Solar cell production process and solar cell production system

Abstract

The invention relates to a solar cell testing process, a solar cell production process, a solar cell testing system and a solar cell production system. The solar cell testing process comprises the following method steps: making at least one primary measurement (110) of substantially each of the plurality of solar cells (11, 12) by a primary measurement method to obtain a primary measurement (115) of each of the plurality of solar cells (11, 12); performing a secondary measurement (120) of at least one of the produced solar cells by a secondary measurement method to obtain a secondary measurement result (125); and classifying (150) each of the plurality of solar cells (11, 12) into a classification category according to a primary measurement (115) and / or a secondary measurement (125) associated with the solar cells (11, 12); wherein a sub-set of the plurality of solar cells (11, 12) is subjected to a secondary measurement (120) by means of a secondary measurement method, and the primary measurement method and / or the primary measurement result (115) is calibrated as a function of the secondary measurement result (125).

Description

[0001] illustrate:

[0002] The invention relates to a solar cell testing procedure, a solar cell production process, a solar cell testing device and a solar cell production plant.

[0003] In modern solar cell production plants, solar cells are subjected to a series of tests or measurements after they have been manufactured. These measurements are used firstly for quality assurance, since poor-quality or defective solar cells can be sorted out as a result. Secondly, the results of such tests can be used to sort the produced solar cells, in particular according to their power or efficiency. Sorting is also known as binning. The solar cells are measured directly at the end of the production line, that is to say in the so-called end-of-line test. This has the advantage that the solar cells do not first have to be transported to another location and can even be packaged in advance in order to test them there.

[0004] Currently, 100% of solar cells are contacted at the end of manufacturing via an IV solar simulator, measured according to the standard and converted to standard values ​​according to standard test conditions (STC). This is therefore referred to here as a 100% STC test. The measurement systems used for this purpose are complex, expensive and their throughput can hardly be increased further. The STC test of the IV characteristic curve (according to IEC 60904) involves standardized illumination of each solar cell under a standard spectrum (AM 1.5G). For the lighting conditions of AM 1.5G, the irradiance of 1 sun is defined as 100 mW / cm 2 During this irradiation, the current-voltage characteristic (IV measurement curve or simply IV curve) is measured in order to determine, among other things, the electrical power of the solar cell under standard illumination. Illumination according to standard spectra and intensities can only be achieved with relatively expensive LED or xenon lighting units, since the requirements for intensity, spectrum, stability and homogeneity are very high.

[0005] EP2823899A1 discloses a method and assembly device for sorting solar cell wafers in suitable containers based on their characteristics. For this purpose, the solar cells are tested and measured by different testing equipment. These tests are used to detect damage to the solar cell wafers, determine their respective colors, and determine their optical and electronic characteristics.

[0006] EP1647827A1 describes a test system for solar cells, which has an optical test device and an electrical test device arranged along a conveyor system. The solar cells driven by the conveyor system pass through the test device and undergo optical and electrical tests or measurements.

[0007] The object of the present invention is to reduce the operating costs per produced solar cell. The present invention achieves this in particular by reducing factory complexity and increasing the throughput of cell testing and sorting at the end of the production line.

[0008] According to the invention, this object is achieved by a solar cell testing procedure having the features of claim 1, by a solar cell production process having the features of claim 8, by a solar cell testing device having the features of claim 9 and by a solar cell production plant having the features of claim 10. Advantageous developments of the invention are listed in the dependent claims.

[0009] According to one aspect of the invention, a solar cell testing procedure is proposed, which feeds a plurality of previously produced solar cells for testing. For example, the solar cells can be supplied in groups, for example in packaged form, and are subjected to the solar cell testing procedure. According to another aspect of the invention, a solar cell production process is proposed, in which solar cells are produced and subsequently fed to the solar cell testing procedure. Preferably, the solar cells are produced continuously and are also continuously subjected to the testing or measurement scheme described below. For example, the produced solar cells can be tested in the order in which they are produced. The produced groups of solar cells referred to as "plurality of solar cells" can also be solar cells produced within a period of time, for example within an hour, a day or a week.

[0010] First, at least one primary measurement is performed on the solar cell by means of a primary measurement method. This primary measurement is performed on substantially every solar cell produced. Here, "substantially" means that no more than a subset of solar cells which is negligible compared to the total number of solar cells are not measured by the primary measurement method, such as solar cells which are obviously defective, in particular broken, or have scratches or color defects. Alternatively, the automatic detection of obvious defects such as these can also be considered as a result of the primary measurement or of a partial measurement part of the primary measurement method.

