Photovoltaic cell energy efficiency diagram detection method based on near-end gradient method

Abstract

The invention discloses a photovoltaic cell energy efficiency diagram detection method based on a near-end gradient method. The method comprises the following steps: determining a projection matrix according to the area of a to-be-tested area on the surface of a photovoltaic cell and the pixel size of a projector; after the projection matrix combination is projected, measuring vectors are collected; constructing a total variation disturbance compressed sensing model; representing a disturbance compressed sensing problem as a total least square problem; the solving difficulty is reduced by introducing convex relaxation; the iteration efficiency in the solving process is accelerated by introducing an alternating direction multiplier method; and processing the solved measurement vector to obtain an energy efficiency diagram of the to-be-tested area. According to the invention, the number of rows and columns of the projection matrix can be randomly adjusted according to the area of the to-be-tested region, and the disturbance compressed sensing model is solved by adopting the near-end gradient method, so that the interference of noise can be well overcome, and a more accurate and high-quality photovoltaic cell energy efficiency diagram can be obtained.

Description

Technical Field

[0001] The present invention belongs to the field of electrical measurement technology, and further relates to a photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method in the field of photovoltaic cell detection technology. The present invention can provide a photovoltaic cell energy efficiency detection result energy efficiency diagram reflecting the health status and photoelectric conversion efficiency of the photovoltaic cell by detecting defects of the photovoltaic cell. Background Art

[0002] The results of photovoltaic cell energy efficiency testing can directly reflect the health status and photoelectric conversion efficiency of photovoltaic cells. Photovoltaic cells are extremely fragile in actual production and use, and are particularly prone to defects such as fragmentation, hidden cracks, broken grids, flower pieces, scratches, etc., which will seriously affect the health status of photovoltaic cells and reduce the photoelectric conversion efficiency of photovoltaic cells. An accurate, practical and comprehensive photovoltaic cell defect detection method can provide valuable energy efficiency diagrams, which plays an important role in improving the factory yield and quality classification in industrial production. Because the collected photovoltaic cell output current is the sum of the output of the entire battery, it is difficult to determine the specific location of the defect on the battery, and it is also difficult to determine the output current value of each point on the battery. If the output current value of each point on the photovoltaic cell can be accurately obtained, the photoelectric conversion efficiency of the point can be obtained. The energy efficiency diagram of the photovoltaic cell can be obtained by calculation, which can reflect the specific location of the photovoltaic cell defect.

[0003] Demirci et al. disclosed an electroluminescence EL (Electroluminescence) detection method in their published document "Efficient deep feature extraction and classification for identifying defective photovoltaic module cells in Electroluminescence images." (2021. Expert Syst. Appl. 175, 114810.). This method mainly relies on high-resolution infrared cameras and image analysis software. By applying voltage to the photovoltaic material, using an infrared camera to take a near-infrared image of the photovoltaic material, and then using image analysis software to process and analyze the image, defects and problems inside the photovoltaic material can be found, potential safety hazards can be prevented, and photovoltaic materials can also be regularly inspected and maintained. However, the disadvantage of this method is that it is necessary to detect specific photovoltaic materials and requires infrared cameras and image analysis software, which will have a high construction cost and will misjudge quality defects that are not easy to find.

[0004] Xidian University disclosed a method for measuring the energy efficiency diagram of photovoltaic cells based on orthogonal modulation in its patent document "Measurement method of energy efficiency diagram of photovoltaic cells based on orthogonal modulation" (application number: 202110981312.5, application publication number: CN 113824400 A, application date: 2021.08.25). The specific implementation steps of this method include the following: Step 1, determine the number of projection pixels according to the range of the photovoltaic cell area to be measured and the size of the projection pixels, and determine the projection matrix based on the number of projection pixels; Step 2, construct a sub-projection matrix, and use the sub-projection matrix to control the structured light to illuminate the area to be measured to obtain the measurement value; Step 3, arrange the measurement values ​​in the column direction to form a measurement vector, and post-process the measurement vector to obtain the energy efficiency diagram of the area to be measured. The disadvantage of this method is that the projection matrix is ​​a Hadamard matrix, the dimension of which is affected by the exponent, and the size of the measurement area cannot be adjusted arbitrarily, which has limitations in practical applications. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and to provide a photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method, aiming to solve the high construction cost associated with the traditional EL detection method, the problem that the Hadamard matrix cannot arbitrarily adjust the size of the measurement area due to the influence of dimension, and the problem of being affected by ambient light noise in the detection environment.

