Solar panel inspection device and solar panel inspection method

Abstract

To provide a solar panel inspection device that can detect defects without interfering with the power supply from the solar panel. [Solution] A solar panel inspection device 1 for detecting defects in a solar panel 100 comprises: an imaging means 2 that photographs the solar panel 100 outdoors in daylight and acquires a first image which is an image in the wavelength range λ1 and a second image which is an image in the wavelength range λ2; and a generation means 4 that generates a diagnostic image from the first image and the second image, wherein the wavelength range λ1 is a wavelength range that includes the emission wavelength of the PL light L1 of the solar panel 100, and the wavelength range λ2 is a wavelength range that does not include the emission wavelength of the PL light L1 of the solar panel 100.

Description

Technical Field

[0001] The present invention relates to a solar cell panel inspection apparatus and a solar cell panel inspection method for detecting defects in solar cell panels.

Background Art

[0002] Solar cell panels are inspected for defects during manufacturing and before shipment, but even after shipment and installation outdoors, it is necessary to regularly inspect whether any defects have occurred in the solar cell panels. Conventionally, as one method for inspecting defects in solar cell panels, a PL inspection method using photoluminescence (PL) is known (see, for example, Patent Document 1). The PL inspection method is an inspection method that utilizes the phenomenon of light emission from a solar cell when it is irradiated with excitation light (this phenomenon is referred to as the photoluminescence phenomenon). If there is a defect or the like in the solar cell panel, the PL emission intensity decreases at the defective portion, so defects in the solar cell panel can be detected from information obtained by PL emission. Normally, PL inspection is performed in a dark room environment, which enables high-precision detection of defects. On the other hand, to inspect a solar cell panel that has been installed outdoors using the PL inspection method, it is necessary to remove noise such as reflected sunlight on the surface of the solar cell panel and scattered light from the environment.

[0003] The inspection apparatus described in Patent Document 1 generates a diagnostic image used for defect determination of a solar cell panel from an open-state PL image obtained by imaging the solar cell panel with the switch open (a state in which no current flows) and a closed-state PL image obtained by imaging the solar cell panel with the switch closed (a state in which current flows). In this inspection apparatus, since the diagnostic image is a difference image obtained by subtracting the closed-state PL image from the open-state PL image, or a division image obtained by dividing the open-state PL image by the closed-state PL image, the diagnostic image has reduced noise and can be used for defect determination of solar cell panels.

Prior Art Literature

Patent Literature

[0004] [Patent Document 1] Patent No. 6999911 [Overview of the project] [Problems that the invention aims to solve]

[0005] However, the inspection device described in Patent Document 1 requires the operation of a switch when generating diagnostic images used for defect detection in solar panels. When the switch is open, no current flows, thus hindering the power supply from the solar panel. Consequently, the generated power cannot be extracted, causing a temporary interruption of power supply during solar panel inspection and reducing the overall operational efficiency of the photovoltaic power generation system. Furthermore, in perovskite-type solar panels, which have attracted attention for their high energy conversion efficiency, light weight, and flexibility compared to conventional silicon-type panels, it is known that interrupting the current or switching the energizing state can cause a delay in the recovery of power generation efficiency due to material properties. Therefore, an inspection method that does not involve the operation of a switch is desired.

[0006] This invention has been made in view of the above-mentioned problems, and aims to provide a solar panel inspection device and a solar panel inspection method that can detect defects without interfering with the power supply from the solar panel. [Means for solving the problem]

[0007] The characteristic configuration of the solar panel inspection apparatus according to the present invention, which solves the above problems, is as follows: A solar panel inspection device for detecting defects in solar panels, A photographing means for photographing the solar panel outdoors during the daytime and acquiring a first image which is an image in the wavelength range λ1 and a second image which is an image in the wavelength range λ2, A generation means for generating a diagnostic image from the first image and the second image, Equipped with, The wavelength range λ1 is a wavelength range that includes the emission wavelength of PL light from the solar cell panel. The wavelength range λ2 is a wavelength range that does not include the emission wavelength of PL light from the solar cell panel.

[0008] With this configuration of solar panel inspection device, a solar panel can be photographed outdoors during the daytime, and a first image, which is an image in the wavelength range λ1, and a second image, which is an image in the wavelength range λ2, can be obtained. At this time, the solar panel is excited by sunlight, and PL light is generated. Here, since the wavelength range λ1 is the wavelength range that includes the emission wavelength of the PL light from the solar panel, the first image, which is the image in the wavelength range λ1, is an image in which the PL light image and an image of noise such as diffusely reflected light are superimposed. Also, since the wavelength range λ2 is the wavelength range that does not include the emission wavelength of the PL light from the solar panel, the second image, which is the image in the wavelength range λ2, is an image of noise such as diffusely reflected light. The diagnostic image is an image in which the noise has been removed (reduced). The generation means generates a diagnostic image from the first and second images that have been taken without interrupting the current or switching the power supply state, so defects can be detected while maintaining power supply from the solar panel.