[0011] The primary measurement is used to determine the primary measurement result for each solar cell. Multiple primary measurements can be performed on each solar cell in parallel or continuously by different primary measurement methods. Accordingly, multiple primary measurement results can be determined. It should be pointed out here that the terms primary measurement and secondary measurement (defined later) are not used to specify the time sequence or other order between the measurements. Primary measurement refers to the sum of the first measurements and secondary measurement refers to the total number of second measurements. The difference between primary measurement and secondary measurement is that basically all the considered number of solar cells are subject to primary measurement, and only a subset is subject to both primary measurement and secondary measurement.

[0012] After the primary measurement or the plurality of primary measurements, or alternatively at a time between the first primary measurement and the second primary measurement, a secondary measurement is performed on at least one of the produced solar cells by a secondary measurement method to obtain a secondary measurement result. Here, a plurality of secondary measurements can also be performed on the solar cell simultaneously or successively by different secondary measurement methods. Accordingly, a plurality of secondary measurement results can be determined accordingly.

[0013] The primary measurement result / primary measurement results and / or the secondary measurement result / secondary measurement results may each include an indicator, a qualification and / or a set of indicators or qualifications. In particular, the primary measurement result and / or the secondary measurement result may include a measurement curve, such as a current-voltage measurement curve (IV measurement curve) or a photoluminescence (PL) image, an electroluminescence (EL) image or an infrared (IR) image.

[0014] After the solar cells have been measured, they are sorted into the corresponding classification classes, that is to say a so-called sorting takes place. This is achieved by selecting a suitable classification class based on the primary measurement results belonging to the solar cells. In addition or as an alternative, the classification class can also be selected based on the secondary measurement results. In particular, the primary measurement results can serve as a basis for determining the power value or the efficiency of the solar cell. Secondary measurement results, that is to say, in particular the IV characteristic curve, can also preferably be used to determine the power value or the efficiency of the solar cell.

[0015] The invention is based on the idea that only a subset of solar cells that have been measured by the primary measurement method is additionally measured by a secondary measurement method and that the primary measurement method and / or the primary measurement result are calibrated based on the primary measurement results and the secondary measurement results. In other words, secondary measurements by the secondary measurement method are performed (only) on a subset of a plurality of solar cells and the primary measurement method and / or the primary measurement result are calibrated based on the secondary measurement results. Solar cells that have not been measured by the secondary measurement method are sorted based on the primary measurement results determined thereon. This method has the advantage that the solar cells can be measured and sorted faster than if all solar cells were measured simultaneously using the primary measurement method and the secondary measurement method.

[0016] The primary measurement method can in particular be a faster and / or more cost-effective measurement method than the secondary measurement method. In contrast, the secondary measurement method should be designed so that it can be used to calibrate the primary measurement method. According to the invention, the secondary measurement method is therefore regarded as a normal measurement method or a standard measurement method, while the primary measurement method is used as a faster and / or cheaper method to replace the secondary measurement method. The calibration is performed in order to match the information value of the primary measurement result with the information value of the secondary measurement result or to take its information value. Preferably, the calibration is performed regularly. The calibration is assessed based on precision (i.e. the variability between the primary measurement and the secondary measurement) and accuracy (i.e. the average difference between the primary measurement and the secondary measurement), the average being calculated over a specified production or measurement time period, for example 1, 10 or 100 minutes.

[0017] A primary measurement result is determined on the solar cell using a primary measurement method. In this case, a conversion algorithm can be performed as part of the primary measurement method to derive the primary measurement result from the raw data obtained by the primary measurement method. This conversion algorithm can be a mathematical formula or a machine learning algorithm. In all cases, the conversion algorithm can be defined by model parameters. As described below, these model parameters can be adjusted as part of the calibration process.

[0018] Preferably, the primary measurement method is calibrated based on the secondary measurement results. This may involve adjusting the model parameters of the primary measurement method in such a way that the characteristic value of the solar cell obtained from the primary measurement result substantially corresponds to the characteristic value of the same solar cell obtained from the secondary measurement result. Alternatively or in addition, the primary measurement result may be calibrated based on the secondary measurement result. This involves converting the primary measurement result itself or the characteristic value of the solar cell obtained therefrom in such a way that the primary measurement result substantially corresponds to the secondary measurement result and / or the characteristic value of the solar cell obtained from the secondary measurement result substantially corresponds to the characteristic value of the solar cell obtained from the primary measurement result. Thus, the secondary measurement result can be inferred with a certain accuracy using only the primary measurement result. For this purpose, the model parameters may be adjusted in particular in the algorithm for converting from the primary measurement result to the characteristic value.