[0006] The technical idea for achieving the purpose of the present invention is that the present invention uses the proximal gradient method based on total variation regularization to solve the constructed perturbation compressed sensing model. Because the adaptive rule is used to select the step size in the iterative solution process, the fluctuation of the gradient can be reduced, and the regularization part can help reduce overfitting and improve the generalization ability of the model, thereby improving the anti-noise performance. In this way, the problem of being affected by ambient light noise in the actual detection of the prior art is solved, and the present invention can efficiently realize the reconstruction of the gradient sparse image, thereby providing a more accurate and high-quality photovoltaic cell energy efficiency diagram. The present invention uses the alternating direction multiplier method to accelerate the solution of the perturbation compressed sensing model. The method splits the original problem into several simple sub-problems, solves the sub-problems in parallel, and can also adjust the convergence speed by adjusting the penalty coefficient. Therefore, the detection can be completed and the energy efficiency diagram can be restored without upgrading the hardware, thereby solving the problem of high construction cost of the prior art. The present invention can determine the dimension size of the generated projection matrix according to the size of the test area, thereby solving the problem that the Hadamard matrix cannot adjust the size of the measurement area at will due to the influence of the dimension.

[0007] The specific steps of the present invention include the following:

[0008] Step 1, determine and generate a projection matrix according to the area of ​​the photovoltaic cell to be tested, and generate a projection matrix combination (Φ+E), where Φ represents the matrix combination, and E represents the noise, and its value range is [-20,-30] dB;

[0009] Step 2, constructing a total variation perturbation compressed sensing model based on the generated measurement vector;

[0010] Step 3, using the proximal gradient method based on total variation regularization to solve the perturbation compressed sensing model;

[0011] Step 4: Introduce convex relaxation to reduce the difficulty of solving the model;

[0012] Step 5, using the alternating direction multiplier method to accelerate the solution of the perturbation compressed sensing model;

[0013] Step 6, determine whether the current iteration meets the convergence condition. If so, execute step 7 after obtaining an N*1 reconstruction vector, otherwise, execute step 4;

[0014] Step 7: Generate an energy efficiency diagram of the photovoltaic cell test area.

[0015] Compared with the prior art, the present invention has the following advantages:

[0016] First, the present invention can determine the dimension size of the generated projection matrix according to the size of the test area, thereby overcoming the problem that when the Hadamard matrix is ​​used as the projection matrix, the matrix cannot adjust the size of the measurement area at will due to the influence of the dimension. In actual photovoltaic cell detection applications, detection can be performed according to the size of the test area without being affected by the dimension of the projection matrix.

[0017] Second, since the present invention adopts the proximal gradient method based on total variation regularization to solve the constructed perturbation compressed sensing model, it overcomes the problem of being affected by ambient light noise in actual detection of the prior art, so that the present invention can show stronger anti-noise performance in actual testing and efficiently realize the reconstruction of gradient sparse images, thereby providing more accurate and high-quality photovoltaic cell energy efficiency diagrams.

[0018] Third, the present invention utilizes the alternating direction multiplier method to accelerate the solution of the disturbance compressed sensing model. It can complete the detection and solve the model to restore the energy efficiency diagram without upgrading the hardware, thus overcoming the problem of high construction cost of the existing technology. It can realize the energy efficiency diagram detection of photovoltaic cells at low cost in actual industrial detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flow chart of the present invention;

[0020] Figure 2is a schematic diagram of a photovoltaic cell test area according to an embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the present invention simplified into a 2*2 disturbance compression sensing mathematical model;

[0022] Figure 4 It is the energy efficiency diagram of the photovoltaic cell test area obtained by the actual test of the present invention;

[0023] Figure 5 is a simulation diagram of the present invention, wherein, Figure 5 (a) is a grayscale image of a photovoltaic cell simulated by the simulation experiment of the present invention, Figure 5 (b) is a comparison diagram of the images reconstructed by PCLBIC and CLBIC of the present invention, Figure 5 (c) is a comparison chart of PSNR and SSIM of the output images of PCLBIC and CLBIC of the present invention. DETAILED DESCRIPTION

[0024] The following is a detailed description of the embodiments of the present invention in conjunction with the accompanying drawings, and the described embodiments are only a part of the present invention, not all of them. It can be clearly seen from the accompanying drawings that the present invention has more obvious advantages for the actual measurement of photovoltaic cells.