[0009] In the solar cell panel inspection apparatus according to the present invention, The aforementioned solar cell panel is preferably a perovskite type solar cell panel.

[0010] In perovskite-type solar panels, when irradiated with light having energy exceeding the band gap (such as sunlight), PL light is emitted from areas without defects (normal areas). However, when the panel deteriorates due to humid heat or other factors and defects occur, PL light is not emitted from the defective areas. With this solar panel inspection device, since the solar panel is a perovskite-type solar panel, areas where PL light is not emitted in the diagnostic image can be detected as defects.

[0011] In the solar cell panel inspection apparatus according to the present invention, The wavelength range λ1 is preferably 759 to 770 nm.

[0012] With this configuration of solar panel inspection device, since the wavelength range λ1 is within the above range, the first image, which is an image of the wavelength range λ1, is an image in which the image of PL light emitted from the perovskite-type solar panel in a wavelength range where the influence of noise is small is superimposed with the image of noise. Therefore, even when the solar panel is generating power, weak PL light can be reliably detected.

[0013] In the solar cell panel inspection apparatus according to the present invention, The wavelength range λ2 is preferably 660 to 710 nm.

[0014] With this configuration of solar panel inspection apparatus, because the wavelength range λ2 is within the above range, the second image, which is the image in the wavelength range λ2, is a noisy image that does not contain PL light emitted from the perovskite-type solar panel. Furthermore, although the focal position of the image captured by the imaging means may shift depending on the wavelength range, because the wavelength range λ2 is within the above range, the wavelength range λ2 is close to the wavelength range λ1, and the effect of the focal position shift is reduced. Therefore, the diagnostic image generated by the generation means becomes clear, and defects can be easily detected.

[0015] In the solar cell panel inspection apparatus according to the present invention, The aforementioned solar cell panel is preferably a silicon crystal type or a CIS type solar cell panel.

[0016] When silicon crystal type and CIS type solar panels are irradiated with light (such as sunlight) that has energy exceeding the band gap, PL light is emitted from areas without defects (normal areas). However, when the panel deteriorates due to humid heat or other factors and defects occur, PL light is not emitted from the defective areas. With the solar panel inspection device configured in this way, since the solar panel is a silicon crystal type or CIS type solar panel, areas in the diagnostic image where PL light is not emitted can be detected as defects.

[0017] In the solar cell panel inspection apparatus according to the present invention, It is preferable that the wavelength range λ1 is 1110 to 1160 nm.

[0018] According to the solar cell panel inspection apparatus of the present configuration, when the wavelength range λ1 is within the above range, the first image, which is an image in the wavelength range λ1, becomes an image obtained by overlapping a PL light image emitted from crystalline silicon type and CIS type solar cell panels with a noise image in a wavelength range less affected by noise. Therefore, even when the solar cell panel is in a power generating state, weak PL light can be reliably detected.

[0019] In the solar cell panel inspection apparatus according to the present invention, It is preferable that the wavelength range λ2 is 700 to 900 nm.

[0020] According to the solar cell panel inspection apparatus of the present configuration, when the wavelength range λ2 is within the above range, the second image, which is an image in the wavelength range λ2, becomes a noise image that does not include PL light emitted from crystalline silicon type and CIS type solar cell panels. In addition, although there is a risk that the focal position of an image captured by the imaging means shifts depending on the wavelength range, when the wavelength range λ2 is within the above range, the wavelength range λ2 becomes a wavelength range close to the wavelength range λ1, and the influence of focal position shift is reduced. Therefore, the diagnostic image generated by the generating means becomes clear, and defects can be easily detected.

[0021] In the solar cell panel inspection apparatus according to the present invention, The imaging means is comprising a first bandpass filter that selectively transmits light in the wavelength range λ1, and a second bandpass filter that selectively transmits light in the wavelength range λ2, it is preferable that the first image is captured through the first bandpass filter, and the second image is captured through the second bandpass filter.

[0022] According to the solar cell panel inspection device of this configuration, an image in the wavelength range λ1 can be easily obtained by capturing the first image through the first bandpass filter. Further, an image in the wavelength range λ2 can be easily obtained by capturing the second image through the second bandpass filter.

[0023] In the solar cell panel inspection device according to the present invention, It is preferable to comprise a light amount adjusting means that adjusts the light amount when the imaging means captures the second image.

[0024] According to the solar cell panel inspection device of this configuration, the light amount adjusting means adjusts the light amount when the second image is captured, thereby limiting the light amount reaching the imaging means and suppressing the mixing of unnecessary high-luminance components caused by sunlight. As a result, the diagnostic image generated by the generating means becomes clearer, and defects can be detected efficiently.

[0025] In the solar cell panel inspection device according to the present invention, It is preferable that the imaging means captures the first image and the second image while the solar cell panel is generating power.

