Method for determining a degree of soiling of a surface of a photovoltaic module
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- DEUTSCHES ZENTRUM FÜR LUFT UND RAUMFAHRT E V
- Filing Date
- 2024-06-20
- Publication Date
- 2026-04-29
AI Technical Summary
Current methods for determining the degree of contamination on photovoltaic module surfaces in solar power plants are time-consuming, costly, and lack accuracy due to the need for manual measurements and reliance on gray values, which result in high signal-to-noise ratios and measurement errors, especially when dealing with large areas and varying contamination levels across surfaces.
A method utilizing a camera with multiple color channels to record image data of a defined dark area on the photovoltaic module, analyzing image channel information from the red or green color spectrum to determine contamination levels by comparing scattered radiation patterns, allowing for automated, spatially resolved, and precise contamination assessment.
This method provides a fast, accurate, and cost-effective means to assess contamination across large areas, enabling precise localization of contamination levels and optimizing cleaning processes by using image channel information to quantify the degree of contamination with improved signal-to-noise ratios and reduced measurement errors.
Abstract
Description
[0001] Method for determining the degree of contamination of a surface of a photovoltaic module
[0002] The present invention relates to a method for determining a degree of contamination of a surface of a photovoltaic module of a solar system or a solar power plant.
[0003] Solar power plants can have photovoltaic systems, which usually consist of several photovoltaic modules that convert solar radiation directly into electricity. The sun shines on large areas, and the solar radiation is absorbed by the solar power plant's photovoltaic modules.
[0004] Such large areas, which are usually set up in open fields, tend to become dirty, which limits the performance of the photovoltaic module.
[0005] Photovoltaic modules have a transparent layer, usually a transparent substrate, covering the photovoltaic or solar cells. Contamination reduces the transmission of solar radiation through the substrate and thus the power yield.
[0006] Therefore, the surfaces of photovoltaic modules in solar power plants are often cleaned at regular intervals. However, cleaning these surfaces, which can cover several thousand square meters, is a significant effort. To determine the optimal time for cleaning, attempts are made to measure the contamination.
[0007] Currently, contamination measurements on the mirror surfaces of solar power plants are performed using a handheld reflectivity measurement device. The handheld device measures the reflection of an artificially generated light beam at a fixed angle. The reflection measurement with this device is performed on a spot less than 1 cm in diameter on the mirror and cannot be automated. Therefore, to create a series of measurements, numerous measurements must be performed manually and on-site, which is time-consuming and costly, as five measurement points on a mirror sample are typically required with the handheld device in approximately 15 minutes. In a power plant, this working time is therefore a critical cost factor.
[0008] DE 10 2017 211 466 B3, published by the applicant, discloses a method in which the degree of soiling of a reflective surface is determined using a camera. A gray value of the reflective surface determined by the camera is compared with reference measurements of a soiling degree. From this comparison, the degree of soiling of the reflective surface can be determined.
[0009] A disadvantage of the previously known method is that, due to the use of gray values to determine the degree of contamination of the reflecting surface, a high signal-to-noise ratio is created when determining the degree of contamination, which means that the measurement results obtained are only partially accurate.
[0010] Contamination of the surfaces of photovoltaic modules in solar power plants is primarily caused by dust. Due to varying wind conditions around the surface of the photovoltaic module, varying degrees of contamination can occur at different locations on the surface of the photovoltaic module.
[0011] Furthermore, there are significant differences in the degree of contamination between the center of a photovoltaic module's surface and its edges. A quantitative assessment of the local degree of contamination by interpolating between individual measurement points is time-consuming and often subject to measurement errors.
[0012] The degree of contamination is determined for reflective surfaces of solar power plants as the quotient of the directed reflectance of a contaminated surface to that of a clean surface. In the case of photovoltaic power plants, where solar radiation is transmitted through a substrate, the degree of contamination is usually the ratio of the electrical output of a contaminated module to a cleaned module and thus essentially corresponds to the ratio of the transmittances of the contaminated and cleaned substrates. To optimize the cleaning of the reflective surface or the entire solar field, a fast, large-area, spatially resolved, and qualitative measurement method would be advantageous, with the measurement being carried out as automatically as possible.