[0019] In an advantageous embodiment, it is provided that the secondary measurement method comprises a contact connection measurement method. This means that the measurement involves electrically contacting the solar cell to be measured. The solar cell can then be illuminated or irradiated by an illumination device, and electrical variables, in particular current and / or voltage, are measured at the contact connection during or after the irradiation. The latter measurement can be performed once or over a period of time. It is also possible to determine the average value of the electrical variable over a specific time period. For example, the current or voltage generated by the solar cell due to illumination can be measured over a period of time. Alternatively or in addition, the contact connection can be used to regulate one electrical variable while measuring another electrical variable at the contact connection. For example, this allows the determination of a current-voltage characteristic curve (IV characteristic curve). This can be done with or without illumination.

[0020] The secondary measurement method is preferably a current-voltage solar simulator measurement under standard test conditions (STC). This STC measurement of the IV characteristic curve (according to the IEC 60904 standard) involves standardized illumination of the solar cell under a standard spectrum (AM1.5G). The current-voltage characteristic is measured to determine the electrical power of the solar cell under standard illumination. Illumination with a standard spectrum and standard intensity can be achieved, for example, by LED or xenon lighting units, because the requirements for intensity, spectrum, stability and uniformity are high.

[0021] The primary measurement method is preferably a faster and / or cheaper measurement method than the secondary measurement method. In particular, the primary measurement method dispenses with illumination according to standard test conditions (STC). Performing cell measurements in combination with such non-STC measurements and STC measurements allows increased throughput, in particular when the non-STC measurements are performed in a contactless manner, for example, in the case of photoluminescence measurements. Installation engineering for non-STC measurements is significantly more cost-effective.

[0022] Preferably, the primary measurement method comprises one or more luminescence measurements, in which a luminescence image of the solar cell surface is captured, in particular an electroluminescence measurement and / or a photoluminescence measurement. Electroluminescence measurement involves contact-connecting the solar cell and exciting it by an electrical signal, that is, in particular by the flow of an electric current or an applied voltage. Photoluminescence measurement involves exciting the solar cell by irradiation, in particular laser irradiation. In both cases, the luminescence image can be captured by a camera. In this case, it can be chosen that the entire surface of the solar cell is captured by the camera at once. Alternatively, the solar cell surface can be scanned by the camera in regions.

[0023] The luminescence measurement is in equilibrium within the excess charge carrier lifetime, which is approximately 1 millisecond (ms) or of the order of magnitude of about 1 ms. Since the external covering does not change over time, hysteresis behavior (known, for example, from IV characteristic curve measurements of high-power silicon solar cells) does not occur. Thus, even measurements on moving solar cells can be performed, so-called "on-the-fly" measurements, if the local exposure time is in the range of a few milliseconds. Such luminescence measurements are therefore at least an order of magnitude faster than power measurements derived from IV characteristic curves.

[0024] Preferably, the power value, power level and / or efficiency of the solar cell is determined from one or more luminescence images captured from the solar cell and used as a basis for sorting. In addition, characteristic curve parameters, in particular the short-circuit current Isc, the open-circuit voltage Voc and / or the fill factor can also be determined. The one or more luminescence images are preferably recorded or captured in such a way that they contain relevant information in order to derive the power value, power level and / or efficiency of the solar cell therefrom, such as resistance effects, or this requires further primary measurements. Preferably, the luminescence images are evaluated for this purpose, for example by feature recognition using a machine learning algorithm, in particular using an artificial neural network.

[0025] For the recording of luminescence images, the following variants are particularly possible:

[0026] • Recording of multiple luminescence images, each at different injection levels. This variant can be used both for photoluminescence measurements (PL) and for electroluminescence measurements (EL).

[0027] • Recording of multiple luminescence images, each at a different illumination wavelength. This variant can be used in particular for photoluminescence measurements (PL).

[0028] Record multiple luminescence images at different detection wavelength ranges.

[0029] Recording of multiple luminescence images, in each case with front and / or back illumination (PL) and detection (PL, EL).

[0030] In a preferred embodiment, provision is made for the primary measurement method to comprise a contactless measurement method. In addition to the photoluminescence measurement, this option also comprises the contactless measurement of the external quantum efficiency EQE ("PL-QE"), the contactless measurement of the series resistance and the so-called "suns-PL" measurement. The common feature of all these measurements is that the signal is transmitted via the photoluminescence radiation and only the way in which the radiation is excited spatially, spectrally or temporally can establish the different properties. Furthermore, the contactless measurement facilitates fast and non-destructive measurement of a large fraction of the solar cells produced. This offers advantages, in particular, for thin, large and / or busbar-free solar cells. This can result in an increased throughput.