[0026] According to the solar cell panel inspection device of this configuration, since the imaging means captures the first image and the second image while the solar cell panel is generating power, the inspection of the solar cell panel can be performed in a state where the generated power can be continuously extracted, and the solar cell panel inspection device can detect defects without interfering with the power supply from the solar cell panel.

[0027] In the solar cell panel inspection device according to the present invention, It is preferable that the diagnostic image is a difference image obtained by subtracting the second image from the first image, or a division image obtained by dividing the first image by the second image.

[0028] With this configuration of solar panel inspection device, by using a difference image obtained by subtracting the second image from the first image, or a division image obtained by dividing the first image by the second image, as the diagnostic image, a clear image with noise removed (reduced) can be obtained, enabling high-precision detection of defects in solar panels.

[0029] In the solar cell panel inspection apparatus according to the present invention, In the aforementioned diagnostic image, it is preferable to include a detection means for detecting areas that are darker than the surrounding area as defects.

[0030] When a defect exists in a solar panel, the PL light emitted from that defect is weaker in intensity than the PL light emitted from areas without defects. Therefore, in a diagnostic image, which is an image of PL light with noise removed, the brightness value of the pixel representing the defect will be lower than the brightness value of the pixel representing the area without defects (normal area), resulting in a darker region than the surrounding area. Thus, by detecting this darker region as a defect, defects in solar panels can be detected more reliably.

[0031] The characteristic configuration of the solar panel inspection method according to the present invention, which solves the above problems, is as follows: A solar panel inspection method for detecting defects in solar panels, A shooting process in which the solar panel is photographed outdoors during the daytime, and a first image is obtained which is an image in the wavelength range λ1, and a second image is obtained which is an image in the wavelength range λ2, A generation step of generating a diagnostic image from the first image and the second image, It includes, The wavelength range λ1 is a wavelength range that includes the emission wavelength of PL light from the solar cell panel. The wavelength range λ2 is a wavelength range that does not include the emission wavelength of PL light from the solar cell panel.

[0032] According to this solar panel inspection method, a solar panel can be photographed outdoors during the daytime, and a first image (wavelength range λ1) and a second image (wavelength range λ2) can be obtained. At this time, the solar panel is excited by sunlight, and PL light is generated. Here, since wavelength range λ1 is the wavelength range that includes the emission wavelength of the PL light from the solar panel, the first image (wavelength range λ1) is an image in which the PL light image and an image of noise such as diffusely reflected light are superimposed. Also, since wavelength range λ2 is the wavelength range that does not include the emission wavelength of the PL light from the solar panel, the second image (wavelength range λ2) is an image of noise such as diffusely reflected light. The diagnostic image is an image in which the noise has been removed (reduced). The generation process generates a diagnostic image from the first and second images, which are taken without interrupting the current or switching the power supply state, so defects can be detected while maintaining power supply from the solar panel.

[0033] In the solar cell panel inspection method according to the present invention, Preferably, the diagnostic image includes a detection step in which areas darker than the surrounding area are identified as defects.

[0034] According to this solar panel inspection method, defects in the solar panel can be detected more reliably by detecting areas that are darker than the surrounding area in the diagnostic image as defects. [Brief explanation of the drawing]

[0035] [Figure 1] Figure 1 is a schematic diagram of a solar panel inspection device according to one embodiment of the present invention. [Figure 2] Figure 2 is a graph of the spectral irradiance of light reaching the ground in the wavelength range where oxygen absorption occurs. [Figure 3] Figure 3 shows an example of a diagnostic image in this embodiment. [Figure 4] Figure 4 is a flowchart showing the procedure for the solar panel inspection method according to this embodiment. [Figure 5]Figure 5 is a graph of the spectral irradiance of light reaching the ground in a wavelength range where water vapor absorption occurs. [Figure 6] Figure 6 shows examples of diagnostic images in another embodiment, where (a) is a diagnostic image of a silicon crystal solar cell and (b) is a diagnostic image of a CIS type solar cell. [Modes for carrying out the invention]

[0036] Embodiments of the solar panel inspection apparatus and solar panel inspection method of the present invention will be described in detail below with reference to the drawings. However, the present invention is not intended to be limited to the configurations described below.

[0037] [Solar panels] Before describing the solar panel inspection apparatus and solar panel inspection method of this embodiment, we will describe the solar panels that the solar panel inspection apparatus and solar panel inspection method of this embodiment target for inspection. The solar panels that the solar panel inspection apparatus and solar panel inspection method of this embodiment target for inspection are perovskite type solar panels (perovskite solar cells).

