and Fourier transform is performed on the aligned spectral data to denoise and extract the band feature peaks. The key peak positions are identified, and the peak height and half-peak width are marked to obtain a spectral feature map. The spectral feature map is then matched with a standard spectral library to identify the dust components. The corresponding power generation loss coefficient is extracted from the loss coefficient database based on the dust composition.

[0009] Furthermore, the specific process for calculating the comprehensive score of the ash accumulation area using the dynamic weighted average method is as follows: The power generation loss coefficient corresponding to the dust component is assigned a value to obtain the final loss coefficient. The dust accumulation thickness, dust accumulation pixel ratio and the final loss coefficient are then standardized. By dynamically assigning standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, a comprehensive score for the dust accumulation area can be obtained.

[0010] Furthermore, the dynamic assignment of standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient includes: Collect real-time environmental data, classify different environmental patterns using the K-means clustering algorithm, and obtain different optimal weight combinations based on different environmental patterns; Furthermore, it is fine-tuned by combining standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, with the following formula: ; In the formula: This represents the value of the i-th weight component at time t. Indicates the adaptive learning rate. This indicates that the mean absolute error (MAE) of power generation is relative to the weight. The partial derivatives, This is the physical constraint factor.

[0011] Furthermore, the specific steps for calculating the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area are as follows: Apply the infrared radiation attenuation formula to each pixel within each dust accumulation area to obtain the radiation attenuation value; The average temperature is extracted from the calibration temperature matrix, and the temperature efficiency is calculated based on the temperature loss coefficient. Wind speed efficiency is derived by combining real-time wind speed with wind speed loss coefficient.

[0012] Furthermore, the error-triggered dynamic calibration process includes: Calculate the absolute error and relative error, set an error threshold of n, and compare it with the relative error. If the relative error > n, then trigger the dynamic standard process. The dynamic standard procedure includes recalibrating the loss factor and updating the STC parameters.

[0013] An electronic device includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the above-described method for predicting power generation based on the dust accumulation status of a photovoltaic panel.

[0014] A computer-readable storage medium storing a computer program, which, when executed, implements the above-described method for predicting power generation based on the dust accumulation state of a photovoltaic panel.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This application discloses a method for predicting power generation based on the dust accumulation status of photovoltaic panels. The method includes: obtaining a reference temperature; calculating the dust accumulation area of ​​the photovoltaic panel and obtaining the percentage of dust-accumulated pixels using a calibration temperature matrix; performing thickness inversion on the dust accumulation area to obtain the dust accumulation thickness; identifying dust components and correlating them with the power generation loss coefficient; calculating a comprehensive score for the dust accumulation area using a dynamic weighted average method; calculating the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area; and deriving the total power generation efficiency and a total efficiency distribution map based on these efficiency values. The prediction method provided by this invention overcomes the limitations of traditional machine learning black-box models, directly deriving power generation using infrared thermal imaging and physical formulas. It possesses high interpretability and traceability, meeting the needs of industrial applications. By dynamically adjusting the weighting of dust accumulation scores based on real-time data and combining it with a closed-loop verification mechanism to continuously calibrate parameters, this method effectively achieves high-precision prediction with errors kept at a low level. It enhances the accuracy and operability of the physical mechanism by coupling multi-physical field effects such as radiation attenuation, temperature loss, and wind speed loss, while simultaneously utilizing infrared spectroscopy to identify dust components. Furthermore, by combining multi-source data such as infrared, meteorological, and spectral data, this method exhibits high adaptability in complex environments, effectively coping with extreme conditions such as high humidity, strong wind speeds, and various types of dust accumulation, reducing the risk of bias from a single data source, and demonstrating strong robustness and generalization ability. Attached Figure Description

[0016] Figure 1 A flowchart of a method for predicting power generation based on the dust accumulation status of photovoltaic panels provided by the present invention; Figure 2 A schematic diagram of the structure of an electronic device provided by the present invention; Figure 3 A schematic diagram of the structure of a computer-readable storage medium provided by the present invention; Figure 4 This invention provides a flowchart for calculating the comprehensive score of ash accumulation areas. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Please see Figure 1 As shown in the figure, this embodiment discloses a method for predicting power generation based on the dust accumulation status of photovoltaic panels. The method includes: S101: Acquire infrared radiation images and spectral data of photovoltaic panels, perform preliminary processing on infrared radiation images to obtain calibration temperature matrix and preprocessed image set; In this embodiment, the infrared radiation image is automatically triggered to be collected every 15 minutes by programming the thermal imager SDK. The installation position of the thermal imager is determined according to the power plant layout diagram (it needs to cover all photovoltaic panels, with a height of ≥3m and a downward angle of ≤45° to avoid shading), and the equipment coordinates are calibrated to align with the power plant geographic information system.

