TWI777866B - Photovoltaic module detecting system and method thereof - Google Patents

Abstract

A photovoltaic module detecting system includes a drone and a processing terminal. The drone includes an infrared camera lens assembly and a visible light camera lens assembly, and the camera lens assemblies shoot a picture on a photovoltaic module and outputs an image. The processing terminal is signally connected to the drone, and includes an image clarifying module, an image blackscaling module, a negative image module and a state detecting module. The image clarifying module receives the image and outputs a clarified image. The image blackscaling module is signally connected to the image clarifying module and outputs a blackscaled image. The negative image module is signally connected to the image blackscaling module and output a negative image. The state detecting module is signally connected to the image clarifying module, the image blackscaling module and the negative image module and output a plurality of characteristics. According to an extension algorithm, the state detecting module output a state signal. Therefore, fault states can be categorized and the accuracy of detecting can be improved.

Description

Solar photovoltaic module detection system and method

The present disclosure relates to a photovoltaic module detection system and method thereof, and more particularly, to a photovoltaic photovoltaic module detection system and method capable of detecting different fault states.

In response to the current trend of environmental protection, solar photovoltaic modules are gradually developing, and their detection systems have also received attention. The conventional detection system analyzes the current, voltage and power changes of the solar photovoltaic module by installing bypass diodes or blocking diodes to determine whether there is a fault. However, due to the influence of environmental changes and The current, voltage and power at the time of failure are different, resulting in an inaccurate judgment in analysis. Later, although the use of drones to take photos was developed to improve the accuracy of detection, the type of fault cannot be determined because the pixels of the lens mounted on the drone are too low, so the accuracy cannot be improved. would result in excessive costs. Therefore, when it is impossible to accurately determine the failure of the photovoltaic modules, the industry often has to replace a whole set of photovoltaic modules, resulting in unnecessary waste of cost.

In view of this, a solar photovoltaic module detection system that can detect the fault states of different photovoltaic modules and improve the detection accuracy is still the goal of joint efforts of related industries.

One of the objectives of the present disclosure is to provide a solar photovoltaic module detection system and a method thereof, wherein the photovoltaic module is photographed by a drone, and the state detection module of the processing terminal outputs the correlation function value according to the extension algorithm, and according to the The largest value of the correlation function outputs a status signal. Thereby, different fault states of the solar photovoltaic module can be classified and detected, and the detection accuracy can be improved.

An embodiment of the present disclosure provides a photovoltaic module detection system, which includes a drone and a processing terminal. The drone includes at least one lens, and the lens captures at least one solar photovoltaic module and outputs at least one image. The processing terminal signal is connected to the drone, and includes a clear image module, an image black level module, an image negative film module and a state detection module. The image clearing module receives the image and outputs a clear image. The signal of the image black level module is connected to the clear image module, and receives the clear image and outputs a black level image. The signal of the image negative film module is connected to the image black level module, and receives the black level image and outputs a negative film image according to an environment setting value. The signal of the state detection module is connected to the image black level module and the image negative film module, receives the black level image and the negative film image and outputs complex features. Each feature includes a field value, and the corresponding complex state of each feature has a complex classical field range. According to an extension algorithm, the distance between the node domain value and the classical domain range corresponds to each state to output a correlation function value, and output a state signal according to the largest correlation function value.

Therefore, by analyzing the image of the solar photovoltaic module captured by the drone through the processing terminal, the status detection module can classify and detect different fault states of the solar photovoltaic module, and improve the detection accuracy.

In the solar photovoltaic module detection system according to the embodiment described in the preceding paragraph, the number of at least one lens may be two, the two lenses are a visible light lens and an infrared lens respectively, and the images include a visible light image and a thermal image.

According to the solar photoelectric module detection system of the embodiment described in the preceding paragraph, the drone can photograph the solar photoelectric module according to a planned path.

The solar photovoltaic module detection system according to the embodiment described in the preceding paragraph may further include an electronic device, the signal of which is connected to the processing terminal and receives the status signal.

