An automatic evaluation system for performance loss due to defects in solar cells
By using electroluminescence imaging detection and image restoration technology, the performance loss caused by defects in solar cells can be automatically assessed, solving the problem of difficulty in quantifying and eliminating the impact of defects in existing technologies, and reducing the R&D cost of high-efficiency solar cells.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-20
- Publication Date
- 2026-03-20
AI Technical Summary
Existing technologies cannot effectively quantify and eliminate the impact of defects introduced during the manufacturing process of solar cells on their performance, which affects the reliability of performance comparisons between different samples and increases the R&D cost of high-efficiency solar cells.
Employing electroluminescence imaging and image restoration technologies, this system automatically assesses the performance loss caused by solar cell defects through cell edge detection, defect recovery, and current-voltage characteristic calculation. It includes an input module, a cell edge detection module, a cell defect recovery module, and a cell performance loss assessment module, quantifying the impact of defects on performance.
This enables quantitative assessment of solar cell defects, reduces the R&D cost of high-efficiency solar cells, and improves the targeted nature of process improvements.
Smart Images
Figure CN114826150B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of battery detection, and relates to an automatic evaluation system for performance loss of a solar cell caused by defects. BACKGROUND
[0002] In order to obtain a solar cell with higher conversion efficiency, process parameters need to be changed constantly in actual production, so as to compare the performance of samples produced by different manufacturing processes. However, some defects are inevitably generated or introduced in the solar cell material itself and in the process of production and manufacturing, which leads to abnormal reduction of conversion efficiency and affects the reliability of performance comparison of different samples. Therefore, it is beneficial to artificially eliminate the influence of defects on the solar cell, quantify the performance loss of the solar cell caused by the defects, and improve the process of the manufacturing personnel, so as to reduce the research and development cost of the high-efficiency solar cell. SUMMARY
[0003] In order to solve the problems in the prior art, the present application proposes an automatic evaluation system for performance loss of a solar cell caused by defects based on electroluminescence imaging detection technology and image restoration technology.
[0004] The present application proposes an automatic evaluation system for performance loss of a solar cell caused by defects, which comprises an input module, a cell edge detection module, a cell defect recovery module, a cell volt-ampere characteristic calculation module and a cell performance loss evaluation module.
[0005] The input module is used to read the electroluminescence spectrum, the external quantum efficiency and a group of electroluminescence images collected under different forward injection current densities of the solar cell wafer.
[0006] The read electroluminescence images of the solar cell are obtained by a CCD camera or a CMOS camera, the electroluminescence spectrum is obtained by a spectrometer, and the external quantum efficiency is obtained by a solar cell external quantum efficiency tester.
[0007] The cell edge detection module is used to determine the accurate position of the solar cell in the electroluminescence image.
[0008] The position of the solar cell in the electroluminescence image is defined as: after thresholding processing of the input electroluminescence image by using a Gaussian filter and an Otsu algorithm, the circumscribed rectangular region corresponding to the white contour in the thresholding image.
[0009] The cell defect recovery module is used to identify the defect area on the surface of the solar cell, calculate the predicted luminous intensity of each defect area and recover the defects, so as to obtain the predicted electroluminescence images under different injection current densities.
[0010] The battery volt-ampere characteristic calculation module obtains the actual or predicted volt-ampere characteristic curve of the solar cell according to the relationship between the actual or predicted electroluminescence intensity and the internal voltage of the battery.
[0011] The battery performance loss evaluation module obtains the actual or predicted performance parameters of the battery and compares them to quantitatively evaluate the performance loss of the solar cell caused by defects according to the actual or predicted volt-ampere characteristic curve.
[0012] The input module reads the electroluminescence spectrum, the external quantum efficiency and the electroluminescence images of the solar cell under different injection current densities, and transmits the above data information to the battery edge detection module.
[0013] The electroluminescence images of the solar cell are obtained by a CCD camera or a CMOS camera, the electroluminescence spectrum is obtained by a spectrometer, and the external quantum efficiency is obtained by a solar cell external quantum efficiency tester; the range of the injection current density is 0 to the short-circuit current density of the solar cell; the short-circuit current density of the solar cell is related to the material type and the manufacturing process of the solar cell.