[0038] A perovskite solar panel is a solar cell that uses a compound with a perovskite structure as its light-absorbing layer. Generally, it consists of a multilayer structure centered around a perovskite light-absorbing layer (perovskite layer). The perovskite layer has the property of efficiently absorbing sunlight and generating electrons and holes. The carriers (electrons and holes) generated by the perovskite layer are separated and guided by the electron transport layer (ETL) and hole transport layer (HTL), and extracted as current to an external circuit. When light with energy exceeding the band gap (excitation light) is irradiated onto the perovskite layer, the generated carriers recombine within the perovskite crystal, and PL light with a peak wavelength in the range of 750-850 nm is emitted. However, if the perovskite layer deteriorates due to moist heat, etc., and defects occur, PL light will not be emitted at the defective areas.

[0039] [Solar panel inspection equipment] Figure 1 is a schematic diagram of a solar panel inspection device 1 according to one embodiment of the present invention. As shown in Figure 1, the solar panel inspection device 1 is a device that inspects a perovskite-type solar panel (hereinafter simply referred to as "solar panel") 100 that is irradiated with sunlight L in the daytime outdoors and has an electromotive force generated, and detects defects in the solar panel 100 using PL light L1 generated by the irradiation of sunlight L, which is the excitation light. As shown in Figure 1, the solar panel inspection device 1 is equipped with an imaging means 2 and a generation means 4. In addition, the solar panel inspection device 1 is optionally equipped with a light intensity adjustment means 3, a detection means 5, and a display means 6. The operation of the imaging means 2, light intensity adjustment means 3, generation means 4, detection means 5, and display means 6 is controlled by a control means 7. The following describes each component of the solar panel inspection device 1 in detail.

[0040] <Method of shooting> The imaging means 2 photographs the solar panel 100 outdoors during the daytime. In this embodiment, the imaging means 2 is a camera capable of photographing the top surface of the solar panel 100, and can be a digital camera having an image sensor in which solid-state image elements such as CCD (Charged-coupled devices) and CMOS (Complementary metal-oxide-semiconductor) are arranged in a two-dimensional array, and the image sensor detects the light incident on the imaging means 2. The imaging means 2 records the captured image on a storage device (not shown), such as a hard disk.

[0041] The imaging means 2 is controlled by the control means 7 to acquire a first image, which is an image in the wavelength range λ1, and a second image, which is an image in the wavelength range λ2. Outdoors during the daytime, sunlight L is irradiated onto the solar cell panel 100, and the solar cell panel 100 is excited by the sunlight L, generating PL light L1. Here, since the wavelength range λ1 is the wavelength range that includes the emission wavelength of PL light L1 from the solar cell panel 100, the first image, which is an image in the wavelength range λ1, is an image in which the PL light L1 image and noise L2 such as diffusely reflected light are superimposed. Also, since the wavelength range λ2 is the wavelength range that does not include the emission wavelength of PL light L1 from the solar cell panel 100, the second image, which is an image in the wavelength range λ2, is an image of noise L2 such as diffusely reflected light.

[0042] The imaging means 2 preferably captures the first and second images while the solar cell panel 100 is generating power. The state in which the solar cell panel 100 is generating power means that sunlight L is irradiating the solar cell panel 100 and the generated power can be extracted. For example, this means that a switch or turntable (not shown) implemented as part of the junction box is closed (current is flowing). By capturing the first and second images while the solar cell panel 100 is generating power, the imaging means 2 enables inspection of the solar cell panel 100 while the generated power can be continuously extracted, resulting in a solar cell panel inspection device 1 that can detect defects without interfering with the power supply from the solar cell panel 100.

[0043] The solar panel 100 to be inspected by the solar panel inspection device 1 of this embodiment is a perovskite-type solar panel. Therefore, the PL light L1 generated by irradiation with sunlight L has a peak wavelength in the range of 750 to 850 nm. However, since PL light L1 is an extremely weak emission, it is easily buried in noise L2. Furthermore, when the solar panel 100 is generating electricity, the carrier density decreases, and the PL emission becomes even weaker. Here, there are many Fraunhofer lines in the spectrum of sunlight L. Fraunhofer lines are numerous dark lines (absorption lines) that appear in the continuous spectrum of sunlight, etc., and are caused by the absorption of specific wavelengths when light passes through a substance. In the wavelength range where Fraunhofer lines exist, the influence of noise L2 becomes relatively small. Among the Fraunhofer lines, the dark lines caused by absorption by oxygen occur around wavelengths of 759 to 770 nm, which partially overlaps with the peak wavelength of PL light L1 generated from the perovskite-type solar panel. Figure 2 is a graph of the spectral irradiance of light reaching the ground in the wavelength range where oxygen absorption occurs. As shown in Figure 2, in the wavelength range of 759-770 nm, the irradiance of sunlight L decreases, thus reducing the influence of noise L2. By utilizing these characteristics of Fraunhofer lines, weak PL light L1 can be detected.

[0044] The wavelength range λ1 is preferably 759 to 770 nm, more preferably 760 to 764 nm, and particularly preferably 760 nm. Because the wavelength range λ1 is within the above range, the first image, which is an image of the wavelength range λ1, is an image in which the image of PL light L1 emitted from the perovskite-type solar cell panel in a wavelength range where the influence of noise L2 is small is superimposed with the image of noise L2. Therefore, even when the solar cell panel 100 is generating power, weak PL light L1 can be reliably detected.