[0019] Meanwhile, a 5×5 median filter is used to perform sliding window processing on the infrared radiation image to eliminate random noise (such as thermal imager electronic noise) and preserve the edge details of the temperature distribution, resulting in a denoised image.

[0020] To facilitate calculations, pixel grayscale values ​​are converted to actual temperature values ​​based on the thermal imager calibration parameters. The conversion formula is as follows: ; In the formula: For pixel temperature, The radiation constant, Indicates infrared radiation flux. Atmospheric transmittance, Here, represents the emissivity, and is the device calibration value, which is taken as 0.95 in this embodiment. denoted as the detector response coefficient.

[0021] After calculating the pixel temperature, pixels with temperature values ​​exceeding a reasonable range are removed, such as abnormally high or low temperatures caused by equipment failure. In this embodiment, the range is ambient temperature ±50℃.

[0022] S102: As Figure 4 As shown, the reference temperature is obtained, and the dust accumulation area of ​​the photovoltaic panel is calculated by combining the calibration temperature matrix and the proportion of dust accumulation pixels is obtained. The thickness of the dust accumulation area is inverted to obtain the dust accumulation thickness. The dust composition is identified and correlated with the power generation loss coefficient. The comprehensive score of the dust accumulation area is calculated by the dynamic weighted average method. Specifically, the step of calculating the dust accumulation area of ​​the photovoltaic panel and obtaining the percentage of dust-accumulated pixels by combining the calibration temperature matrix includes: Set the temperature threshold to T, iterate through each pixel in the calibration temperature matrix, and calculate the deviation between its temperature and the reference temperature, i.e.: ; If the deviation is ≥ T, it is marked as a potential dust accumulation area, and a binary preliminary mask is generated, which is (0 = normal, 1 = potential dust accumulation). The edge detection algorithm is used to extract the edges of the preliminary mask to obtain the continuous dust accumulation area boundary. Then, the boundary is filled by morphological closing operation to form a complete dust accumulation area. The number of pixels within the dust accumulation area is counted to obtain the dust accumulation pixel ratio.

[0023] As a specific example, the thickness inversion of the ash accumulation area is performed using the following formula: ; In the formula: As the reference temperature, For ambient temperature, Here, k represents the pixel temperature in the dust accumulation area, and k is the thermal conductivity of the dust. It is an infrared wavelength.

[0024] It should be noted that the thermal conductivity k of dust refers to the amount of heat conducted per unit temperature gradient during heat conduction. It is a crucial physical parameter characterizing the thermal conductivity of dust materials, directly affecting the temperature distribution and heat transfer efficiency of dust under thermal radiation, temperature differences, and changes in ambient temperature. In laboratory calibration, a standard thermal conductivity experimental setup is typically established. A known heat source temperature difference is applied to a dust sample of a specific thickness, and the temperature difference between the two ends and the heat flux density through the dust layer are measured to calculate the thermal conductivity k. During the experiment, variables such as the thickness, particle size distribution, and density of the dust sample must be strictly controlled to ensure the accuracy of the results. The thermal conductivity of dust is affected by various factors, such as particle size, moisture content, density, composition, and temperature. The value of k may vary under different experimental conditions. A series of calibration experiments are usually conducted to determine the corresponding thermal conductivity range at different temperatures and for different dust types. For example, looser dust or dust with higher moisture content may have lower thermal conductivity, while dense, dry dust may have higher thermal conductivity. Therefore, the specific value of the thermal conductivity of dust needs to be determined by a comprehensive analysis that takes into account the laboratory environmental conditions, the physical properties of the dust sample, and the precision of the experimental setup.

[0025] Meanwhile, the infrared wavelength λ is 8-14μm in the thermal imager band, and in this embodiment, the center wavelength is 11μm.

[0026] For example, identifying dust components and associating them with the power generation loss coefficient includes: The infrared radiation image and spectral data are aligned, and Fourier transform denoising is performed on the aligned spectral data to extract the 600-1200 MHz spectrum. Band characteristic peaks; The key peak positions are identified, and the peak height and half-peak width are marked to obtain a spectral feature map. The spectral feature map is then matched with a standard spectral library to identify the dust components. It should be noted that a standard spectral library refers to a dataset containing characteristic spectra of known substances. These datasets are obtained through experiments or theoretical calculations and cover the absorption, reflection, or transmission spectra of various substances at different wavelengths (or wavenumbers). Each substance has unique spectral characteristic peaks within a specific wavelength range, reflecting the characteristics of its molecular structure or chemical composition. For example, substances such as silicates, carbonates, and organic compounds have characteristic peaks at 1000 nm in their infrared spectra. 850 1450 The isowavenumber position exhibits a specific absorption peak.