[0025] Reference Figure 1 , further describing the implementation steps of the embodiment of the present invention.

[0026] Step 1: Determine the projection matrix according to the area to be measured.

[0027] The number of projection pixels is determined according to the area size of the photovoltaic cell to be tested and the size of the projector pixels, and the number of rows and columns m and n as well as the number N of the projection matrix are further constructed.

[0028] Reference Figure 2 The specific implementation steps of determining the projection matrix according to the area to be measured in the embodiment of the present invention are described as follows:

[0029] The embodiments of the present invention are implemented on a hardware platform: an AMD Ryzen 7 3750H processor, 32GB of memory, a monochrome red light digital industrial projector, and a software platform: MATLAB R2023a and a Windows 10 operating system.

[0030] The photovoltaic cells preset in the embodiment of the present invention include grid lines, buses, artificial defects, scratch defects, welding areas, and the photovoltaic cell test area marked in the red dotted box. The area of ​​the test area is 6mm*6mm, and the size of each projection pixel of the projector is 75μm*75μm. Based on this, the number of rows of the projection matrix is ​​determined to be: 6mm / 75μm=80, and the number of columns is: 6mm / 75μm=80, so the number of rows and columns of the projection matrix is ​​m and n, respectively 80 and 80, and the number N=m*n=6400.

[0031] Step 1.2, use MATLAB R2023a to generate 6400 random matrices with 80*80 rows and columns and elements of only 0 and 1, organize all the random matrices into a matrix combination Φ, and store them.

[0032] Step 1.3, add -30 dB noise E to each matrix in the generated matrix combination Φ to obtain the projection matrix combination (Φ+E), and store it.

[0033] Step 2: Collect measurement vectors.

[0034] In the embodiment of the present invention, a computer is used to control the projector, and the projection matrix combination (Φ+E) obtained in step 1 is projected by structured light, and is sequentially projected onto the area to be tested. The photocurrent values ​​projected sequentially 6400 times form a measurement vector y. The specific process is as follows:

[0035] Step 2.1, use a projector to project the stored projection matrix (Φ+E) onto the photovoltaic cell measurement area in turn.

[0036] Step 2.2: A photocurrent value is generated each time projection is performed, and the current value generated by each projection is recorded by ADC.

[0037] Step 2.3, record the photocurrent value generated by 6400 projections as a 6400*1 measurement vector y.

[0038] The projector is a monochrome red light digital industrial projector.

[0039] Step 3: Construct the total variation perturbation compressed sensing model as follows:

[0040] min||Dx||1 sty+e=(Φ+E)x

[0041] Among them, min represents the minimum operation, ||·||1 represents the L1 norm operation, D represents the total variation operator to be calculated using the number of rows and columns of the projection matrix, x represents the reconstruction vector to be calculated, and e represents the interference noise in the measurement vector.

[0042] Reference Figure 3, further describes the process of simplifying the perturbation compressed sensing mathematical model into 2*2.

[0043] The total variation operator D is calculated according to the number of rows and columns of the projection matrix obtained in step 1, and the perturbation compressed sensing model is constructed by combining the projection matrix group (Φ+E) and the measurement vector y obtained in step 2. Figure 3 The Φ in the example represents the randomly generated matrices Φ1, Φ2…Φ 6400 The random matrix is ​​composed of E, and the embodiment adds -30dB noise E1, E2...E 6400 , (Φ+E) represents the projection matrix finally obtained in the embodiment. The matrix is ​​projected onto the photovoltaic cell in the form of a light beam through a projector to obtain a measurement vector y with noise e. The perturbation compressed sensing model is solved to obtain the reconstructed vector x, and the photovoltaic cell energy efficiency diagram can be obtained.

[0044] Step 4: Formulate the perturbation compressed sensing problem as a total least squares problem

[0045] Since the least squares problem can provide an effective optimization framework and effectively improve the solution efficiency, the perturbation compressed sensing model in step 3 can be expressed as a total least squares problem, taking into account the projection matrix noise and the measurement vector noise. Then, the proximal gradient method based on total variation regularization is used to solve the following equation, which can effectively avoid the interference noise of the system.