[0045] The wavelength range λ2 is preferably 660 to 710 nm, more preferably 675 to 695 nm, and particularly preferably 685 nm. Alternatively, the wavelength range λ2 may be 810 to 840 nm, which is a longer wavelength than the emission wavelength of PL light L1 from the solar cell panel 100. In this case, the wavelength range λ2 is preferably 810 to 830 nm. Because the wavelength range λ2 is within the above range, the second image, which is an image of the wavelength range λ2, becomes an image of noise L2 that does not include the PL light L1 emitted from the perovskite-type solar cell panel. Furthermore, although the focal position of the image captured by the imaging means 2 may shift depending on the wavelength range, because the wavelength range λ2 is within the above range, the wavelength range λ2 is close to the wavelength range λ1, and the effect of the focal position shift is reduced. Therefore, the diagnostic image generated by the generation means 4, which will be described later, becomes clear, and defects can be easily detected.

[0046] In this embodiment, the imaging means 2 has a bandpass filter 21 that selectively transmits light in a specific wavelength range. In this invention, "transmitting light" means that the transmittance is 1% or more, and the transmission wavelength range of the bandpass filter 21 means the wavelength range in which the transmittance is 1% or more. The bandpass filter 21 includes a first bandpass filter and a second bandpass filter, and is configured to be switchable by the control means 7. The first bandpass filter is a filter that selectively transmits light in the wavelength range λ1, and if the solar cell panel 100 is a perovskite type solar cell panel, for example, it is preferable to use a bandpass filter with a transmission band of wavelength 750 to 770 nm. By capturing a first image through the first bandpass filter, an image in the wavelength range λ1 can be easily obtained. The second bandpass filter is a filter that selectively transmits light in the wavelength range λ2, and if the solar cell panel 100 is a perovskite type solar cell panel, for example, it is preferable to use a bandpass filter with a transmission band of wavelength 660 to 710 nm. By capturing a second image through a second bandpass filter, an image in the wavelength range λ2 can be easily obtained.

[0047] The imaging means 2 is set in an optimal position so as to be able to photograph the entire solar panel 100. In Figure 1, only one solar panel 100 is shown for the sake of simplicity, but the solar panel inspection device 1 of the present invention is not limited to a single solar panel 100, but can also be used to inspect a solar string consisting of multiple solar panels 100 connected together. Therefore, when inspecting a solar string consisting of multiple solar panels 100 connected together, it is preferable to set the imaging means 2 in a position that can photograph the entire solar string. For example, if the target of inspection is a solar panel 100 installed on the roof of a house, it is preferable to install the imaging means 2 in the basket of an aerial work platform or the like so that it can photograph the entire solar panel 100. Furthermore, if the solar panel 100 to be inspected is at a higher elevation and the shooting environment is difficult, it is also possible to mount the imaging means 2 on a small unmanned aerial vehicle (drone, etc.). Furthermore, the shooting means 2 may be configured to be movable so as to be able to photograph the top surface of the solar panel 100. It can photograph an area of ​​any position and size on the solar panel 100, take multiple images of a predetermined size while changing the shooting position of the solar panel 100, and combine these images to create an image of the entire solar panel 100.

[0048] In order to generate clearer diagnostic images in the processing by the generation means 4, which will be described later, it is preferable for the shooting means 2 to take multiple shots (for example, 50 times) of each of the first and second images. The shooting of the solar panel 100 by the shooting means 2 is not limited to control by the control means 7, but can also be configured to be controlled by direct operation of the shutter by an inspector, or by synchronization with a mobile information and communication device (smartphone, tablet terminal, etc.) via wireless LAN or IEEE802.15.1 (so-called Bluetooth®) and controlled by the inspector's shutter operation via the mobile information and communication device.

[0049] The imaging means 2 preferably has an autofocus function. Having an autofocus function allows the focal point to be automatically adjusted according to the surface condition of the solar panel 100. Furthermore, the autofocus function maintains the optimal focal state regardless of the distance to the solar panel 100 or its tilt, enabling efficient inspection.

[0050] <Light amount adjustment means> The light intensity adjustment means 3 adjusts the amount of light when the shooting means 2 captures the second image. Examples of the light intensity adjustment means 3 include adjusting the amount of light using an aperture mechanism, adjusting the exposure time, and reducing the sensitivity (ISO sensitivity) of the image sensor. Under direct sunlight or strong backlighting conditions, excessive light input to the shooting means 2 may cause image quality degradation such as overexposure, halation, and reduced contrast. However, by adjusting the amount of light when the shooting means 2 captures the second image using the light intensity adjustment means 3, the amount of light reaching the shooting means 2 can be limited, suppressing the inclusion of unwanted high-luminance components caused by sunlight L. As a result, the diagnostic image becomes clearer, and defects can be detected more efficiently.