[0031] In an advantageous development, provision is made for the primary measurement method to include an imaging measurement method and / or a non-imaging measurement method. The imaging measurement method can in particular be the capture of the above-mentioned luminescence image. The non-imaging measurement method in particular measures the solar cell via contacts, for example measuring an IV characteristic curve. Non-contact non-imaging measurement methods are, for example, non-contact sheet resistance measurement, non-contact PL-QE measurement, non-contact series resistance measurement, non-contact suns-PL measurement.

[0032] As described above, multiple primary measurements can be performed on each solar cell by different primary measurement methods. For example, the first primary measurement method can be a PL method and the second secondary measurement method can be an EL method, which are used successively on the same solar cell.

[0033] In a preferred embodiment, the specified subset includes less than 10% or 1% of the number of solar cells. This means that, calculated over a period of time, for example over a day or one or more hours, less than 10% or 1% of the solar cells measured using the primary measurement method are also measured using the secondary measurement method. The fewer solar cells are measured using the secondary measurement method, the greater the throughput. On the other hand, both measurement methods must be performed regularly on the same solar cell, so that the secondary measurement results can be used to calibrate the primary measurement method and / or the primary measurement results. The above percentages refer in particular to average values ​​over a period of one or more hours. Preferably, before measuring the solar cells by both measurement methods, at least 5, 9, 50 or 80 solar cells are initially measured only by the primary measurement method.

[0034] The subset preferably reaches values ​​of less than 10% or 1% during normal operation after a start-up phase. In other words, a basic calibration or basic modeling is first performed during the first operation, after a product change, after a process change, after a material change and / or after a longer interruption, in particular a test interruption or a production interruption. This basic calibration involves first measuring a large portion of the produced solar cells using both the primary measurement method and the secondary measurement method until the sorting accuracy (i.e. variability) of the primary measurement method is within specification. The subset here is therefore very large and preferably approaches or reaches 100%.

[0035] Preferably, a basic calibration is provided, which is preferably performed during the start-up or new commissioning of a solar cell production process and / or is repeated regularly, for example weekly. Such a basic calibration or basic modeling is preferably performed based on sufficiently extensive and representative primary and secondary measurements in order to derive training data sets and calibration data sets for calibration therefrom. For this purpose, a large portion of the produced solar cells is initially preferably subjected to primary and secondary measurements, so that the subset is very high, preferably more than 90% or almost 100%.

[0036] If the production process and the substrate material fluctuate only within a low specification range, a regular basic calibration is preferably unnecessary. This means that continuous production can also be run without a basic calibration, which involves constantly subjecting a small number of solar cells to secondary measurements. This is because each individual secondary measurement already involves a primary measurement result and thus also a check on the primary measurement method. A basic calibration is particularly necessary if a too large proportion of the primary measurement results suddenly no longer match the secondary measurement results, for example because a completely different grade of silicon material is suddenly introduced during production.

[0037] After such a startup phase, the solar cell production process is preferably switched to continuous operation, during which a recalibration is performed regularly or occasionally based on the measurement results of the secondary measurement. This prevents a drift in the accuracy of the primary measurement compared to the secondary measurement, i.e. the average measurement value of a subset of at least 10 samples. During the recalibration, the model parameters are changed, which are used to interpret the primary measurement results and subsequently classify and sort the measured solar cells. It is ensured that both the primary measurement and the secondary measurement remain within the target specifications in terms of precision and accuracy, resulting in measurement results that are comparable in a deterministic manner over the long term.

[0038] According to a preferred development, provision is made for a calibration of the primary measurement method and / or the primary measurement results, including the application of a machine learning algorithm. This involves a plurality of pairs of primary measurement results and secondary measurement results, which in each case were previously determined on a solar cell, being provided to the algorithm. One or more such result pairs are determined for each solar cell for training the algorithm and for determining the model parameters of the algorithm. Alternatively, in the case of a result pair comprising a primary measurement result and a secondary measurement result available for the solar cell, the secondary measurement result can be used to determine a characteristic of the solar cell, such as a power value and / or an efficiency. The primary measurement results and the determined characteristics can then be used to train the algorithm and to determine the model parameters.