An embodiment of the present disclosure provides a solar photovoltaic module detection method, which includes a photographing step, an image processing step, and a state detection step. The photographing step uses a drone to photograph at least one solar photovoltaic module and outputs at least one image to a processing terminal. The image processing step includes an image sharpening step, an image black leveling step, an image negative filming step and a feature outputting step. The image sharpening step is to output a clear image through an image sharpening module of the processing terminal according to the image. The step of image black leveling is to output a black level image through an image black level module of the processing terminal according to the clear image. The step of image negative filming is to output a negative film image through an image negative film module of the processing terminal according to the black level image. The feature output step is to output complex features through a state detection module of the processing terminal according to the black level image and the negative image, wherein each feature includes a domain value, and each feature corresponding to the complex state has a complex classical domain range. The state detection step is based on an extension algorithm, outputs a correlation function value corresponding to each state corresponding to the distance between the node domain value and the classical domain range, and outputs a state signal according to the maximum value of the correlation function value.

In this way, by taking images of the photovoltaic modules by drones, different fault states of the photovoltaic modules can be classified and detected, and the detection accuracy can be improved.

According to the solar photovoltaic module detection method of the embodiment described in the preceding paragraph, the state detection step may include setting a weight value for each feature, and outputting each correlation function corresponding to the weight value of each state corresponding to the distance between the node domain value and the classical domain range value.

According to the solar photovoltaic module detection method of the embodiment described in the preceding paragraph, the image negative filming step may include that the image negative film module sets an environmental setting value according to an environmental temperature, and outputs a negative film image according to the environmental setting value.

According to the solar photovoltaic module detection method of the embodiment described in the preceding paragraph, the feature output step may include that the feature includes a heating feature, and the state detection module outputs a heating point image according to the negative film image, and the heating feature corresponds to the heating point image.

The solar photovoltaic module detection method according to the embodiment described in the preceding paragraph may further include a path planning step. In the path planning step, the drone shoots the solar photovoltaic module according to a planned path.

The solar photovoltaic module detection method according to the embodiment described in the preceding paragraph may further include a status reporting step. In the status reporting step, when the status signal is a fault signal, the processing terminal transmits the fault signal to an electronic device.

Please refer to FIG. 1 and FIG. 2, wherein FIG. 1 shows a schematic diagram of a photovoltaic module detection system 100 according to an embodiment of the present disclosure, and FIG. 2 shows a processing terminal 120 according to the embodiment of FIG. 1 and a schematic diagram of the electronic device 130 . As shown in FIG. 1 and FIG. 2 , the photovoltaic module detection system 100 includes a drone 110 and a processing terminal 120 . The drone 110 includes at least one lens, and the lens captures at least one solar photovoltaic module 140 and outputs at least one image (not shown). The signal processing terminal 120 is connected to the drone 110 and includes an image clearing module 121 , an image black level module 122 , an image negative film module 123 and a state detection module 124 . The image clearing module 121 receives the image and outputs a clear image. The signal of the image black level module 122 is connected to the clear image module 121, and receives the clear image and outputs a black level image. The signal of the image negative film module 123 is connected to the image black level module 122, and receives the black level image and outputs a negative film image according to an environment setting value. The signal of the state detection module 124 is connected to the image black level module 122 and the image negative film module 123 , receives the black level image and the negative film image, and outputs complex features. Each feature includes a field value, and the corresponding complex state of each feature has a complex classical field range. According to an extension algorithm, the distance between the node domain value and the classical domain range corresponds to each state to output a correlation function value respectively, and output a state signal S1 according to the maximum of the correlation function values.

By analyzing the image of the solar photovoltaic module 140 captured by the drone 110 through the processing terminal 120, the detection accuracy can still be maintained without using a drone equipped with a high-pixel lens, and the solar photovoltaic module detection system 100 of the present disclosure can The different fault states of the photovoltaic modules 140 are classified and detected, and the detection accuracy is improved, so as to improve the efficiency of maintenance, and the faulty photovoltaic modules 140 can be replaced according to the fault state, thereby reducing unnecessary solar photovoltaic modules. Loss of power generation. The structural details of the solar photovoltaic module detection system 100 will be described in detail as follows.