[0014] The battery edge detection module obtains the circumscribed rectangular region corresponding to the white contour in the threshold image after threshold processing of the input electroluminescence image by Gaussian filtering and Otsu algorithm, which is the accurate position of the solar cell in the electroluminescence image.
[0015] The Gaussian filtering refers to a linear smoothing algorithm used in the noise reduction process of image processing, thereby improving the burr of the edge of the solar cell and obtaining an electroluminescence image with higher signal-to-noise ratio; the window size of the Gaussian filtering is S, and S is a positive integer; the Otsu algorithm is an algorithm for adaptively selecting a threshold, which can automatically set a threshold for the electroluminescence image after Gaussian filtering by using the method, thereby obtaining a threshold image; the setting of the threshold is related to the gray characteristics of the image, and when the threshold is adopted, the inter-class variance between the black and white regions of the threshold image satisfies the maximum, and the error probability of the thresholding is the minimum. Specifically, the Otsu algorithm is considered to be the best algorithm for threshold selection in image segmentation, which is simple to calculate and is not affected by image brightness and contrast, and therefore has been widely used in digital image processing.
[0016] The Otsu algorithm divides the image into background (black) and foreground (white) parts according to the gray characteristics of the image. Generally, "variance" is used as a measure of the uniformity of gray distribution, and the greater the inter-class variance between the background and the foreground, the greater the difference between the two parts (i.e. the black and white parts) constituting the image. Therefore, the segmentation (i.e. threshold selection) that maximizes the inter-class variance means the minimum misclassification probability.
[0017] Therefore, the Otsu algorithm makes the threshold selection of the maximum inter-class variance.
[0018] According to the obtained accurate position of the solar cell, the cell defect recovery module selects an electroluminescence image under a certain injection current density, traverses the entire electroluminescence image to obtain all defect regions D1 to D i ;
[0019] The definition of the defect region is that the electroluminescence intensity of the region satisfies the following conditions simultaneously: 1) there is a point P i with the lowest electroluminescence intensity in the region, and the intensity value of the point P i is lower than 1 / k1 of the average intensity value of a specified region N1, k1 being a given coefficient; 2) the intensity values of all points in the region are lower than 1 / k2 of the average intensity value of a specified region N2, k2 being a given coefficient; i being a positive integer.
[0020] The specified regions N1 and N2 are square or circular regions with different side lengths or radii centered on the point P i , and the defect region D i is contained in N1 and N2; for the same electroluminescence image, the side lengths or radii of multiple defect regions remain consistent, and N1 of different defect regions or N2 of different defect regions do not overlap with each other; the coefficients k1 and k2 are both real numbers greater than 1.
[0021] According to the obtained all defect regions D1 to D i , the cell defect recovery module replaces the electroluminescence intensity of the defect region with the predicted electroluminescence intensity of the defect region one by one to obtain a predicted electroluminescence image of the solar cell;
[0022] The predicted electroluminescence intensity of the defect region D i is defined as the average electroluminescence intensity of the part of the region N2 other than the defect region D i .
[0023] The cell defect recovery module sequentially selects corresponding electroluminescence images in order of increasing injection current density, obtains actual or predicted electroluminescence images under each injection current density, calculates the electroluminescence intensity of the cell under each injection current density, obtains the relationship curve of the injection current density and the actual or predicted internal voltage according to the reciprocal relationship between the actual or predicted electroluminescence intensity of the solar cell and the internal voltage, and outputs the actual or predicted volt-ampere characteristic of the solar cell.
[0024] The cell performance loss evaluation module obtains actual or predicted performance parameters of the cell according to the actual or predicted volt-ampere characteristic curve and compares them to quantitatively evaluate the difference between the actual performance of the cell and the predicted performance of the cell, i.e., the performance loss of the cell caused by defects.
[0025] The actual or predicted electroluminescence intensity of the solar cell is defined as the average electroluminescence intensity of all pixels in the actual or predicted electroluminescence image of the solar cell.