[0039] Subsequently, the algorithm with the determined model parameters can respond to the primary measurement results by determining an approximation of the solar cell characteristics as input, such as a power value approximation and / or an efficiency approximation. The better the algorithm works, the smaller the difference between the approximation and the characteristics of the solar cell obtained if the solar cell is measured by the secondary measurement method and the characteristics of the solar cell are obtained from the secondary measurement results determined thereby. As described above, recalibration is preferably performed regularly, for example at fixed time intervals or after a certain number of solar cells have been tested or after the estimated accuracy of the primary measurement has decreased, by subjecting at least one solar cell to primary and secondary measurements simultaneously. The (post)calibration serves to minimize the difference between the approximation of the characteristics of the solar cell obtained by the algorithm from the primary measurement results and the characteristics of the solar cell obtained from the secondary measurement results.

[0040] If the primary measurement method comprises one or more optical measurement methods (e.g. photoluminescence measurement method and / or electroluminescence measurement method), calibration is preferably used to adjust or modify calibration factors in order to track changes in the transfer of electrical properties to the optical recording technology, for example due to changes in the optical structure.

[0041] Preferably, provision is made for calibration of the primary measurement method and / or the primary measurement results, including the use of an artificial neural network. Thus, the above algorithm preferably includes an artificial neural network, and the model parameters are parameters of the artificial neural network, in particular weights of the artificial neural network.

[0042] The artificial neural network preferably comprises a recurrent neural network (RNN) and / or a convolutional neural network (CNN), in which the activity of neurons is calculated by discrete convolution. The artificial neural network may comprise multiple layers, in which one, two or more layers are convolved in this manner. If this is an RNN, then it is a convolutional recurrent neural network.

[0043] method:

[0044] I. Calibration of optical constants (Ai) for optical measurements (EL, PL or similar). Quantification of measured values, e.g. by STC measurements or by other reference measurements.

[0045] II. Ranking high-dimensional (eg, image), qualitative and quantitative measurement data by using statistical models (eg, machine models, ML) calibrated with relevant STC measurements.

[0046] 1. Imaging luminescent images:

[0047] i. Rating of image features, i.e. trained defect objects, and correlation with STC IV parameters.

[0048] ii. Based on the luminescence images, use statistical methods (e.g., machine learning (ML)) to predict STC IV parameters. In particular, deep learning and convolutional neural network (CNN) methods can be used here.

[0049] iii. In addition, an anomaly detection network (CNN or other) is optionally used to selectively trigger STC measurements. Anomaly detection detects outliers that can only be accurately measured using STC.

[0050] 2. Non-imaging measurement sequences include PL, PL-QE, non-contact RS (according to Kasemann et al.), suns-PL.

[0051] According to another aspect of the present invention, a solar cell testing device is provided. All configurations related to the solar cell testing program described above or below can be implemented accordingly in the solar cell testing device. In addition, all embodiments related to the solar cell production process described above or below about the solar cell testing program can be applied to the solar cell testing program accordingly.

[0052] According to another aspect of the present invention, a solar cell production plant is provided. All configurations related to the solar cell production process described above or below can be implemented accordingly in the solar cell production plant. In addition, all embodiments related to the solar cell testing device described above or below in relation to the solar cell production plant can be applied to the solar cell testing device accordingly.

[0053] Specifically, a solar cell production plant has a production part, a primary measurement device, a secondary measurement device and a sorting device. The production part is designed to produce a plurality of solar cells. The primary measurement device is designed to perform at least one primary measurement on substantially each of the plurality of solar cells produced by a primary measurement method, and determine a primary measurement result for each of the plurality of solar cells. The secondary measurement device is designed to perform at least one secondary measurement on at least one of the produced solar cells by a secondary measurement method, thereby determining a secondary measurement result. The sorting device is designed to assign each of the plurality of solar cells to a classification category according to the primary measurement result and / or the secondary measurement result associated with the solar cell. According to the present invention, the solar cell production plant is designed to perform secondary measurement on a subset of the plurality of solar cells by a secondary measurement method, and calibrate the primary measurement method and / or the primary measurement result according to the secondary measurement result.

[0054] A distribution system, for example comprising a robot arm and / or a conveyor belt, can ensure that all solar cells reach the primary measuring device and a subset of the solar cells reach the secondary measuring device. Alternatively or cumulatively, the device can be designed such that, although all produced solar cells pass through the primary measuring device and the secondary measuring device, the secondary measuring device is controlled in such a way that only a subset of the solar cells that pass through the secondary measuring device are also measured by the secondary measurement method.