In the embodiment of FIG. 1, the number of lenses may be two, and the two lenses are a visible light lens 111 and an infrared lens 112 respectively, and the images include a visible light image and a thermal image, but the disclosure is not limited thereto. The drone 110 can photograph the solar photovoltaic module 140 according to a planned path. In this way, the drone 110 can take pictures periodically, thereby saving time and cost during operation. Specifically, the image clearing module 121 may include a visible light image output module 1211 and a thermal image output module 1212 . The visible light image output module 1211 receives the visible light image and the signal is connected to the image black level module 122 . The thermal image output module 1212 receives the thermal image and the signal is connected to the image black level module 122, but the present disclosure is not limited thereto.

The electronic device 130 of the photovoltaic module detection system 100 is signal-connected to the processing terminal 120 and receives the status signal S1, and the signal connection can be a wired signal connection or a wireless signal connection. The electronic device 130 may be a mobile phone or a computer, but the present disclosure is not limited thereto. In this way, the status of the photovoltaic module 140 can be sent back to the user in real time.

…(1)。 Specifically, the processing terminal 120 can be a computer or a cloud system, and the image clearing module 121 , the image black level module 122 , the image negative film module 123 and the state detection module 124 can perform the aforementioned functions according to the instructions of MATLAB. The disclosure content is not limited to this. The image clearing module 121 can sharpen the image through the second derivative according to the Laplace image detection method to output a clear image, but the present disclosure is not limited thereto. Further, the clear image may include a clear visible light image and a clear thermal image. The visible light image output module 1211 and the thermal image output module 1212 are respectively sharpened by the second derivative to output a clear visible light image and a clear thermal image. The image black level module 122 can respectively output black level images according to the clear visible light image and the clear thermal image. In other embodiments, the thermal image output module 1212 can also directly output the thermal image without sharpening through the second derivative. The Laplacian image detection method is a method for realizing edge detection by performing second-order differential processing on the image, as shown in formula (1):

…(1).

…(2)；以及

…(3)。 The second derivative is more sensitive to discrete points and noise. Therefore, the image is first subjected to Gaussian convolution filtering and noise reduction, and then the Laplacian formula is used for edge detection. This filtering combination is called Laplacian of Gaussian filtering, and its equations are as shown in Equation (2) And formula (3):

…(2); and

…(3).

…(4)。 The sharpness value F _AMP of a clear image can be obtained by formula (4):

…(4).

The clear image is an image composed of N*N points, I(x,y) is the image value of the clear image at each point, and u is the average value of the entire clear image. When the sharpness value F _AMP is larger, it represents the sharper image sharpened by the second derivative.

Please refer to FIGS. 3A and 3B, wherein FIG. 3A shows a black level image B10 according to the embodiment of FIG. 2 and a schematic diagram of a negative image N10 output according to an environmental setting value level1, and FIG. 3B shows a schematic diagram according to the first 2 is a schematic diagram of a black-level image B10 and another negative image N20 output according to another environment setting value level2 in the embodiment. As shown in FIG. 3A , the image black level module 122 and the image negative film module 123 can respectively receive a clear thermal image to output a black level image B10 and output a negative film image N10 according to an environmental setting value level1. In the embodiment of FIG. 3A, the image negative film module 123 can set level1=0.65 according to the instruction of MATLAB, and obtain the high temperature characteristics C1 and C2 through the negative film image N10. As shown in FIG. 3B , the image negative film module 123 can output the negative film image N20 according to another environment setting value level2. In the embodiment of Fig. 3B, level2=0.75, and the high temperature characteristic C3 is obtained through the negative image N20. Further, the image negative film module 123 can set an environmental setting value according to the environmental temperature of the solar photovoltaic module 140 captured by the drone 110, and output the negative film images N10 and N20 according to the environmental setting value. By comparing Fig. 3A and Fig. 3B, it can be seen that the high temperature characteristic C2 is not the part of the solar photovoltaic module 140 that actually generates high temperature, and by adjusting the environmental setting value from level1 to level2, it can be known that the photovoltaic module 140 generates a high temperature The high temperature portion is in high temperature feature C1 (ie, high temperature feature C3). In this way, the environment setting value is adjusted according to the ambient temperature of the photovoltaic module 140 , so that the location of the photovoltaic module 140 where high temperature is generated can be more accurately determined and the accuracy of fault determination can be improved.