[0013] It is therefore the object of the present invention to provide an improved method for determining the degree of contamination of a surface of a photovoltaic module.
[0014] The invention is defined by the features of claim 1.
[0015] The method according to the invention for determining the degree of contamination of a surface of a photovoltaic module of a solar power plant comprises the following steps: a) selecting and providing a defined dark area,
[0016] Positioning a camera having multiple color channels or adjusting the surface of the photovoltaic module such that an image of the dark area from the perspective of a camera having multiple color channels appears on the surface of the photovoltaic module, and
[0017] Illuminating the surface of the photovoltaic module with light, whereby a known or predetermined lighting situation is created, b) Recording image data of the surface of the photovoltaic module with the camera, c) Determining areas of an image consisting of the image data or several images consisting of the image data that show the reflection of the dark surface on the photovoltaic module, d) Determining image channel information from the image data by means of at least one first color channel of the camera that images the red or a part of the red color spectrum or by means of at least one first color channel of the camera that images the green or a part of the green color spectrum, wherein image information of the image data for the areas or sub-areas of the areas determined in step c) is determined by means of the image channel information,and e) determining the degree of contamination of the surface of the photovoltaic module from the image information for the areas determined in step c) or the sub-areas of the areas determined in step c); wherein, in a calibration step, the image information is compared with reference values for the degree of contamination.
[0018] In photovoltaic power plants where solar radiation is transmitted through a substrate, the degree of contamination is determined as the ratio of the electrical output power of a contaminated module to a cleaned module and thus essentially corresponds to the ratio of the transmittances of the contaminated and cleaned substrate.
[0019] The invention is based on the knowledge that the camera detects the reflected image of the dark area at a point on the surface of the photovoltaic module. Within the context of the invention, a "dark area" is understood to mean an area where the radiation flux of the dark area reflected in a directed manner by the photovoltaic module is smaller than the radiation flux scattered into the camera due to the illumination from dirt on the surface of the photovoltaic module. The dark area can be easily determined, for example, during calibration.
[0020] Contamination of the corresponding surface causes radiation reflected by the photovoltaic module to be scattered forwards and backwards. The camera records a portion of the scattered radiation depending on the illumination angle and the observation angle. In photovoltaic modules, the covering transparent substrate has a very high degree of transparency, resulting in very low reflection. The influence of the directly reflected portion of the brightness detected by the camera is therefore comparatively small and can be neglected, especially at higher levels of contamination. The brightness distribution of the absorbing solar cell surface located behind the substrate has a greater influence.The image of the dark area appearing on the surface of the photovoltaic module and captured by the camera is thus a reflection of the dark area on the covering transparent substrate together with an image of the solar cell surface.
[0021] In the images captured by the camera, a dirty surface of a photovoltaic module appears brighter than a clean surface, due to the scattering of solar radiation by the dirt particles described above. The degree of contamination can thus be determined by comparing a measurement of a dirty surface of a photovoltaic module with information from a reference measurement.
[0022] If the surface of the photovoltaic module's substrate is contaminated, the solar radiation incident on the substrate's surface and thus reaching the solar cell surface is reduced by the degree of contamination, which is assumed to be an average or a weighted average across the solar spectrum. The degree of contamination depends on the angle of incidence of the incoming solar radiation, since oblique incidence increases the number of dirt particles per projected area. The projected area corresponds to the area recorded by the camera. Additionally, as a simple approximation, a proportionality between the contamination losses and the number of particles can be assumed. Other relationships between contamination losses and the number of particles are also possible.
[0023] By using a defined dark area reflected by the photovoltaic module, the proportion of radiation transmitted from the dark area through the dirt, reflected, and retransmitted is known or determinable for the subsequent determination of the image channel information. For example, in a clean or cleaned module, it is possible to exclude the reflection of the defined dark area in order to perform a calibration using this background signal, which contains information about the transmission and reflection of the light coming from the dark area.