[0026] The solar cell volt-ampere characteristic evaluation module traverses the actual electroluminescence image under each injected current density, and according to the reciprocity relation between the actual electroluminescence intensity of the solar cell and the internal voltage, outputs the actual volt-ampere characteristic of the solar cell, and compares and evaluates it with the ideal volt-ampere characteristic obtained in step five; and obtains the difference between the actual volt-ampere characteristic and the ideal volt-ampere characteristic.
[0027] The degree to which the volt-ampere characteristic is superior to the actual volt-ampere characteristic represents the influence of defects on the quality of the solar cell; and the actual electroluminescence intensity of the solar cell is defined as the average electroluminescence intensity of all pixels in the actual electroluminescence image of the solar cell.
[0028] The relationship between the actual or predicted electroluminescence intensity and the internal voltage of the solar cell is:
[0029]
[0030] wherein V(I) is the internal voltage of the solar cell at the injected current I, k is the Boltzmann constant, T is the Kelvin temperature, q is the electronic charge, is the actual or predicted electroluminescence intensity, F(I) is the attenuation factor, A(I) is the calibration factor, R n (I) is the pixel value of pixel n, R min (I) is the lowest pixel value of the image, t(I) is the exposure time, S is the area of the solar cell, and EQE is the experimentally measured external quantum efficiency of the solar cell, <eqe> EL the average quantum efficiency of the battery in the wavelength range of the electroluminescence spectrum, the distribution of the photon density of the blackbody radiation spectrum with the photon energy, h is the Planck constant, c is the speed of light in vacuum, E g is the band gap of the battery, E is the energy; under the light working condition of the solar cell, Isun=q∫EQE(E)S AM1.5G (E)dE, I sun is the current density under the air mass AM1.5G spectrum S AM1.5G condition, the relationship between the internal voltage V(I) and (I sun ) is the volt-ampere characteristic of the solar cell under the light condition.
[0031] The solar cell performance parameters used for comparative evaluation include one or more of the following parameters: volt-ampere characteristic curve, open circuit voltage, fill factor and conversion efficiency, etc.
[0032] Wherein the computer automatically finds the intersection of the volt-ampere characteristic curve and the voltage axis, which is the open circuit voltage;
[0033] Fill factor = the area of the maximum rectangle formed by each point on the volt-ampere characteristic curve and the coordinate axis / (open circuit voltage x I sun );
[0034] Conversion efficiency = fill factor x (open circuit voltage x I sun ) / power density of air mass AM1.5G solar irradiation; the power density of the air mass AM1.5G solar irradiation is 1000W / m 2 .
[0035] The beneficial effects of the present application are that the influence of defects on the solar cell is artificially eliminated by using electroluminescence image recovery technology, the performance loss of the solar cell caused by defects can be quantified, which is beneficial to the targeted improvement of the process by the manufacturing personnel, and the research and development cost of high-efficiency solar cells is reduced. BRIEF DESCRIPTION OF DRAWINGS
[0036] Figure 1 The system block diagram of the automatic evaluation system of the performance loss of the solar cell caused by the defects of the solar cell according to the present application.
[0037] Figure 2 The flow chart of the automatic evaluation system of the performance loss of the solar cell caused by the defects of the solar cell according to the present application.
[0038] Figure 3 The actual electroluminescence image of the GaAs solar cell wafer collected by the CCD camera when the forward injection current density is 20mA / cm 2 .
[0039] Figure 4 The ideal electroluminescence image of the GaAs cell piece output by the cell defect recovery module when the forward injection current density is 20 mA / cm 2
[0040] Figure 5 The performance loss curve of the GaAs solar cell defect predicted by the present application.
[0041] Figure 6 The ideal volt-ampere characteristic curve (part) of the GaAs solar cell piece and the comparison diagram of the ideal open circuit voltage and the actual result. DETAILED DESCRIPTION
[0042] The application will be further described in conjunction with the following specific examples and drawings. The process, conditions, experimental methods, etc. for implementing the present application are generally known and commonly understood, except for the following specifically mentioned content, and the present application does not have special restrictions.