Please refer to FIG. 4 , which is a schematic diagram of a clear thermal image H30 combined with a negative image N30 according to the embodiment of FIG. 2 . As shown in FIG. 4 , the state detection module 124 obtains the hot spot from the black level image and outputs a hot spot image, and outputs a clear thermal image H30 for the convenience of identification, and the features output by the state detection module 124 include three heat generation features HP , the heat-generating feature HP is the three heat-generating points of the solar photovoltaic module 140 and corresponds to the image of the heat-generating points. In this way, the state detection module 124 can determine whether the solar photovoltaic module 140 is faulty through the heating characteristic HP.

Please refer to Figures 5A, 5B, 6A and 6B, wherein Figure 5A shows the visible light image of the photovoltaic module 140 according to the embodiment of Figure 1, and Figure 5B shows the image according to Figure 5A The thermal image of the photovoltaic module 140 in the embodiment, FIG. 6A shows another visible light image of the photovoltaic module 140 in the embodiment according to FIG. 1 , and FIG. 6B shows the photovoltaic module in the embodiment according to FIG. 6A . Thermal image of group 140. As shown in FIG. 5A and FIG. 5B , the state detection module 124 receives the visible light image and the black-level image output from the thermal image, and can detect and output the contamination feature C4 and the heating feature C6 caused by cracking, and the heating feature of high temperature caused by dust. C5 and the heating characteristic C7 caused by the diode short circuit. Furthermore, as shown in FIGS. 6A and 6B , the state detection module 124 can also detect and output the contamination feature C8 and the heating feature C9 caused by the tape. Thereby, the failure of the solar photovoltaic module 140 can be classified into different states through the visible light image and the thermal image.

Specifically, the complex features output by the state detection module 124 can be heating features, heating area features, module temperature features, and dirt features, wherein the heating features, heating area features, and module temperature features are based on the black level output by the thermal image. For images and negative images, the contamination characteristics are based on the black-level images output by the visible light images, but the present disclosure is not limited thereto. The heating feature, heating area feature, module temperature feature and dirt feature respectively contain a section of domain values. In this embodiment, the section field value is set at 1-10, but the present disclosure is not limited to this. In the present disclosure, the states of the solar photovoltaic module 140 can be classified into a bypass diode short-circuit fault state, a snail fault state, a dust fault state, a cracked fault state, a normal state, and a fully damaged state. The heating characteristics, heating area characteristics, module temperature characteristics and dirt characteristics corresponding to the aforementioned states can respectively have classical domain ranges as shown in Table 1: Table 1. The classical domain range of each feature corresponding to the state Feature status fever Heating area Module temperature Degree of dirt Bypass Diode Shorted 10 3-5 9-10 2-7 snail pattern 10 1-7 6-7 1-2 dust 10 1-3 4-5 5-9 cracked 10 4-8 5-6 1-3 normal 1 1-2 2-3 1-3 all damaged 10 8-10 7-8 2-6

…(5)；

…(6)；以及

…(7)。 The section field value and the classical field range are concepts of extension theory describing extension set and correlation function, and will not be repeated here. In the extension theory, the correlation function is used to express the degree of influence of the characteristics of things, so the state of the solar photovoltaic module 140 can be judged through the distance relationship between the node domain value of the complex feature and the classical domain range, and formula (5), (6), (7) means:

...(5);

…(6); and

…(7).

where ϕ is the distance between the node domain value and the classical domain, F ₀ is the interval of one of the states of the photovoltaic module 140 , f is the node domain value, v _a , v _b are the classical domain range, and F is the photovoltaic module In another state interval of 140, D is the position value of F ₀ and F, and K is the correlation function.

When the correlation function K>0, it means that f falls in the interval F ₀ , which means that the photovoltaic module 140 is in the state of F ₀ ; when -1<K<0, f is located in the extension domain of F ₀ , through The correlation functions of different states are compared, and the state of the solar photovoltaic module 140 is determined according to the largest value of the correlation function. In this way, by classifying the failures of the photovoltaic module 140 into different states, the detection accuracy can be further improved.