[0046]

[0047] Wherein, λ represents a regularization parameter, which controls the strength of the regularization term. λ>0. A value that is too large may make the model too simple, affecting the model effect. In the embodiment of the present invention, λ is taken as 0.02, and ||·||2 is an L2 norm operation.

[0048] Step 5: According to the following formula, introduce convex relaxation to reduce the difficulty of the model:

[0049]

[0050] in, Indicates that x n Get the minimum value operation, d and p are auxiliary variables constructed by the alternating direction multiplication method for the convenience of subsequent steps. In step 6, there will be an expression, δ is the Lagrangian penalty term, δ>0, in the embodiment of the present invention, it is 0.111, x n+1 、x n Respectively represent the reconstruction vectors after the n+1th and nth iterations, μ n represents the iteration step size. The smaller the step size, the better the iterative optimization effect. In the embodiment of the present invention, it is set to 0.0000005. nrepresents the gradient of the total least squares problem after each iteration. In the embodiment of the present invention, each iteration needs to calculate g according to the total least squares problem after iterative update. n .

[0051] The difficulty of solving the model in step 4 is reduced by introducing a convex relaxation coefficient. Introducing convex relaxation can reduce the complexity of the problem, simplify the solution process, ensure the global optimal solution, and prepare for the introduction of the alternating direction multiplier method to speed up the iteration speed.

[0052] Step 6: Use the following alternating direction multiplier method to solve the perturbation compressed sensing model to speed up the iteration efficiency:

[0053]

[0054] p n+1 =p n +(Dx n+1 -d n+1 )

[0055] Among them, d n+1 and d n Respectively represent the intermediate variables d and p after the n+1th and nth iterations n+1 、p n They represent the intermediate variable p after the n+1th and nth iterations respectively.

[0056] The embodiment of the present invention introduces the alternating direction multiplier method to split the function of step 5 into three low-difficulty sub-functions, and efficiently implements the total variation regularized reconstruction algorithm by accelerating iteration.

[0057] The alternating direction multiplier method splits the problem in step 5 into three simple sub-problems, solves the sub-problems in parallel, and can also adjust the convergence speed by adjusting the penalty coefficient, greatly speeding up the iteration speed.

[0058] Step 7: Determine whether the current iteration meets the convergence condition. If so, get a 6400*1 reconstruction vector and execute step 8. Otherwise, continue to iterate from step 5.

[0059] The convergence conditions are as follows:

[0060]

[0061] The superscript T represents the transpose operation.

[0062] Step 8: Arrange the reconstruction vectors into a reconstruction matrix and represent it with pixel values ​​to obtain an energy efficiency diagram of the photovoltaic cell test area.

[0063] Use the imadjust function in MATLAB to arrange the reconstruction vectors into an 80*80 reconstruction matrix in a column-first manner, and use the imshow function in MATLAB to convert the reconstruction matrix into the final energy efficiency diagram of the photovoltaic cell test area, such as Figure 4 As shown in the figure, the grid lines, structural defects, scratch defects of the photovoltaic cell, and the defects of the photovoltaic cell itself marked with a red circle in the upper right corner (defects that are invisible to the naked eye) can be clearly observed.

[0064] The effect of the present invention is further described below in conjunction with simulation experiments:

[0065] 1. Simulation experiment conditions:

[0066] The simulation experiment hardware platform of the present invention has an AMD Ryzen 7 3750H processor and 32GB of memory.

[0067] The simulation experiment software platform of the present invention: MATLAB R2023a and Windows 10 operating system.

[0068] 2. Analysis of simulation content and results.

[0069] The image used in the simulation experiment of the present invention is a simulated photovoltaic cell grayscale image, such as Figure 5 (a) shows the image. The image includes simulated photovoltaic cell grid lines and simulated photovoltaic cell defects. The imaging time is July 2024, the image size is 64*64 pixels, and the image format is bmp.

[0070] The simulation experiment of the present invention adopts the method of the present invention, referred to as PCLBIC, and an existing technology (a traditional algorithm for solving the compressed sensing model CLBIC) to reconstruct the input simulated photovoltaic cell grayscale image to obtain a reconstructed image.