[0024] The determined image channel information is therefore essentially influenced by the radiation fluxes of the light illuminating the surface scattered by the dirt. In this context, the sky, for example, can be used as a defined dark area. Thus, the degree of contamination can be advantageously deduced from the image channel information. In the method according to the invention, in particular, the radiation flux scattered by the dirt is evaluated, so that the measurements according to the invention are particularly sensitive to contamination on the surface of the photovoltaic module.
[0025] In the method according to the invention, the multiple color channels of the camera can, for example, have a red channel, a green channel and a blue channel of a conventional digital camera.
[0026] In the context of the invention, a known or predetermined lighting situation is understood to mean that, in the case of directed light radiation, the radiant flux and the radiation direction are determinable and thus known or predetermined. Diffuse light with a known or predetermined radiant flux can also be generated or used as a lighting situation.
[0027] The method according to the invention can provide that the lighting situation is created artificially or occurs through natural radiation. Artificial lighting has the advantage that it can be created as very constant, i.e. non-fluctuating lighting, which can also be specified very precisely or is known. The use of natural radiation has the advantage that additional device-related expenditure is avoided. When using natural radiation, such as solar light, light reflected from the moon or starlight, it is advantageous to measure the radiation on or near the surface of the photovoltaic module for standardization purposes. The radiation can be measured using known methods. Preferably, the image data is recorded as raw image data and used accordingly in steps b) and c) or in the further method steps.Acquiring raw image data of the surface of the photovoltaic module offers the advantage of bypassing automatic camera corrections such as white balance or gamma correction. These corrections are undesirable in the method according to the invention, as they can influence the subsequent determination of the degree of contamination of the surface of the photovoltaic module. It can also be provided that the raw image data is acquired in the form of a series of raw image data.
[0028] The image data of the surface of the photovoltaic module can contain both areas that show a reflection of the dark area in the surface of the photovoltaic module and areas without this reflection. By determining the areas that show a reflection of the dark area in the surface of the photovoltaic module, the section of the image consisting of the image data that is essential for the method is defined. The areas not relevant to the method can be removed from the image data, so that the image consisting of the image data advantageously has a reduced amount of data. In principle, however, it is also possible for the image data of an existing image to contain only areas that show a reflection of the dark area in the surface of the photovoltaic module. In this case, in step d), either the image consisting of the image data can be used as a specific area or just a partial area.
[0029] The image channel information from the image data provides the image information that is recorded exclusively via one channel. The method according to the invention can use the image channel information from a red channel. The red color spectrum can also include a region of the infrared spectrum. It has been found that image channel information recorded via a red channel is significantly more advantageous for the method according to the invention than image channel information recorded, for example, via a blue channel. Image information from a clean surface has a comparatively low red component, which is recorded via the red channel. For dirty surfaces, the proportions of all color values increase, at least in all cases investigated by the applicants.Thus, the image channel information of the red color channel is particularly advantageous for the method according to the invention, since the signal of the red color channel changes significantly from a clean to a dirty surface.
[0030] It has thus been found that the image information of the red color channel of a clean surface of the photovoltaic module in comparison to a dirty surface of the photovoltaic module has greater deviations than when using a gray value for determining contamination according to the prior art and is therefore more advantageously suitable for determining contamination. The first color channel, which depicts the red or part of the red color spectrum, is thus particularly advantageous for a spectral analysis of the image data and for the method according to the invention.
[0031] In principle, the method according to the invention can also use the image channel information of a green channel. Just as with the red channel, the blue channel is also excluded when using the green channel. Since the defined dark area often has a high blue channel, such as when using the sky as a defined dark area, excluding the blue channel, as occurs when using the green channel, often leads to an improved signal-to-noise ratio.