[0043] As shown in Figure 1 The automatic evaluation system for performance loss of solar cell defects of the present application includes an input module, a cell edge detection module, a cell defect recovery module and a cell volt-ampere characteristic evaluation module.
[0044] The input module is used to read the electroluminescence spectrum, external quantum efficiency and a group of electroluminescence images collected under different forward injection current densities of the solar cell piece.
[0045] The cell edge detection module is used to determine the accurate position of the solar cell in the electroluminescence image.
[0046] The cell defect recovery module is used to identify the defect area on the surface of the solar cell, predict the ideal luminous intensity of each defect area and recover the defects, obtain the ideal volt-ampere characteristic curve of the solar cell according to the relationship between the electroluminescence intensity and the internal voltage of the cell.
[0047] The cell volt-ampere characteristic evaluation module is used to call the relationship between the electroluminescence intensity and the internal voltage of the cell, obtain the actual volt-ampere characteristic curve of the solar cell from the original electroluminescence image, and compare the ideal and actual volt-ampere characteristics of the solar cell to evaluate the ideal quality of the solar cell under a specific process.
[0048] The solar cell volt-ampere characteristics used for comparison and evaluation include one or more of the following parameters: volt-ampere characteristic curve, open circuit voltage, fill factor and conversion efficiency.
[0049] EMBODIMENT
[0050] The automatic evaluation of the ideal volt-ampere characteristic of a GaAs solar cell with a size of 4cm x 2cm is taken as an example to illustrate the present application, and the specific steps are as follows:
[0051] First, the input module is imported with the current density of 0 to 25mA / cm 2 The electroluminescence images of a group of GaAs solar cell pieces taken by a CCD camera under forward injection current density, and the electroluminescence spectrum and external quantum efficiency of the cell pieces.
[0052] Further, the cell edge detection module uses a Gaussian filter with a window size of 25 to improve the burr of the solar cell edge and obtain electroluminescence images with higher signal-to-noise ratio; then uses the Otsu algorithm to automatically set the threshold value of the electroluminescence image, and realizes the thresholding processing of the input electroluminescence image; then intercepts the circumscribed rectangle region corresponding to the white contour in the thresholding image as the actual position of the GaAs solar cell piece in the electroluminescence image, as shown in Figure 3
[0053] Further, the cell defect recovery module uses the automatic evaluation method of the performance loss caused by the solar cell defect to identify the luminescence intensity of the defect area on the surface of the solar cell, and predict the ideal luminescence intensity, and according to the relationship between the electroluminescence intensity and the internal voltage of the cell, the ideal volt-ampere characteristic curve of the solar cell is obtained. Including the following steps:
[0054] Step one: transmit the input electroluminescence spectrum, external quantum efficiency and electroluminescence images under different injection current density to the cell edge detection module to obtain a group of GaAs solar cell electroluminescence images;
[0055] Step two: specify the electroluminescence image corresponding to the injection current density of 20mA / cm 2 as shown in Figure 3 If the electroluminescence intensity of a region on the image meets the following conditions at the same time, it is defined as a defect area D i (i is an integer greater than 1): first, there is a point P i with the lowest electroluminescence intensity in the region, and the intensity value of point P i is lower than 0.85 of the average intensity value of the square region with P i as the center and a side length of 10; second, the intensity values of all points in the region are lower than 0.95 of the average intensity value of the square region with P i as the center and a side length of 6;
[0056] Step three: calculate the average electroluminescence intensity of the square region with P i as the center and a side length of 6 excluding the defect area D i , and use it as the ideal electroluminescence intensity of the defect area D i Predicted electroluminescence intensity;
[0057] Step 4: Traverse the image and find all defect regions D1 to D2 that meet the conditions. i Transfer D1 to D i One by one, the electroluminescence intensity is restored to the predicted value, and the following is obtained: Figure 4 The GaAs solar cell shown is injected with an injection current density of 20 mA / cm². 2 Ideal electroluminescent image at that time;
[0058] Step 5: Select the corresponding electroluminescence images sequentially according to the order of injection current density from smallest to largest. Following the methods in Steps 3 and 4, obtain the ideal electroluminescence image for each injection current density. Based on the reciprocal relationship between the electroluminescence intensity and internal voltage of the solar cell, obtain the following... Figure 5 The curve showing the relationship between the injected current density and the ideal internal voltage outputs the ideal current-voltage characteristics of the solar cell.