[0051] <Generation means> The generation means 4 generates a diagnostic image used for defect detection of the solar panel 100 from the first and second images captured by the shooting means 2 and stored as data. The diagnostic image is an image from which noise has been removed (reduced). Since the generation means 4 generates the diagnostic image from the first and second images captured without interrupting the current or switching the power supply state, it is possible to detect defects while maintaining power supply from the solar panel 100. The generation means 4 can realize its function by executing a program on a computer having a CPU, memory, storage, etc.

[0052] The diagnostic image is preferably a difference image obtained by subtracting the second image from the first image, or a division image obtained by dividing the first image by the second image. The first image is an image in which the PL light L1 image and the noise L2 image are superimposed, and the second image is the noise L2 image. Therefore, by using this difference image obtained by subtracting the second image from the first image, or the division image obtained by dividing the first image by the second image, as the diagnostic image, a clear image with noise removed (reduced) can be obtained, and defects in the solar cell panel 100 can be determined with high accuracy. In order to make the diagnostic image clearer, it is preferable to use an integrated image obtained by image integration processing of multiple difference images pixel by pixel, or an integrated image obtained by image integration processing of multiple division images pixel by pixel, as the diagnostic image. Since the diagnostic image obtained by image integration processing of multiple difference images or multiple division images is a clearer image than a single difference image or a single division image, the condition of the solar cell panel 100 can be grasped more accurately.

[0053] When generating a subtraction image, generation means 4 preferably sets the value of pixels in the subtraction image to 0 if the value obtained by subtracting the second image from the first image is negative. Also, when generating a division image, generation means 4 preferably sets the value of pixels in the division image to 1 if the value obtained by dividing the first image by the second image is less than 1.

[0054] It is preferable to apply smoothing processing using a moving average filter or a Gaussian filter to the difference images, division images, and integrated images used as diagnostic images to adjust the image roughness. By applying appropriate processing to the difference images, division images, and integrated images in this way, it is possible to generate clearer diagnostic images that can determine defects in the solar cell panel 100 with even greater accuracy.

[0055] <Detection method> The detection means 5 detects areas that are darker than the surrounding area in the diagnostic image as defects. When a defect exists in the solar cell panel 100, the PL light L1 emitted from the defect has a weaker emission intensity than the PL light L1 emitted from areas without defects. Therefore, in the diagnostic image, which is an image of PL light L1 with noise L2 removed, the brightness value of the pixel representing the defect becomes lower than the brightness value of the pixel representing the area without defects (normal area), resulting in an area that is darker than the surrounding area. Thus, by detecting this darker area as a defect, defects in the solar cell panel 100 can be detected more reliably.

[0056] The detection means 5 may detect defects by identifying pixels in the diagnostic image. For example, if the brightness value of a pixel is below a threshold, it is considered a defective pixel, and based on that defective pixel, a portion where defective pixels are clustered can be detected as a defect. The threshold and range of brightness values ​​for determining a pixel as a defective pixel can be set by the operator, but it is preferable that they be set automatically. Known methods can be used to set the threshold brightness value for determining a pixel as a defective pixel.

[0057] Figure 3 shows an example of a diagnostic image in this embodiment. Figure 3 is a diagnostic image generated by acquiring a first image with a wavelength range of 756 to 776 nm and a second image with a wavelength range of 675 to 725 nm in order to detect a defect in a perovskite solar cell (UtomoLight, model UL-M23-G1). In a perovskite solar cell, if an electrode is short-circuited, the electrodes between unit cells are short-circuited, so the light emission of the unit cell weakens due to the short circuit, and a darker region than the surrounding area is created as shown by the arrow in Figure 3, which can be detected as a defect by the detection means 5.

[0058] <Display means> Display means 6 is a display that shows the diagnostic image generated by generation means 4 and the defects detected by detection means 5. By displaying the diagnostic image on display means 6, inspectors can visually diagnose the presence or absence of defects and identify the location of the defects. Display means 6 displays the diagnostic image of the solar cell panel 100 to be inspected, highlighting the defects detected by detection means 5 by surrounding them with rectangles, making it easy to determine the presence or absence and extent of defects in the solar cell panel 100. As a result, the accuracy of defect detection in the solar cell panel 100 and the reliability of the inspection are improved.

[0059] <Control means> The control means 7 implements the function of controlling the operation of each component of the solar panel inspection device 1 described above by executing a program on a computer having a CPU, memory, storage, etc. By the control means 7 controlling the operation of each component of the solar panel inspection device 1, it becomes possible to perform defect inspection of the solar panel 100 quickly and easily.

[0060] [Solar Panel Inspection Method] Figure 4 is a flowchart showing the procedure for the solar panel inspection method of the present invention. In the inspection method of the present invention, the imaging step and the generation step are performed sequentially using the solar panel inspection apparatus 1 shown in Figure 1, and optionally, a detection step is performed after the generation step. In the flowchart of Figure 4, each step of the solar panel inspection method is indicated by the symbol "S".