Furthermore, in order to improve the detection accuracy of the state detection module 124, a weight value can be set for each feature. In this embodiment, the setting of each feature weight value is as shown in Table 2: Table 2. Weights set for each feature feature Weights fever 0.30 Heating area 0.20 Module temperature 0.30 dirty 0.20

Table 3 shows the node domain values corresponding to different characteristics of the solar photovoltaic module 140 in a known fault state: Table 3. pen count fever Heating area Module temperature dirty known state 1 10 4 9 3 Bypass Diode Shorted 2 10 5 8 6 Bypass Diode Shorted 3 10 3 6 1 with snail pattern 4 10 4 7 2 with snail pattern 5 10 2 4 6 dusty 6 10 3 6 9 dusty 7 10 6 5 1 Crack occurs 8 10 7 6 3 Crack occurs 9 1 2 3 1 normal 10 1 1 2 2 normal 11 10 8 8 2 all damaged 12 10 10 7 5 all damaged

In this embodiment, the solar photovoltaic module 140 in the known fault state is detected through the state detection module 124 and outputs the correlation function value and the state signal as shown in Table 4: Table 4. Correlation function values and status signals output by the status detection module 124 pen count Bypass Diode Shorted with snail pattern dusty Crack occurs normal all damaged status signal 1 0.13 -0.08 -0.30 -0.18 -0.57 -0.14 Bypass Diode Shorted 2 -0.04 -0.07 -0.16 -0.12 -0.60 -0.05 Bypass Diode Shorted 3 -0.18 0.06 -0.17 -0.07 -0.41 -0.23 with snail pattern 4 -0.1 0.06 -0.22 -0.01 -0.31 -0.11 with snail pattern 5 -0.19 -0.15 0.15 -0.23 -0.29 -0.27 dusty 6 -0.21 -0.08 -0.05 -0.20 -0.52 -0.28 dusty 7 -0.26 -0.01 -0.23 0.10 -0.42 -0.21 Crack occurs 8 -0.13 -0.05 -0.23 0.05 -0.47 -0.10 Crack occurs 9 -0.63 -0.38 -0.40 -0.48 0.10 -0.68 normal 10 -0.63 -0.46 -0.53 -0.49 0.10 -0.65 normal 11 -0.17 -0.12 -0.39 -0.02 -0.47 0.06 all damaged 12 -0.18 -0.22 -0.27 -0.25 -0.64 0.05 all damaged

From Tables 3 and 4, it can be known that the photovoltaic module detection system 100 of the present disclosure can not only classify the fault state of the photovoltaic module 140, but also maintain high detection accuracy.

Please refer to FIG. 7 , which shows a flow chart of steps of a solar photovoltaic module detection method S100 according to an embodiment of the present disclosure. In particular, the photovoltaic module detection method S100 of the embodiment of FIG. 7 can be performed in conjunction with the photovoltaic module detection system 100 of the embodiment of FIG. 1, but the present disclosure is not limited thereto. The solar photovoltaic module detection method S100 includes a photographing step S120, an image processing step S130 and a state detection step S140. In the photographing step S120 , a drone 110 is used to photograph at least one solar photovoltaic module 140 and output at least one image to a processing terminal 120 . The image processing step S130 includes an image sharpening step S131 , an image black leveling step S132 , an image negative filming step S133 and a feature outputting step S134 . The image clearing step S131 is to output a clear image through an image clearing module 121 of the processing terminal 120 according to the image. The image sharpening step S131 may include a visible light image sharpening step S1311 and a thermal image sharpening step S1312. The clear image may include a clear visible light image and a clear thermal image. The visible light image sharpening step S1311 is to output a clear visible light image through a visible light image output module 1211 of the processing terminal 120 according to a visible light image of the image. The thermal image sharpening step S1312 is to output a clear thermal image through the thermal image output module 1212 of the processing terminal 120 according to a thermal image of the image. The image black leveling step S132 is to output a black level image through an image black level module 122 of the processing terminal 120 according to the clear image. The image negative filming step S133 is to output a negative film image through an image negative film module 123 of the processing terminal 120 according to the black level image. Further, the image black leveling step S132 is performed according to the clear visible light image and the clear thermal image in the clear image, and the image negative filming step S133 is performed based on the black level image output from the clear thermal image. The feature output step S134 is to output complex features through a state detection module 124 of the processing terminal 120 according to the black level image and the negative image, wherein each feature includes a domain value, and each feature corresponding to the complex state has a complex classical domain range respectively. The state detection step S140 is based on an extension algorithm, outputs a correlation function value corresponding to each state corresponding to the distance between the node domain value and the classical domain range, and outputs a state signal S1 according to the largest correlation function value.