[0027] Based on the dust composition, the corresponding power generation loss coefficients are extracted from the loss coefficient database. It should be noted that the specific dust composition and corresponding loss coefficients are shown in Table 1. Table 1: Dust Mapping Table The coefficients for silicates, carbonates, and organic matter are derived from the "Silicate Handbook" and the photovoltaic industry loss coefficient database; the coefficients for sulfates and metal oxides are referenced from infrared spectroscopy studies and X-ray fluorescence spectroscopy analysis of mine dust. The values ​​for mixtures are determined using a weighted average method, ensuring that the spectral characteristics of each component are distinguishable (e.g., silicate 1000). 850 carbonates To avoid calculation errors caused by overlapping characteristic peaks, in actual use, the loss coefficient needs to be updated by fitting historical data with real-time spectral analysis (such as a portable infrared spectrometer). For example, the silicate coefficient may be adjusted to 0.94-0.96 due to changes in ambient humidity.

[0028] Meanwhile, the specific process of calculating the comprehensive score of the ash accumulation area using the dynamic weighted average method is as follows: The power generation loss coefficient corresponding to the dust component is assigned a value to obtain the final loss coefficient. The dust accumulation thickness, dust accumulation pixel ratio and the final loss coefficient are then standardized. By dynamically assigning standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, a comprehensive score for the dust accumulation area can be obtained.

[0029] Furthermore, the dynamic assignment of standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient includes: Collect real-time environmental data, classify different environmental patterns using the K-means clustering algorithm, and obtain different optimal weight combinations based on different environmental patterns; Furthermore, it is fine-tuned by combining standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, with the following formula: ; In the formula: This represents the value of the i-th weight component at time t. Indicates the adaptive learning rate. This indicates that the mean absolute error (MAE) of power generation is relative to the weight. The partial derivatives, This is the physical constraint factor.

[0030] It should be further explained that, The adaptive learning rate is dynamically adjusted over time, and its formula is: ; In the formula: The initial learning rate, This represents the historical maximum thickness of the accumulated dust. The minimum learning rate is set to ensure that the learning rate adaptively increases when the ash accumulation parameter changes drastically.

[0031] and As a physical constraint factor, its coupled ash accumulation thickness deviation Deviation from ambient temperature The formula is as follows: ; In the formula: , The physical coupling coefficients are set to values ​​of 0.2 and 0.1, respectively. This represents the historical average thickness of the dust accumulation. This represents the historical average ambient temperature.

[0032] When the dust accumulation thickness deviation Δδ increases, the learning rate Increase the learning rate to accelerate weight adjustments and address significant changes in dust accumulation; as the bias decreases, the learning rate approaches a certain value. To avoid over-adjustment, physical constraint factors: through By linearly coupling the ash thickness deviation and the ambient temperature deviation into the gradient update, the weight adjustment not only depends on the power generation error, but also reflects the combined influence of the ash physical state and environmental conditions, thus enhancing interpretability. Furthermore, by calculating the gradient based on the real-time power generation error, the weight adjustment is ensured to directly improve the accuracy of power generation prediction, forming a closed-loop optimization of "error-gradient-weight".

[0033] As a specific example, assume the temperature is 32℃ (historical average 30℃, deviation +2℃). Wind speed: 3.5 m / s (historical average 2.5 m / s, deviation +1.0 m / s); Humidity: 45% (historical average 50%, deviation -5%) GHI: 850W / m² (timestamp alignment error ±1 second); Ash accumulation physical parameters: Dust accumulation thickness δ=115μm (historical average 100μm, deviation +15%, historical maximum 200μm). The percentage of pixels with accumulated dust is 15.6%. Loss factor = 0.92 (silicate-carbonate mixed type); By standardizing the parameters, we can obtain: Temperature normalization: (32−20) / (40−20)×100=60.0 minutes; Normalized wind speed: (3.5−0) / (5−0)×100=70.0 minutes; Humidity normalization: (45−30) / (70−30)×100=37.5 minutes; Using the K-means clustering algorithm, the input temperature, wind speed, and humidity vectors are used to determine the "high temperature, low humidity, and strong wind" mode. The resulting weight combination is: thickness 50%, pixel ratio 20%, and loss coefficient 30%, with fine-tuning performed if the deviation is greater than 15%.