[0061] <Filming process: S1~S4> The solar panel inspection method of the present invention is performed while the solar panel 100 is generating power. In the imaging step, the imaging means 2 photographs the solar panel 100 outdoors in daylight and acquires a first image which is an image in the wavelength range λ1 and a second image which is an image in the wavelength range λ2.

[0062] First, the imaging means 2 photographs the solar panel 100 through a first bandpass filter that selectively transmits light in the wavelength range λ1, and acquires a first image, which is an image in the wavelength range λ1 (S1). To generate a clear diagnostic image, the imaging is repeated multiple times (for example, 50 times) (S2), and the first image is acquired again. The acquired first image is stored as data in storage such as non-volatile memory.

[0063] Next, the imaging means 2 photographs the solar cell panel 100 through a second bandpass filter that selectively transmits light in the wavelength range λ2, and acquires a second image, which is an image in the wavelength range λ2 (S3). To generate a clear diagnostic image, the imaging is repeated multiple times (for example, 50 times) (S4) to acquire a second image. The acquired second image is stored as data in storage such as non-volatile memory.

[0064] <Generation process: S5> In the generation process, the generation means 4 generates a diagnostic image from the first image and the second image (S5). For example, 50 first images and 50 second images are read from storage, and 50 difference images are generated by subtracting the second image from the first image pixel by pixel, with the first and second images as one pair. The integrated image obtained by image integration processing of these 50 difference images pixel by pixel is used as the diagnostic image. Alternatively, the diagnostic image may be obtained by reading 50 first images and 50 second images from storage, and generating 50 division images by calculating the ratio between the first and second images (dividing the first image by the second image), and then using the integrated image obtained by image integration processing of these 50 division images pixel by pixel. Or, for example, the diagnostic image may be obtained by reading 1 first image and 1 second image from storage and using difference images or division images generated from them.

[0065] <Detection process: S6> In the detection step, the detection means 5 detects areas that are darker than the surrounding area in the diagnostic image as defects (S6). Since the diagnostic image generated in the generation step (S5) is an image of PL light L1 with noise L2 removed, the brightness value of pixels representing defects becomes lower than the brightness value of pixels representing areas without defects (normal areas), resulting in areas that are darker than the surrounding area. Therefore, by detecting areas that are darker than the surrounding area in the diagnostic image as defects, defects in the solar cell panel 100 can be detected.

[0066] The solar panel inspection apparatus and solar panel inspection method of the present invention have been described above based on embodiments. However, the present invention is not limited to the configuration described in the above embodiments, and its configuration can be modified as appropriate without departing from the spirit of the invention. Examples of such modifications will be described as alternative embodiments. In the following alternative embodiments, parts that are the same as or similar to those in the above embodiments will be denoted by the same reference numerals, and detailed descriptions will be omitted. The following descriptions will focus on parts that differ from the above embodiments.

[0067] [Another embodiment] In the above embodiment, the solar panel 100 to be inspected by the solar panel inspection device and the solar panel inspection method was a perovskite type solar panel, but it is not limited to this and may be other types of solar panels. In another embodiment, the solar panel to be inspected by the solar panel inspection device is a silicon crystal type or a CIS type.

[0068] [Solar panel inspection equipment] <Method of shooting> In another embodiment, the imaging means 2 is not particularly limited as long as it is sensitive to the wavelength of PL light L1 in silicon crystal type or CIS type solar cell panels, but it is preferable to use a digital camera equipped with a photodetector such as an InGaAs (indium gallium arsenide) semiconductor photodetector having a sensitivity wavelength range of 800 to 2000 nm. When an InGaAs semiconductor photodetector is used as the photodetector, it becomes possible to capture the PL light L1 more clearly.

[0069] In another embodiment, since the solar panel 100 to be inspected by the solar panel inspection device 1 is a silicon crystal type or CIS type solar panel, the PL light L1 generated by irradiation with sunlight L has a peak wavelength in the range of 1000 to 1400 nm. Here, among the absorption lines in the spectrum of sunlight L, the absorption line caused by absorption by water vapor occurs around a wavelength of 1135 nm, which partially overlaps with the peak wavelength of PL light L1 generated from silicon crystal type or CIS type solar panels. Figure 5 is a graph of the spectral irradiance of light reaching the ground in the wavelength range in which water vapor absorption occurs. As shown in Figure 5, in the wavelength range of 1110 to 1160 nm, the irradiance of sunlight L is small, so the influence of noise L2 is reduced. By utilizing these absorption line characteristics, even weak PL light L1 can be detected in silicon crystal type or CIS type solar panels.