[0055] The present invention will be explained below based on exemplary embodiments with reference to the accompanying drawings, in which:

[0056] Figure 1 A sequence diagram showing a testing and sorting method according to the prior art;

[0057] Figure 2 shows a solar cell production plant according to a preferred embodiment;

[0058] Figure 3 A flow chart showing a solar cell production process according to a preferred embodiment;

[0059] Figure 4 A flow chart showing a solar cell production process according to another preferred embodiment with multiple primary measurements performed simultaneously; and

[0060] Figure 5 A flow chart of a solar cell production process according to another preferred embodiment is shown.

[0061] The testing and sorting methods according to the prior art are based on Figure 1The flow chart shown is explained. Previously produced solar cells 11, 12 are fed to a standard measuring device 102. The standard measuring device 102 is a tester for measuring solar cells under standard conditions (STC standard test conditions). These are referred to as STC testers for short. The standard conditions (according to IEC 60904) in particular include standardized illumination / irradiation of the solar cells under a standard spectrum (AM1.5G). In addition, each solar cell is electrically contacted and its current-voltage measurement curve (IV curve) is measured during the irradiation.

[0062] like Figure 1 As shown, two or more standard measurement devices 102 are operated in parallel to increase the throughput. Typically, the standard measurement device 102 is arranged directly at the end of the production line and is therefore used as a so-called end-of-line test. Each produced solar cell 11, 12 is subjected to a standard measurement 102. For example, if the production line produces 8000 solar cells per hour, each of the two standard measurement devices 102 measures 4000 solar cells per hour.

[0063] It should be noted here that the rectangles shown in the diagrams in the accompanying drawings can represent both method steps in the method and corresponding device modules in the device. Figure 1 Two standard measurement devices 102 can be shown, each of which is fed with a portion of the produced solar cells 11, 12, or two standard measurements 102 which are carried out in parallel at one time in two produced solar cells 11, 12. This dichotomy is used below without explicit reference.

[0064] After the standard measurement 102 has been performed, the measured solar cells 11, 12 are fed to a sorting 150, the so-called sorting. This involves sorting the solar cells 11, 12 according to the measurement results determined in the standard measurement 102, in particular sorting the solar cells 11, 12 according to the cell power or cell efficiency that has been measured or derived from the measurement results.

[0065] According to a preferred embodiment, a solar cell production plant is as follows Figure 2 The entire production 100 of solar cells 11 and 12 is simplified in Figure 2 . However, the actual implementation naturally requires many production steps to produce the solar cell. It should also be pointed out here that the solar cells produced according to any configuration described herein can be only non-contact-connected solar cells. The produced solar cells therefore do not necessarily have to be contact-connected. Alternatively, the solar cell can already be fully contact-connected before being tested, that is to say electrically connected to the contacts of the measuring electronics.

[0066] like Figure 2 As shown, all solar cells 11, 12 initially undergo a primary measurement 110 in a primary measuring device 110. For each solar cell 11 measured here, the primary measurement 110 is used as a basis for determining a primary measurement result 115, which is fed to an evaluation module 130 (or evaluation). After the primary measurement 110, the first subset of solar cells 11 is fed directly to a sorting 150. The primary measurement result 115 is used for the sorting 150. The sorting control 137 is in Figure 2 This is indicated by a dashed arrow between the evaluation module 130 and the sorting container 150. The evaluation module 130 and the sorting container 150 are intended to identify the module having electronic and mechanical components in order to be able to carry out the corresponding method steps for evaluation or sorting.

[0067] The second subset of solar cells 12 is additionally subjected to secondary measurements 120 in a secondary measuring device 120, which involve the generation of secondary measurement results 125. These secondary measurement results 125 can also be provided to the evaluation module 130. A subsequent sorting 150 of the solar cells of the second subset 12 is also carried out under the control 137 of the evaluation module 130 as a function of the primary measurement results 115 and / or the secondary measurement results 125.

[0068] The secondary measurement 120 in the configuration discussed here is a standard measurement 102 according to the prior art. However, the secondary measurement can also be realized by another reliable secondary measurement method. In contrast, the primary measurement 110 is realized by a primary measurement method that is faster and / or cheaper than the secondary measurement method. In order to also allow an accurate power classification of the measured solar cells 11, 12 according to the primary measurement 110, some of the solar cells, i.e. the solar cells of the second subset 12, are subjected to secondary measurements 120. This can be achieved, for example, by subjecting every x-th solar cell to a secondary measurement 120, or by subjecting one or more solar cells to a secondary measurement 120 after a certain period of time. The secondary measurement results 125 are then used to regularly calibrate the primary measurement method and / or the primary measurement results 115, such as in Figure 2 As shown by the dashed arrow 133.