[0032] The degree of contamination of the surface of the photovoltaic module is determined using information provided by the calibration step. The reference values used can, for example, be measurements of a cleaned surface of a photovoltaic module or measurements of a surface with a known degree of contamination. Furthermore, the reference values can, for example, be generated during calibration under laboratory conditions or carried out in situ. In this case, a section of the surface of the photovoltaic module can be used, whereby the degree of contamination is measured at specified time intervals and the image channel information is determined and stored together with the lighting situation. In principle, an estimation of the reference values is also possible; for example, the estimated values could be quantified using a scattering simulation.As an alternative to the lighting conditions according to step a), the calibration step can also be performed under conditions of a calibration lighting situation, which in turn can be laboratory conditions. It is preferably provided that the lighting situation and the calibration lighting situation are identical, or that the lighting situation is determined in step a), and differences between the lighting situation and the calibration lighting situation are taken into account when determining the image information in step d) using one or more correction factors.
[0033] The calibration step can generally be used for multiple measurements performed according to the method. Updating a calibration may be necessary, for example, if the lighting situation changes or if the type of contamination, such as different types of dust, causes contamination.
[0034] The method according to the invention allows the degree of contamination of a photovoltaic module's surface to be determined particularly easily. The degree of contamination is determined particularly accurately by using the first color channel, which represents the red or part of the red color spectrum.
[0035] Preferably, in step d), the image channel information is determined from the image data using at least the first color channel of the camera, which images the red or part of the red color spectrum, or using at least the first color channel of the camera, which images the green or part of the green color spectrum, and additionally using a further color channel of the camera, wherein the image channel information of the first color channel and the image channel information of the further color channel are weighted relative to one another. The further color channel can, for example, be a color channel that images the blue color spectrum or part of the blue color spectrum. The further color channel can be used if the defined dark area is unusually bright or if the reflections of the photovoltaic module are unusually dark.Furthermore, by weighting the first color channel and the further color channel to each other, the image information of the image data can be adapted so that the degree of contamination of the surface can be determined particularly accurately.
[0036] Preferably, in step d), the at least one first color channel imaging the red or a portion of the red color spectrum images a portion of the red color spectrum in the wavelength range 590 nm < λ < 1400 nm. It has been shown that this wavelength range is particularly advantageous for determining image channel information from the image data.
[0037] Preferably, after step b), a vignetting correction of the image data is performed to compensate for peripheral light falloff caused by the camera optics. Depending on the lens and aperture used on the camera, unwanted brightness artifacts occur in the peripheral regions of the raw images. These unwanted artifacts caused by peripheral light falloff can be compensated for by means of vignetting correction. Thus, after vignetting correction, the image data exhibit fewer brightness artifacts and are subsequently particularly advantageously suited for the method according to the invention.
[0038] Preferably, the areas or sub-areas determined in step c) are divided into sub-areas, with steps d) and e) being performed for each sub-area. By dividing the determined areas or sub-areas into sub-areas, the degree of contamination of the individual sub-areas can be determined, thus enabling a higher-resolution spatial assignment of the individual degrees of contamination to the respective positions on the surface of the photovoltaic module. In other words, by dividing the areas into sub-areas, the degree of contamination can be localized more precisely.
[0039] Preferably, the sub-areas are adapted to sections of the photovoltaic module. A photovoltaic module can, for example, have multiple solar cells. By adapting the sub-areas to sections of the photovoltaic module, the contamination of individual solar cells in a photovoltaic module can be determined. Thus, the contamination of a photovoltaic module can be particularly advantageously assigned to individual solar cells, so that, for example, a cleaning or maintenance process can be carried out specifically in the area of the respective contaminated solar cell.
[0040] In a preferred embodiment of the method, selected image data are excluded after step b). If multiple image data are recorded, it may happen that individual image data are unsuitable for determining a degree of contamination due to isolated measurement errors. Such a measurement error can be, for example, a direct solar reflection reflected by a module or another undesired reflection. By excluding selected unsuitable image data, method steps c) to e) are only carried out with image data that do not exhibit undesired reflections, thus avoiding measurement errors. The method according to the invention can thus be carried out particularly accurately and efficiently.