Thereby, the solar photovoltaic module detection method S100 can classify and detect different fault states of the photovoltaic module 140, and improve the detection accuracy, thereby improving the efficiency during maintenance and replacing the faulty photovoltaic module according to the fault state. Group 140, reducing unnecessary loss of power generation of the solar photovoltaic module.

The solar photovoltaic module detection method S100 may further include a path planning step S110 and a status reporting step S150. The path planning step S110 is to shoot the solar photovoltaic module 140 by the drone 110 according to a planned path. In the status reporting step S150 , when the status signal S1 is a fault signal, the processing terminal 120 transmits the fault signal to an electronic device 130 .

Please refer to FIG. 8 , which shows a detailed step diagram of the solar photovoltaic module detection method S100 in the embodiment according to FIG. 7 . As shown in FIG. 8, the path planning step S110 may include sub-steps S111 and S112. In sub-step S111, after the UAV 110 inputs the planned path, the UAV 110 is driven to perform the flight operation according to the planned path. Next, sub-step S112 is executed, and the drone 110 moves to the location where the photovoltaic module 140 is photographed according to the planned path. The photographing step S120 may include sub-steps S121, S122, S123, and S124. In sub-step S121, the drone 110 shoots the photoelectric module 140 with the camera lens; when the drone 110 successfully shoots, the sub-step S122 is executed to wirelessly transmit the image to the cloud system and back to the processing terminal 120, and Execute sub-step S123 to confirm whether the drone 110 has photographed all the photovoltaic modules 140. When the drone 110 has not finished photographing, execute sub-step S112, move the drone 110 to the location of photographing other photovoltaic modules 140, And re-execute sub-steps S121 and S122; when the drone 110 shoots all the solar photovoltaic modules 140, move the drone 110 to the drone docking station; when the drone 110 fails to shoot, execute sub-step S124, move the drone 110, and re-execute sub-step S121. The solar photovoltaic module detection method S100 may further include a step S151. After receiving the image, the processing terminal 120 sequentially executes the image processing step S130 , the state detection step S140 , the state reporting step S150 and the step S151 . In step S151 , the fault signal is transmitted to the client or the manufacturer through the electronic device 130 . Thereby, subsequent maintenance or treatment can be performed.

Although the present disclosure has been disclosed as above in embodiments, it is not intended to limit the present disclosure. Anyone skilled in the art can make various changes and modifications without departing from the spirit and scope of the present disclosure. The protection scope of the disclosed content shall be determined by the scope of the appended patent application.

100: Solar Photoelectric Module Detection System 110: Drone 111: Visible light lens 112: Infrared lens 120: Process Terminal 121: Image clear module 1211: Visible light image output module 1212: Thermal image output module 122: Image black level module 123: Image negative film module 124: Status detection module 130: Electronic Devices 140: Solar photovoltaic module B10: Black Level Image C1, C2, C3: high temperature characteristics C4, C8: Dirty features C5, C6, C7, C9: Thermal characteristics H30: Clear thermal image HP: heat characteristics N10, N20, N30: Negative image S1: Status signal S100: Detection method of solar photovoltaic modules S110: Path Planning Step S111, S112, S121, S122, S123, S124: Sub-steps S120: Shooting steps S130: Image processing step S131: Image sharpening step S1311: Visible light image sharpening steps S1312: Thermal image sharpening steps S132: Image black leveling step S133: Image negative filming step S134: Feature output step S140: state detection step S150: Status reporting step S151: Steps