[0034] Adaptive learning rate calculation: ; Physical constraint factor calculation: ; The three weights are as follows: ; ; ; The three weights were normalized, and the adjusted weight combination was 51% for thickness, 21% for pixel ratio, and 28% for loss coefficient. Thickness score: 115 / 200 × 100 = 57.5 points; Pixel ratio score: 15.6 points; Loss coefficient score: 0.92 × 100 = 92.0 points; Overall score = 57.5 0.51 + 15.6 0.21+92 0.28≈58.36.

[0035] S103: Calculate the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area, and derive the total power generation efficiency and the total efficiency distribution map based on these efficiency values.

[0036] It should be noted that the specific steps for calculating the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area are as follows: Applying the infrared radiation attenuation formula to each pixel within each dust accumulation area yields the radiation attenuation value, as expressed below: ; In the formula: This represents the transmitted radiation flux (W / m²). This represents the radiant flux when there is no dust accumulation. This is the dust attenuation coefficient. Indicates the thickness of the dust accumulation. This represents the power loss rate due to dust accumulation.

[0037] The average temperature is extracted from the calibration temperature matrix, and the temperature efficiency is calculated based on the temperature loss coefficient. The calculation formula is as follows: ; In the formula: This is the temperature loss coefficient. The average temperature. The standard test temperature is 25℃.

[0038] The wind speed efficiency is calculated based on real-time wind speed and the wind speed loss coefficient, and the formula is as follows: In the formula: denoted as the wind speed loss coefficient, and v is the real-time wind speed.

[0039] As one specific implementation method, its overall efficiency is: ; In the formula: This represents the final loss coefficient.

[0040] For example, the expression for predicting power generation is: ; In the formula: GHI represents the rated power of the photovoltaic panel and the total solar radiation. In this embodiment, the temperature coefficient is set to -0.45% / ℃.

[0041] As a specific example, the dust attenuation coefficient β = 0.13 / mm Temperature loss coefficient γ = 0.0006 / ℃). Wind speed loss coefficient μ = 0.025 / m / s; Temperature coefficient α = -0.0045 / ℃; Infrared radiation attenuation efficiency is: Transmitted radiation flux =850× = 837.39 W / m² =1-(850-837.39) / 850×100≈98.52%; Temperature efficiency is: =1-0.0006×(32-25)=99.58%; Wind speed efficiency is: =1 - 0.025 × 3.5 = 91.25%; Final loss factor: =0.92×100=92.00%; Total power generation efficiency = 98.52% × 99.58% × 91.25% × 92.00% ≈ 82.36%.

[0042] The total power generation can then be obtained: =82.36%×98×(850 / 1000)×[1+(-0.0045)×(32-25)]≈66.44kW.

[0043] S104: After visible light or ultraviolet light is converted into electrical energy, the measured power generation efficiency is collected to calculate the error. Based on the error, a dynamic calibration process is triggered to obtain the calibrated predicted power generation value.

[0044] Specifically, the error-triggered dynamic calibration process includes: Calculate the absolute error and relative error, set an error threshold of n, and compare it with the relative error. If the relative error > n, the dynamic standard process is triggered. In this embodiment, the value of n can be set according to the actual situation. This application does not limit this. The dynamic standard procedure includes recalibrating the loss factor and updating the STC parameters.

[0045] Furthermore, it should be added that recalibrating the loss coefficients includes: updating the temperature loss coefficient γ (e.g., adjusting it from 0.05 to 0.06) or the wind speed loss coefficient μ (e.g., adjusting it from 0.02 to 0.025) based on historical data fitting, or recalibrating the dust composition and updating the β value through real-time spectral analysis (calling data from a portable infrared spectrometer).

[0046] The STC parameter update is as follows: if panel aging causes... Attenuation is addressed by recalibrating the rated power using historical power data.

[0047] Example 2 Please see Figure 2As shown, this embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the power generation prediction method based on the dust accumulation state of the photovoltaic panel provided by the above methods.

[0048] Since the electronic device described in this embodiment is the electronic device used to implement the power generation prediction method based on the dust accumulation state of photovoltaic panels in the embodiments of this application, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the power generation prediction method based on the dust accumulation state of photovoltaic panels described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the power generation prediction method based on the dust accumulation state of photovoltaic panels in the embodiments of this application falls within the scope of protection of this application.

[0049] Example 3 Please see Figure 3 As shown, this embodiment discloses a computer-readable storage medium, including a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, it implements the power generation prediction method based on the dust accumulation state of the photovoltaic panel provided by the above methods.