[0041] Preferably, steps a) to e) are performed repeatedly. For example, the positioning of the camera or the adjustment of the surface of the photovoltaic module according to step a) can be performed at a different camera angle during each repetition. Repeating the calibration step according to step e) is not absolutely necessary; however, a more precise determination of the degree of contamination is achieved by repeating the calibration step.
[0042] The calibration step can be performed separately from the evaluation or as part of the evaluation. It would also be conceivable to simultaneously acquire image data of the surface of the photovoltaic module and another module used to determine the reference values.
[0043] Since the respective image channel information of the image data of the surface of the photovoltaic module depends on the camera's recording angle, the proportion of light scattered by the contamination on the surface of the photovoltaic module received by the camera changes accordingly with a changing camera's recording angle. The image data of dirty photovoltaic modules can be compared with image data from a clean module. This generally allows conclusions to be drawn about the proportion of scattered light. By comparing the image to a photovoltaic module with known contamination, the relationship between contamination and scattered light can be modeled as a function of the camera's recording angle and the lighting situation. Using this calibration, a contamination level can be calculated for each area to be evaluated.This allows the degree of contamination of the surface of the photovoltaic module to be determined particularly accurately and independently of the camera's angle.
[0044] Preferably, it is provided that a power loss caused by contamination of the surface of the photovoltaic module is determined by means of the degree of contamination. Contamination generally leads to a power loss of the photovoltaic module, which, however, does not exclusively correlate with the degree of contamination of the surface of the photovoltaic module. For example, power losses can also be caused by shading of the photovoltaic module, corroded contacts of the photovoltaic module or other defects in the photovoltaic module. Nevertheless, regular contamination, for example due to natural events such as dust formation on the surface of the photovoltaic module, leads to power losses. In particular, the contamination pattern is also relevant, i.e. how a given amount of dirt is distributed on the photovoltaic module.In this case, for example, the electrical power loss of the photovoltaic module is generally not equal to the averaged optical loss. If the MPP (maximum power point) is tracked for each photovoltaic module, the nonlinear relationship can be calculated at the module level. If, for example, several modules are connected together (usually in series), the electrical loss can be calculated at the module level.
[0045] If the power loss of the photovoltaic module caused by contamination of the surface of the photovoltaic module is determined, it can be evaluated whether the total power loss of a photovoltaic module is caused by contamination or by other influences. This advantageously makes it possible to decide whether the contaminated photovoltaic module requires cleaning or maintenance. In a preferred embodiment of the method, the calibration step is carried out using reference measurements of the degree of contamination taken from a test surface with a predetermined or known level of contamination. Such a test surface can, for example, be the surface of a reference photovoltaic module. The reference measurements of the degree of contamination can have been generated by environmental influences over a specific period of time.For example, the reference photovoltaic module may not be cleaned for a period of one month. A reference measurement of the degree of contamination is then generated on this reference photovoltaic module. In the method according to the invention, the image information is compared with the reference measurement of the reference photovoltaic module. This allows the degrees of contamination of the surfaces of photovoltaic modules that are arranged in different positions, for example, to be compared with one another. Based on this comparison, conclusions can then advantageously be drawn from the contamination of the test surface to the contamination of the surface of a photovoltaic module to be examined, for example, whether the surface to be examined is more or less heavily soiled than the test surface.
[0046] Preferably, the specified or known contamination of the test surface is distributed as homogeneously as possible. The distribution of contamination on the surface of the photovoltaic module caused by environmental influences is usually not homogeneous. For example, more dust may be deposited in one corner of the surface of the photovoltaic module than in another corner, due to natural air turbulence that occurs in the area of the surface of the photovoltaic module. However, a homogeneous contamination distribution is suitable for improving the comparability of the contamination of the reference measurements with the image information of the surface of the photovoltaic module. The homogeneous contamination distribution thus creates a uniform comparison value with which the degree of contamination of the surface of the photovoltaic module can be determined particularly advantageously.