[0069] Figure 3A flow chart of a solar cell production process according to a preferred embodiment is shown. Here, solar cells are initially produced 100 and then all undergo primary measurement 110. A first subset 11 of solar cells is fed directly to sorting 150, while a second subset 12 is subjected to secondary measurement 120. The secondary measurement 120 also involves an evaluation 130 of the primary measurement results and the secondary measurement results of the solar cells from the second subset 12. In response to the results of this evaluation, the primary measurement method is calibrated 133. In some cases, calibration may be dispensed with, for example, if the primary measurement already produces completely correct results, or if it is specified that the primary measurement does not need to be recalibrated immediately for a first secondary measurement that is unrelated to the primary measurement.

[0070] Figure 4 The solar cell production process shown is similar to Figure 3 The difference shown is that here three primary measurements 110 are performed simultaneously in order to increase the throughput of the solar cells. Here too, the primary measurements 110 are performed on all solar cells 11, 12, while the secondary measurements 120 are performed only on a subset 12 of the solar cells. However, the produced solar cells 11, 12 are divided into three groups, which are subjected to the primary measurements 110 simultaneously. Although when the primary measurements 110 are performed much faster than the secondary measurements 120, Figure 3 The process shown is particularly advantageous, however, when the primary measurement 110 or the primary measurement device 110 is particularly cheap compared to the secondary measurement 120. Figure 4 The process shown is particularly advantageous. In the latter case, a plurality of primary measuring devices 110 can then be installed inexpensively, which leads to a further increase in throughput.

[0071] Figure 5 A flow chart of a solar cell production process according to another preferred embodiment is shown. An advantage of this embodiment is that association of a solar cell with its production line may be easier and the primary measurements may be calibrated specifically for each production line compared to a global calibration of all primary measurement stations.

[0072] As mentioned above, Figure 2-5 The elements of the diagrams can represent devices or method steps depending on their function. If they represent method steps, the flow of solar cells indicated by the arrows can be interpreted in such a way that the relevant solar cells undergo or do not undergo the method steps shown. In other words, a solar cell can also pass through a device without undergoing a method step that is performed with this device. In particular, this means that, according to one configuration, although all produced solar cells pass through the primary measuring device and the secondary measuring device, the secondary measuring device then only carries out secondary measurements on said subset of solar cells.

[0073] The calibration of the primary measurement by the secondary measurement according to a preferred embodiment will be explained below. The calibration method is based on the method known from the publication "End-of-Line Binning of Full and Half-Cut Cells using Deep Learning on Electroluminescence Images" by Y. Buratti et al., in the 2020 47th IEEE Photovoltaic Specialists Conference (PVSC), June 2020, No. 1, pp. 133-138, which can be divided into four steps, and subsequent references to the method steps "Step 1", "Step 2" and "Step 3" refer to the publication. Figure 1 :

[0074] a) First, the optical constants are determined as primary measurement results by means of secondary measurements, by virtue of the correlation between the open-circuit terminal voltage and the signal strength of the primary measurement (e.g., for PL measurements, by the formula Voc = kT / q*ln(C*I PL ), where C is the optical calibration constant).

[0075] b) The CNN is then trained. This can be done automatically or by a human (e.g., a quality engineer). To train the CNN, defect structures in primary measurements (e.g., PL and / or EL images) are detected, labeled, and classified. The CNN thus learns to automatically detect and classify similar defect structures on unknown images.

[0076] At the same time, the CNN also learns the primary measurement that defines a defect-free solar cell. Furthermore, it can use the existing quantitative signal calibration (factor C, see step a)) to determine whether the absolute signal of a defect-free solar cell has changed (step 1).

[0077] c) Next, the primary measurement results of the solar cell (information signal strength vector, defect list) are additionally correlated with the secondary measurement results (such as STC power measurement). This data correlation is used to perform a high-dimensional regression (machine model) that assigns power values ​​to detected defects and signal strengths as unambiguously as possible (step 2).

[0078] d) Perform steps a) to c) until the primary measurement is able to predict the secondary measurement to a specified precision and accuracy related to the sorting parameters (Isc, Voc, FF, potency). This completes the calibration of the primary measurement.

[0079] A partial recalibration process according to an advantageous configuration is described below. This is a recalibration that is performed regularly or occasionally during normal operation. It can have the following steps:

[0080] a. A certain proportion of solar cells is always directed to the secondary measurement, that is, regardless of the primary measurement result. This ensures that the predicted part is checked in time.