[0050] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters, weights, and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0051] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired or wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0052] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0053] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0054] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0055] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0056] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0057] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0058] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting power generation based on the dust accumulation status of photovoltaic panels, characterized in that, The method includes: Infrared radiation images and spectral data of photovoltaic panels are acquired, and the infrared radiation images are preliminarily processed to obtain a calibration temperature matrix and a preprocessed image set. The reference temperature is obtained, and the dust accumulation area of ​​the photovoltaic panel is calculated by combining the calibration temperature matrix and the proportion of dust accumulation pixels is obtained. The thickness of the dust accumulation area is inverted to obtain the dust accumulation thickness. The dust composition is identified and correlated with the power generation loss coefficient. The comprehensive score of the dust accumulation area is calculated by the dynamic weighted average method. Calculate the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area, and derive the total power generation efficiency and the total efficiency distribution map based on these efficiency values. After visible light or ultraviolet light is converted into electrical energy, the measured power generation efficiency is collected to calculate the error. Based on the error, a dynamic calibration process is triggered to obtain the calibrated predicted power generation value.

2. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 1, characterized in that, The calculation of the dust accumulation area of ​​the photovoltaic panel and the percentage of dust accumulation pixels by combining the calibration temperature matrix includes: Set the temperature threshold to T, traverse each pixel in the calibration temperature matrix, and calculate the deviation between its temperature and the reference temperature. If the deviation is ≥ T, it is marked as a potential dust accumulation area, a binary preliminary mask is generated, the edge detection algorithm is used to extract the edge of the preliminary mask to obtain the continuous dust accumulation area boundary, and then the boundary is filled by morphological closing operation to form a complete dust accumulation area. The number of pixels within the dust accumulation area is counted to obtain the dust accumulation pixel ratio.

3. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 1, characterized in that, The thickness inversion of the ash accumulation area is performed using the following formula: ； In the formula: As the reference temperature, For ambient temperature, Here, k represents the pixel temperature in the dust accumulation area, and k is the thermal conductivity of the dust. It is an infrared wavelength.

4. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 3, characterized in that, The process of identifying dust components and correlating them with the power generation loss coefficient includes: The infrared radiation image is aligned with the spectral data, and Fourier transform is performed on the aligned spectral data to denoise and extract the band feature peaks. The key peak positions are identified, and the peak height and half-peak width are marked to obtain a spectral feature map. The spectral feature map is then matched with a standard spectral library to identify the dust components. The corresponding power generation loss coefficient is extracted from the loss coefficient database based on the dust composition.

5. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 4, characterized in that, The specific process for calculating the comprehensive score of the ash accumulation area using the dynamic weighted average method is as follows: The power generation loss coefficient corresponding to the dust component is assigned a value to obtain the final loss coefficient. The dust accumulation thickness, dust accumulation pixel ratio and the final loss coefficient are then standardized. By dynamically assigning standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, a comprehensive score for the dust accumulation area can be obtained.

6. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 5, characterized in that, The dynamic assignment of standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient includes: Collect real-time environmental data, classify different environmental patterns using the K-means clustering algorithm, and obtain different optimal weight combinations based on different environmental patterns; Furthermore, it is fine-tuned by combining standardized dust accumulation thickness, dust accumulation pixel ratio, and final loss coefficient, with the following formula: ； In the formula: This represents the value of the i-th weight component at time t. Indicates the adaptive learning rate. This indicates that the average absolute error of power generation is related to the weight. The partial derivatives, This is the physical constraint factor.

7. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 1, characterized in that, The specific steps for calculating the radiation attenuation value, temperature efficiency, and wind speed efficiency of each pixel in the dust accumulation area are as follows: Apply the infrared radiation attenuation formula to each pixel within each dust accumulation area to obtain the radiation attenuation value; The average temperature is extracted from the calibration temperature matrix, and the temperature efficiency is calculated based on the temperature loss coefficient. Wind speed efficiency is derived by combining real-time wind speed with wind speed loss coefficient.

8. The method for predicting power generation based on the dust accumulation status of photovoltaic panels according to claim 1, characterized in that, The error-triggered dynamic calibration process includes: Calculate the absolute error and relative error, set an error threshold of n, and compare it with the relative error. If the relative error > n, then trigger the dynamic standard process. The dynamic standard procedure includes recalibrating the loss factor and updating the STC parameters.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the power generation prediction method based on the dust accumulation state of the photovoltaic panel as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed, implements the power generation prediction method based on the dust accumulation state of a photovoltaic panel as described in any one of claims 1 to 8.

Images