[0070] In another embodiment, the wavelength range λ1 is preferably 1110 to 1160 nm, more preferably 1125 to 1145 nm, and particularly preferably 1135 nm. Because the wavelength range λ1 is within the above range, the first image, which is an image of the wavelength range λ1, is an image in which the image of PL light L1 emitted from a silicon crystal type or CIS type solar cell panel in a wavelength range where the influence of noise L2 is small is superimposed with the image of noise L2. Therefore, even when the solar cell panel 100 is generating power, weak PL light L1 can be reliably detected.

[0071] In another embodiment, the wavelength range λ2 is preferably 700 to 900 nm, and more preferably 750 to 850 nm. Because the wavelength range λ2 is within the above range, the second image, which is an image of the wavelength range λ2, is an image of noise L2 that does not include PL light L1 emitted from silicon crystal type or CIS type solar cell panels. Furthermore, although the focal position of the image captured by the imaging means 2 may shift depending on the wavelength range, because the wavelength range λ2 is within the above range, the wavelength range λ2 is close to the wavelength range λ1, and the effect of the focal position shift is reduced. Therefore, the diagnostic image generated by the generation means 4 becomes clear, and defects can be easily detected.

[0072] <Detection method> Figure 6 shows an example of a diagnostic image in another embodiment, where (a) is a diagnostic image of a silicon crystal solar cell and (b) is a diagnostic image of a CIS type solar cell. Figure 6(a) is a diagnostic image generated by acquiring a first image with a wavelength range of 1125 to 1175 nm and a second image with a wavelength range of 756 to 776 nm in order to detect defects in a silicon crystal solar cell (GW-SOLAR, model GW-100A). In a silicon crystal solar cell, if a cell cracks, the crack becomes a darker area than the surrounding area, as indicated by the arrow in Figure 6(a), and can be detected as a defect by the detection means 5.

[0073] Figure 6(b) shows diagnostic images generated by acquiring a first image with a wavelength range of 1125 to 1175 nm and a second image with a wavelength range of 756 to 776 nm in order to detect defects in a CIS type solar cell (Solar Frontier model SF-106F). In a CIS type solar cell, if there is uneven film deposition (uneven coating), the area with uneven film deposition becomes a darker region than the surrounding area, as indicated by the arrow in Figure 6(b), and can be detected as a defect by the detection means 5. [Industrial applicability]

[0074] The solar panel inspection apparatus and solar panel inspection method of the present invention can be used to detect defects in operational solar panels installed on the roofs of houses or the like. [Explanation of Symbols]

[0075] 1. Solar panel inspection device 2. Method of filming 3 Light amount adjustment means 4 Generation means 5. Detection means 100 solar panels L1 PL light

Claims

1. A solar panel inspection device for detecting defects in a perovskite-type solar panel, A photographing means for photographing the solar panel outdoors during the daytime and acquiring a first image which is an image in the wavelength range λ1 and a second image which is an image in the wavelength range λ2, A generation means for generating a diagnostic image from the first image and the second image, Equipped with, The wavelength range λ1 is a wavelength range that includes the emission wavelength of PL light from the solar cell panel. A solar panel inspection device in which the wavelength range λ2 is a wavelength range that does not include the emission wavelength of PL light from the solar panel.

2. The solar cell panel inspection apparatus according to claim 1, wherein the wavelength range λ1 is 759 to 770 nm.

3. The solar cell panel inspection apparatus according to claim 1, wherein the wavelength range λ2 is 660 to 710 nm.

4. The aforementioned photographic means is It comprises a first bandpass filter that selectively transmits light in the wavelength range λ1, and a second bandpass filter that selectively transmits light in the wavelength range λ2. The solar cell panel inspection apparatus according to claim 1, wherein the first image is captured via the first bandpass filter and the second image is captured via the second bandpass filter.

5. The solar panel inspection apparatus according to claim 1, further comprising a light intensity adjustment means for adjusting the amount of light when the shooting means captures the second image.

6. The solar panel inspection apparatus according to claim 1, wherein the imaging means captures the first image and the second image while the solar panel is generating power.

7. The solar panel inspection apparatus according to claim 1, wherein the diagnostic image is a difference image obtained by subtracting the second image from the first image, or a division image obtained by dividing the first image by the second image.

8. The solar cell panel inspection apparatus according to claim 1, further comprising a detection means for detecting areas that are darker than the surrounding area in the diagnostic image as defects.

9. A solar panel inspection method for detecting defects in a perovskite-type solar panel, A shooting step involves photographing the solar panel outdoors during the daytime and obtaining a first image which is an image in the wavelength range λ1 and a second image which is an image in the wavelength range λ2. A generation step of generating a diagnostic image from the first image and the second image, It includes, The wavelength range λ1 is a wavelength range that includes the emission wavelength of PL light from the solar cell panel. A solar panel inspection method wherein the wavelength range λ2 is a wavelength range that does not include the emission wavelength of PL light from the solar panel.

10. The solar panel inspection method according to claim 9, which includes a detection step of detecting a region that is darker than the surrounding area in the diagnostic image as a defect.

Images