[0081] b. If the inspection detects an increase in variance, i.e. poor precision, or drift, i.e. excessive deviations from the mean measured values, such as Isc, Voc, FF or efficacy, e.g. caused by changes in optical constants in the primary measurements, then relearn / recalibrate the prediction model. In case of very large deviations, recalibration is possible (e.g. for new raw wafer material, different coating, different texture, etc.).

[0082] The selective recalibration process according to an advantageous configuration is described below. This is a recalibration performed during normal operation, if the primary measurement results justify this. It can have the following steps:

[0083] a. If the CNN of the primary measurement does not achieve sufficient accuracy in detecting defects or predicting the machine model, or an anomaly is detected, the secondary measurement is automatically activated and the CNN or ML model is relearned (automatically or with the assistance of a quality assurance engineer if labeling is required). However, the accuracy of the CNN is not defined here as a range of power parameters, but as a value between 0 and 1, and the frequency of correctly predicted defects and their type is related to the number of defects of that type that exist.

[0084] List of reference numerals:

[0085] 11 The first subset of solar cells

[0086] 12 The second subset of solar cells

[0087] 100 Steps for producing solar cells, production parts of production equipment

[0088] 102 Standard measurements, standard measuring devices

[0089] 110 Primary measurements, primary measuring devices

[0090] 115 Primary measurement results

[0091] 120 Secondary measurement, secondary measuring device

[0092] 125 Secondary measurement results

[0093] 130 Evaluation Modules

[0094] 133 Calibration

[0095] 137 Sorting Control

[0096] 140 Solar Cell Selection

[0097] 150 Sorting, sorting containers

Claims

1. A solar cell testing procedure for testing a plurality of produced solar cells (11, 12), comprising the following steps: - performing at least one primary measurement (110) on substantially each of the plurality of solar cells (11, 12) by means of a primary measurement method to determine a primary measurement result (115) for each of the plurality of solar cells (11, 12); - performing a secondary measurement (120) on at least one of the produced solar cells (11, 12) by means of a secondary measurement method to determine a secondary measurement result (125); and - assigning (150) each of the plurality of solar cells (11, 12) to a classification category based on the primary measurement results (115) and / or secondary measurement results (125) associated with the solar cells (11, 12), The invention is characterized in that the secondary measurement (120) is performed on a subset of the plurality of solar cells (11, 12) by the secondary measurement method, and the primary measurement method and / or the primary measurement result (115) is calibrated according to the secondary measurement result (125).

2. The solar cell testing procedure according to claim 1, characterized in that: The primary measurement method includes a non-contact measurement method.

3. The solar cell testing procedure according to claim 1 or 2, characterized in that: The primary measurement method includes an imaging measurement method and / or a non-imaging measurement method.

4. Solar cell testing procedure according to one of the preceding claims, characterized in that The secondary measurement method includes a contact connection measurement method.

5. Solar cell testing procedure according to one of the preceding claims, characterized in that The calibration of the primary measurement method and / or the primary measurement result (115) comprises the application of an artificial intelligence algorithm, in particular a machine learning algorithm.

6. The solar cell testing procedure according to claim 5, characterized in that: The calibration of the primary measurement method and / or the primary measurement results (115) comprises the use of an artificial neural network.

7. Solar cell testing procedure according to one of the preceding claims, characterized in that The subset comprises less than 10%, 1% or 0.1% of the number of the plurality of solar cells (11, 12).

8. A solar cell production process, comprising producing (100) a plurality of solar cells (11, 12) and testing the produced plurality of solar cells (11, 12) using a solar cell testing procedure according to one of the preceding claims.

9. A solar cell testing device, comprising: - a primary measurement device for performing at least one primary measurement (110) on substantially each of the plurality of solar cells (11, 12) by means of a primary measurement method to determine a primary measurement result (115) for each of the plurality of solar cells (11, 12); - a secondary measurement device for performing a secondary measurement (120) on at least one of the produced solar cells by means of a secondary measurement method to determine a secondary measurement result (125); and - sorting means for assigning (150) each of the plurality of solar cells (11, 12) to a classification category according to the primary measurement result (115) and / or the secondary measurement result (125) associated with the solar cell (11, 12), It is characterized in that The solar cell production plant is designed to perform the secondary measurement (120) on a subset of the plurality of solar cells (11, 12) by means of the secondary measurement method and to calibrate the primary measurement method and / or the primary measurement result (115) based on the secondary measurement result (125).

10. A solar cell production plant having a production section designed for producing (100) a plurality of solar cells (11, 12) and having a solar cell testing device according to claim 9.

Images