[0071] The prior art CLBIC refers to a compressed light beam induced current detection method disclosed by Quan Lei et al. in their paper “Compressive light beam induced current sensing for fast defect detection in photovoltaic cells.” (2017.Sol.Energy 150,345–352.), referred to as CLBIC.

[0072] The simulation experiment of the present invention adds three kinds of noise of -20dB, -25dB and -30dB to the grayscale image of the simulated photovoltaic cell, and then inputs the three images after adding noise into the PCLBIC model and CLBIC model of the present invention to reconstruct the image. The comparison of the reconstruction results is shown in the figure Figure 5 (b) as shown.

[0073] The effects of the present invention are further described below in conjunction with the accompanying drawings.

[0074] Figure 5 The first column in (b) shows the reconstructed images of the grayscale image of the simulated photovoltaic cell with -20 dB noise added, output by the conventional CLBIC and the PCLBIC of the present invention. Figure 5 It can be seen from the first column in (b) that the image output by the PCLBIC of the present invention can roughly distinguish the grid lines and defects, while the image output by the CLBIC is seriously polluted by noise and is so blurred that the grid lines and defects cannot be distinguished. Figure 5 The second column in (b) is respectively CLBIC and PCLBIC of the present invention, and the reconstructed image output of the simulated photovoltaic cell grayscale image with -25dB noise added is Figure 5 From the second column in (b), it can be seen that the image output by the PCLBIC of the present invention can distinguish grid lines and defects more clearly, while the image output by the CLBIC is very fuzzy and cannot accurately distinguish grid lines and defects. Figure 5 The third column in (b) is respectively CLBIC and PCLBIC of the present invention, and the reconstructed image output of the simulated photovoltaic cell grayscale image with -30dB noise added is Figure 5 From the third column in (b), it can be seen that the reconstructed image output by PCLBIC of the present invention is almost unaffected by noise, while the reconstructed image output by CLBIC still has a small amount of blurred boundaries.

[0075] In order to objectively evaluate the image quality, the present invention introduces the image's Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) as image quality evaluation indicators. The PSNR and SSIM values ​​of the output image can be directly calculated by the psnr and ssim functions in MATLAB.

[0076] Figure 5 (c) is a comparison chart of PSNR and SSIM of the reconstructed images output by PCLBIC and CLBIC of the present invention. Figure 5 The horizontal axis of (c) represents the noise of -20dB, -25dB, and -30dB added to the simulated photovoltaic cell grayscale image, the vertical axis on the left represents the PSNR value, and the vertical axis on the right represents the SSIM value. Figure 5The orange triangle curve in (c) is the PSNR value of the reconstructed image output by PCLBIC of the present invention under different noise conditions, and the orange circle curve is the PSNR value of the reconstructed image output by CLBIC under different noise conditions. The blue triangle curve is the SSIM value of the reconstructed image output by PCLBIC of the present invention under different noise conditions, and the blue circle curve is the SSIM value of the reconstructed image output by CLBIC under different noise conditions. By comparing PSNR and SSIM, it can be found that the PSNR and SSIM of PCLBIC of the present invention are better than those of the traditional compressed sensing model CLBIC under four noise conditions, among which the PSNR of PCLBIC of the present invention is 3.5dB higher than that of the traditional compressed sensing model CLBIC on average.

[0077] The simulation experiment of the present invention uses a projector to project the generated random matrix combination to the measurement area of ​​the photovoltaic cell, and the collected photocurrent value is the measurement vector. The measurement vector is processed by the proximal gradient method based on total variation regularization to the perturbation compressed sensing algorithm model. The proximal gradient method uses an adaptive rule to select the step size during the iterative solution process, which can reduce the fluctuation of the gradient, and the regularization part can help reduce overfitting and improve the generalization ability of the model, thereby improving the anti-noise performance. The present invention can efficiently realize the reconstruction of the gradient sparse image, thereby providing a more accurate and high-quality photovoltaic cell energy efficiency diagram.

[0078] The present invention uses the alternating direction multiplier method to accelerate the solution of the perturbation compressed sensing model. The method splits the original problem into several simple sub-problems, solves the sub-problems in parallel, and can also adjust the convergence speed by adjusting the penalty coefficient. Therefore, the present invention can complete the detection and restore the energy efficiency graph without upgrading the hardware.