[0059] Step 6: Iterate through the actual electroluminescence images at each injection current density, output the actual current-voltage characteristics of the solar cell based on the reciprocal relationship between the electroluminescence intensity and the internal voltage, and compare and evaluate them with the ideal current-voltage characteristics obtained in Step 5.
[0060] Furthermore, the battery ideal volt-ampere characteristic evaluation module outputs as follows: Figure 6 The diagram shows a comparison of the ideal and actual current-voltage characteristics of a solar cell, including a partial current-voltage characteristic curve and open-circuit voltage.
[0061] See Figure 6 The defects in the solar cells obtained using the method of this invention cause the performance loss curve to shift to the right compared to the actual current-voltage characteristic curve, and the ideal open-circuit voltage to increase compared to the actual open-circuit voltage. This indicates that defects have a certain impact on the quality of solar cells, and eliminating these defects can bring about an increase in open-circuit voltage of at least 0.17mV.
[0062] The scope of protection of this invention is not limited to the above embodiments. Any variations and advantages that can be conceived by those skilled in the art without departing from the spirit and scope of the inventive concept are included in this invention and are protected by the appended claims.< / eqe>
Claims
1. An automatic assessment system for performance loss caused by defects in solar cells, characterized in that, The system includes: an input module, a battery edge detection module, a battery defect recovery module, a battery volt-ampere characteristic calculation module, and a battery performance loss evaluation module; wherein, The input module is used to read the electroluminescence spectrum, external quantum efficiency, and a set of electroluminescence images acquired under different forward injection current densities of the solar cell. The battery edge detection module is used to threshold the input electroluminescent image through Gaussian filtering and Otsu's algorithm, and then obtain the bounding rectangular region corresponding to the white outline in the thresholded image, which is the accurate position of the solar cell in the electroluminescent image. The battery defect recovery module is used to identify defect areas on the surface of the solar cell, calculate the predicted luminescence intensity of each defect area, and recover the defects to obtain predicted electroluminescence images under different injection current densities. The defect region is defined as the region where the electroluminescence intensity simultaneously satisfies the following conditions: 1) There exists a point P in the region with the lowest electroluminescence intensity. i Point P i 1) The intensity value of the point is lower than 1 / k1 of the average intensity value of the specified area N1, where k1 is a given coefficient; 2) The intensity value of all points in the area is lower than 1 / k2 of the average intensity value of the specified area N2, where k2 is a given coefficient; i is a positive integer; The specified regions N1 and N2 are defined by point P. i Centered on a square or circular area with different side lengths or radii, and the defect area D... i Contained in N1 and N2; the predicted electroluminescence intensity of the defect region is defined as the area of region N2 excluding the defect region D. i Average electroluminescence intensity of the portion other than the specified value; Based on the precise location of the solar cell, the cell defect recovery module selects the electroluminescence image corresponding to a certain injection current density, and traverses the entire electroluminescence image to obtain all defect regions D1 to D2. i According to all defect areas D1 to D i The battery defect recovery module replaces the electroluminescence intensity of the defect area with the predicted electroluminescence intensity of the defect area one by one to obtain the predicted electroluminescence image of the solar cell; the battery defect recovery module selects the corresponding electroluminescence image in order of increasing injection current density to obtain the predicted electroluminescence image under each injection current density. The battery current-voltage characteristic calculation module obtains the actual and predicted current-voltage characteristic curves of the solar cell based on the relationship between the actual and predicted electroluminescence intensity and the internal voltage of the battery. The battery performance loss assessment module obtains and compares the actual and predicted performance parameters of the battery based on the actual and predicted current-voltage characteristic curves, and quantitatively assesses the performance loss of the solar cell caused by defects.