[0047] Preferably, the image channel information determined in step d) from the image data is used by means of the at least one first color channel of the camera to form an average image channel information item for the areas determined in step c), wherein the average image channel information is used to determine the image information of the image data for the areas or sub-areas of the areas determined in step c). It may be expedient to determine the degree of contamination using multiple image channel information items of the first color channel, which are generated, for example, by the repeatedly performed method, since this can reduce the influence of measurement errors.However, as the number of image channel pieces of information increases, so does the computational effort and thus the processing time required to determine the image information from the multiple pieces of image channel information that is then used to determine the degree of contamination. By forming an average image channel information, the image channel information from several measurements is, for example, arithmetically averaged and summarized, thus reducing the computational effort and time required to determine the image information. In addition, the influence of erroneous image channel information, which arises, for example, from the recording of a solar reflection or another reflection in step b) and which is not filtered out or taken into account in any other way, is reduced when forming the average image channel information, as this is averaged out by averaging.By forming the average image channel information, the method according to the invention can thus be carried out with less sensitivity to measurement errors.
[0048] In a preferred embodiment of the method, the camera is positioned on a tower or on a flying object. In principle, any surveillance camera that is already present, for example, in a solar system or a solar power plant can also be used. By positioning the camera on a tower, a constant recording condition is created in which the camera is directed from an elevated position onto the surface of the photovoltaic module. This allows the image data to be recorded particularly advantageously in step b). However, it may also be desirable to be able to position the camera flexibly in certain cases, for example if the image data is to be recorded using a different recording angle of the camera or if the degree of contamination of larger areas is to be determined. For this purpose, the camera can be positioned on a flying object.The flying object can be unmanned and controlled via a remote control. It is also possible for the flying object to automatically follow a predetermined route, such as a dome flight. By positioning the camera on a tower or a flying object, the method according to the invention can thus be carried out under constant recording conditions or flexibly.
[0049] Preferably, the calibration step is performed using reference measurements of the degree of contamination, with the reference measurements taking into account information about the type of contamination. The surface of the photovoltaic module can be contaminated in various ways, for example by dust or rain residue. Particularly in the dry regions of Southern Europe and North Africa, a layer of dust of varying thickness forms on the surface of the photovoltaic module over time. The information about the type of contamination can be used to determine how the determined degree of contamination should be interpreted after comparison with a reference measurement in which a degree of contamination from a surface of a photovoltaic module that has not been cleaned for four weeks, for example, was used.The information on the type of contamination allows the degree of contamination determined using the method according to the invention to be interpreted more meaningfully.
[0050] In a preferred embodiment of the method, it is provided that in the calibration step for determining the reference values, a maximum output power of a contaminated photovoltaic module of a solar power plant is determined and the maximum output power of the contaminated module is compared with the maximum output power of an uncontaminated module and then with the corresponding power ratio when both modules are uncontaminated. In addition to contamination, the output power of a module can also be impaired by other influences such as defective cells, corroded contacts or shading. By determining the maximum output power of a contaminated module and then comparing it with the maximum output power of an uncontaminated module, it can be determined whether a potential loss of performance is caused by contamination or other influences.For this purpose, the uncontaminated module should not only be uncontaminated, but also unaffected by other influences such as defective cells, corroded contacts, or shadows. The uncontaminated module is therefore in an optimal condition. The method according to the invention can thus determine the factors influencing the maximum output power of a module by comparing the maximum output power of a contaminated module with the maximum output power of an uncontaminated module.
[0051] Preferably, modules of the same design, i.e., at least of the same cell type, are used for the contaminated and the uncontaminated photovoltaic module. The maximum output power of a module varies depending on the design of the respective module. If modules of the same design are used, they can be advantageously compared with each other, and the degree of contamination can be determined with greater precision.
[0052] The method according to the invention offers the possibility of calculating the degree of contamination of photovoltaic modules in a simple and highly accurate manner, with a resolution per solar cell of the photovoltaic module possible. The method according to the invention enables improved monitoring of solar power plants.