[0079] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method, characterized in that: The generated projection matrix is ​​determined according to the size of the area to be tested, and the constructed perturbation compressed sensing model is solved by the proximal gradient method based on total variation regularization. The perturbation compressed sensing model is solved and accelerated by the alternating direction multiplier method. The detection steps of this method include the following: Step 1, determine and generate a projection matrix according to the area of ​​the photovoltaic cell to be tested, and generate a projection matrix combination (Φ+E), where Φ represents the matrix combination, and E represents the noise, and its value range is [-20,-30] dB; Step 2, constructing a total variation perturbation compressed sensing model based on the generated measurement vector; Step 3, using the proximal gradient method based on total variation regularization to solve the perturbation compressed sensing model; Step 4: Introduce convex relaxation to reduce the difficulty of solving the model; Step 5, using the alternating direction multiplier method to accelerate the solution of the perturbation compressed sensing model; Step 6, determine whether the current iteration meets the convergence condition. If so, execute step 7 after obtaining an N*1 reconstruction vector, otherwise, execute step 4; Step 7: Generate an energy efficiency diagram of the photovoltaic cell test area.

2. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 1 is characterized in that: Generating the projection matrix in step 1 means calculating the size of the projection matrix according to the length and width of the photovoltaic cell area to be tested and the size of each pixel of the projector, dividing the length of the area to be tested by the length of the pixel to obtain the number of columns n of the projection matrix, dividing the width of the area to be tested by the width of the pixel to obtain the number of rows m of the projection matrix, and the number of projection matrices is equal to the product of m and n.

3. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 1 is characterized in that: Generating the projection matrix combination in step 1 refers to generating N random matrices with m*n rows and columns and elements of only 0 and 1, and forming all the random matrices into a matrix combination Φ; adding noise to each matrix in the matrix combination Φ to obtain a projection matrix combination (Φ+E).

4. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 3 is characterized in that: The measurement vector in step 2 is to project the photovoltaic cell to be measured area in sequence using the projection matrix combination (Φ+E), record the photocurrent value generated by each projection, and form the measurement vector y with all the projected photocurrent values.

5. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 4 is characterized in that: The total variation perturbation compressed sensing model described in step 2 is as follows: min‖Dx‖1s.t.y+e=(Φ+E)x Among them, min represents the minimum operation, ‖·‖1 represents the L1 norm, D represents the total variation operator to be calculated using the number of rows and columns of the projection matrix, x represents the reconstruction vector to be calculated, and e represents the interference noise in the measurement vector.

6. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 5, characterized in that: The proximal gradient method based on total variation regularization described in step 3 is used to solve the perturbation compressed sensing model by the following formula: Among them, λ represents the regularization parameter, which controls the strength of the regularization term. λ>0. Too large a value may make the model too simple and affect the model effect. ‖·‖2 represents the L2 norm.

7. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 6, characterized in that: The introduction of convex relaxation in step 4 to reduce the difficulty of solving the model is obtained from the following formula: in, Indicates that x n The minimum operation is obtained, u represents the optimized reconstruction vector in the iterative process, d and p are auxiliary variables constructed by the alternating direction multiplication method for the convenience of subsequent steps, and there will be an expression in claim 8, δ represents the Lagrangian penalty term, δ>0, x n+1 、x n Respectively represent the reconstruction vectors after the n+1th and nth iterations, μ n It represents the iteration step length. The smaller the step length, the better the iterative optimization effect. n represents the gradient of the overall least squares problem after the nth iteration.

8. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 7 is characterized in that: The use of the alternating direction multiplier method in step 5 to accelerate the solution of the perturbation compressed sensing model is accomplished by the following formula: p n+1 =p n +(Dx n+1 -d n+1 ) Among them, d n+1 and d n Respectively represent the intermediate variables d and p after the n+1th and nth iterations n+1 、p n They represent the intermediate variable p after the n+1th and nth iterations respectively.

9. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 8, characterized in that: The convergence conditions described in step 6 are as follows: Among them, the superscript Represents a transpose operation.

10. The photovoltaic cell energy efficiency diagram detection method based on the proximal gradient method according to claim 1, characterized in that: Generating the energy efficiency diagram of the photovoltaic cell test area described in step 7 refers to using the imadjust function in MATLAB to arrange the reconstruction vector x into an m*n reconstruction matrix in a column-first manner, and using the imshow function in MATLAB to convert the reconstruction matrix into an energy efficiency diagram of the photovoltaic cell test area.

Images