2. The automatic assessment system for performance loss caused by solar cell defects as described in claim 1, characterized in that, The input module reads the electroluminescence spectrum, external quantum efficiency, and electroluminescence images of the solar cell at different injection current densities, and transmits the above data to the cell edge detection module; The battery current-voltage characteristic calculation module calculates the actual and predicted electroluminescence intensity of the battery at each injected current density. Based on the reciprocity relationship between the actual and predicted electroluminescence intensity and the internal voltage of the solar cell, it obtains the relationship curve between the injected current density and the actual and predicted internal voltage, and outputs the actual and predicted current-voltage characteristics of the solar cell. The battery performance loss assessment module obtains and compares the actual and predicted performance parameters of the battery based on the actual and predicted current-voltage characteristic curves, and quantitatively assesses the difference between the actual performance and the predicted performance of the battery, i.e., the battery performance loss caused by defects.
3. The automatic evaluation system as described in claim 2, characterized in that, The electroluminescence image of the solar cell is captured by a CCD camera or a CMOS camera, the electroluminescence spectrum is obtained by a spectrometer, and the external quantum efficiency is obtained by a solar cell external quantum efficiency tester; the injection current density ranges from 0 to the short-circuit current density of the solar cell; the short-circuit current density of the solar cell is related to the material type and manufacturing process of the solar cell.
4. The automatic evaluation system as described in claim 2, characterized in that, The Gaussian filtering refers to a linear smoothing algorithm used in image processing for noise reduction, thereby improving the glitches at the edges of solar cells and obtaining an electroluminescent image with a higher signal-to-noise ratio. The window size of the Gaussian filtering is S, where S is a positive integer. The Otsu algorithm is an adaptive threshold selection algorithm. The program uses the Otsu algorithm to automatically set a threshold for the electroluminescent image, thereby obtaining a thresholded image.
5. The automatic evaluation system as described in claim 1, characterized in that, For the same electroluminescent image, the side length or radius of multiple defect regions is consistent, and the N1 or N2 of different defect regions do not overlap; the coefficients k1 and k2 are both real numbers greater than 1.
6. The automatic evaluation system as described in claim 2, characterized in that, The actual or predicted relationship between electroluminescence intensity and the internal voltage of the solar cell is as follows: ， Where V(I) is the internal voltage of the solar cell when the injected current I is applied, k is the Boltzmann constant, T is the Kelvin temperature, and q is the electron charge. The actual or predicted electroluminescence intensity, F(I) is the attenuation factor, A(I) is the calibration factor, and R is the actual or predicted electroluminescence intensity. n (I) represents the pixel value of pixel n, R min (I) represents the lowest pixel value of the electroluminescent image, t(I) represents the exposure time, S represents the cell area, and EQE represents the experimentally measured external quantum efficiency of the solar cell. <eqe> EL The average quantum efficiency of the battery within the wavelength range of the electroluminescence spectrum. Let be the distribution of photon density as a function of photon energy in the blackbody radiation spectrum, where h is Planck's constant, c is the speed of light in vacuum, and E is... g Where is the bandgap of the battery, and E is the energy; under illumination conditions, the solar cell... I sun For the spectrum S of air quality AM1.5G AM1.5G Current density under certain conditions, internal voltage V(I) and (I) sun The relationship between -I) refers to the current-voltage characteristics of a solar cell under illumination.< / eqe> 7. The automatic evaluation system as described in claim 2, characterized in that, The solar cell performance parameters used for comparative evaluation include one or more of the following parameters: current-voltage characteristic curve, open-circuit voltage, fill factor, and conversion efficiency; The computer automatically finds the intersection of the volt-ampere characteristic curve and the voltage axis, which is the open-circuit voltage; Fill factor = Area of the largest rectangle formed by each point on the volt-ampere characteristic curve and the coordinate axis / (open circuit voltage × I) sun ); Conversion efficiency = Fill factor × (Open circuit voltage × I) sun The power density of AM1.5G solar irradiance in air quality is 1000 W / m³. 2 .