Claims
Patent claims 1. A method for determining the degree of contamination of a surface of a photovoltaic module of a solar system or a solar power plant, comprising the following steps: a) selecting and providing a defined dark area, positioning a camera having multiple color channels or adjusting the surface of the photovoltaic module in such a way that an image of the dark area appears on the surface of the photovoltaic module as seen by a camera having multiple color channels, and Illuminating the surface of the photovoltaic module with light, thereby creating a known or predetermined lighting situation, b) Recording image data of the surface of the photovoltaic module with the camera, c) Determining areas of an image consisting of the image data or several images consisting of the image data that show the reflection of the dark area in the surface of the photovoltaic module, d) Determining image channel information from the image data using at least one first color channel of the camera that images the red or a part of the red color spectrum or using at least one first color channel of the camera that images the green or a part of the green color spectrum, wherein image information of the image data for the areas or sub-areas of the areas determined in step c) is determined using the image channel information,and e) determining the degree of contamination of the surface of the photovoltaic module from the image information for the areas determined in step c) or the sub-areas of the areas determined in step c); wherein, in a calibration step, the image information is compared with reference values of the degree of contamination.
2. Method according to claim 1, characterized in that in step d) the image channel information is determined from the image data by means of at least the first color channel of the camera imaging the red or a part of the red color spectrum or by means of at least the first color channel of the camera imaging the green or a part of the green color spectrum and additionally by means of a further color channel of the camera, wherein the image channel information of the first color channel and the image channel information of the further color channel are weighted relative to one another.
3. Method according to claim 1 or 2, characterized in that in step d) the at least one first color channel imaging the red or a part of the red color spectrum images a section of the red color spectrum in the wavelength range 590 nm < A < 1400 nm.
4. Method according to one of claims 1 to 3, characterized in that after step b) a vignetting correction of the image data is carried out to compensate for edge light fall-off caused by a camera optics.
5. Method according to one of claims 1 to 4, characterized in that the areas or sub-areas determined in step c) are divided into sub-areas, wherein steps d) and e) are carried out for each sub-area.
6. Method according to claim 5, characterized in that the sub-regions are adapted to sections of the photovoltaic module.
7. Method according to one of claims 1 to 6, characterized in that after step b) selected image data are excluded.
8. Method according to one of claims 1 to 7, characterized in that steps a) to e) are carried out repeatedly, wherein at each repetition the positioning of the camera or the adjustment of the surface of the photovoltaic module according to step a) is carried out at a different recording angle of the camera.
9. Method according to one of claims 1 to 8, characterized in that a power loss caused by contamination of the surface of the photovoltaic module is determined by means of the degree of contamination.
10. Method according to one of claims 1 to 9, characterized in that the calibration step is carried out with reference measurements of the degree of contamination taken from a test area with a predetermined level of contamination.
11. Method according to claim 10, characterized in that the determined contamination of the test area has a largely homogeneous distribution of the contamination.
12. Method according to one of claims 1 to 11, characterized in that the image channel information determined in step d) from the image data is used by means of the at least one first color channel of the camera to form an average image channel information for the areas determined in step c), wherein the average image channel information is used to determine the image information of the image data for the areas or sub-areas of the areas determined in step c).
13. Method according to one of claims 1 to 12, characterized in that the camera is positioned on a tower or on a flying object.
14. Method according to one of claims 1 to 13, characterized in that the calibration step is carried out with reference measurements of the degree of contamination, the reference measurements taking into account information on the type of contamination.
15. Method according to one of claims 1 to 14, characterized in that in the calibration step for determining the reference values, a maximum output power of a contaminated photovoltaic module is determined and the maximum output power of the contaminated module is compared with the maximum output power of an uncontaminated module. ratio is set and then compared with the corresponding performance ratio when both modules are uncontaminated, whereby modules of the same design are preferably used for the contaminated and the uncontaminated photovoltaic module.