FIG. 1 is a schematic diagram of a solar photovoltaic module detection system according to an embodiment of the present disclosure; FIG. 2 shows a schematic diagram of the processing terminal and the electronic device according to the embodiment of FIG. 1; FIG. 3A is a schematic diagram of a black-level image and a negative image output according to an environmental setting value in the embodiment of FIG. 2; FIG. 3B is a schematic diagram illustrating a black-level image according to the embodiment of FIG. 2 and outputting another negative image according to another environment setting value; FIG. 4 shows a schematic diagram of a clear thermal image combined with a negative film image according to the embodiment of FIG. 2; FIG. 5A shows a visible light image of the solar photovoltaic module according to the embodiment of FIG. 1; Fig. 5B shows a thermal image of the solar photovoltaic module according to the embodiment of Fig. 5A; FIG. 6A shows another visible light image of the solar photovoltaic module according to the embodiment of FIG. 1; FIG. 6B shows a thermal image of the solar photovoltaic module according to the embodiment of FIG. 6A; FIG. 7 illustrates a flow chart of steps of a method for detecting a solar photovoltaic module according to an embodiment of the present disclosure; and FIG. 8 is a detailed step diagram of the solar photovoltaic module detection method according to the embodiment of FIG. 7 .

100: Solar Photoelectric Module Detection System

110: Drone

111: Visible light lens

112: Infrared lens

120: Process Terminal

130: Electronic Devices

140: Solar photovoltaic module

S1: Status signal

Claims (10)

A solar photovoltaic module detection system, comprising: 1 UAV, including: at least one lens, capturing at least one solar photovoltaic module and outputting at least one image; and A processing terminal, the signal is connected to the drone, including: a clear image module, receiving the image and outputting a clear image; an image black level module, the signal is connected to the clear image module, receives the clear image and outputs a black level image; an image negative film module, the signal is connected to the image black level module, receives the black level image and outputs a negative film image according to an environment setting value; and a state detection module, the signal is connected to the image black level module and the image negative film module, receives the black level image and the negative film image and outputs a complex number of features, wherein each of the features includes a field value, and each of the features corresponds to a complex number The states respectively have complex classical domain ranges; Wherein, according to an extension algorithm, the distances between the node domain values and the classical domain ranges correspond to each of the states to output a correlation function value respectively, and output a state signal according to the largest of the correlation function values. The photovoltaic module detection system of claim 1, wherein the number of the at least one lens is two, the two lenses are a visible light lens and an infrared lens respectively, and the at least one image includes a visible light image and a thermal image . The solar photoelectric module detection system of claim 1, wherein the drone photographs the at least one solar photoelectric module according to a planned path. The solar photovoltaic module detection system as described in claim 1, further comprising: An electronic device, the signal is connected to the processing terminal and receives the status signal. A solar photovoltaic module detection method, comprising: a photographing step of photographing at least one solar photovoltaic module through a drone and outputting at least one image to a processing terminal; An image processing step, including: an image clearing step, outputting a clear image through an image clearing module of the processing terminal according to the image; an image black leveling step, outputting a black level image through an image black level module of the processing terminal according to the clear image; an image negative filming step, outputting a negative film image through an image negative film module of the processing terminal according to the black level image; and A feature output step is to output complex features through a state detection module of the processing terminal according to the black-level image and the negative image, wherein each feature includes a field value, and each feature corresponding to the complex state has a complex classical field respectively scope; and A state detection step is to output a correlation function value according to the distance between the node domain values and the classical domain ranges corresponding to each of the states according to an extension algorithm, and output a correlation function value according to the largest one of the correlation function values. status signal. The solar photovoltaic module detection method as claimed in claim 5, wherein the state detection step comprises: A weight value is set for each of the features, and the distances between the node domain values and the classical domain ranges correspond to the weight values of the states to output the correlation function values respectively. The solar photovoltaic module detection method as claimed in claim 5, wherein the image negative filming step comprises: The image negative film module sets an environmental setting value according to an environmental temperature, and outputs the negative film image according to the environmental setting value. The solar photovoltaic module detection method as claimed in claim 5, wherein the feature output step comprises: The features include a heat-generating feature, and the state detection module outputs a heat-generating spot image according to the negative film image, and the heat-generating feature corresponds to the heat-generating spot image. The solar photovoltaic module detection method as described in claim 5, further comprising: In a path planning step, the drone shoots the at least one solar photovoltaic module according to a planned path. The solar photovoltaic module detection method as described in claim 5, further comprising: In a status reporting step, when the status signal is a fault signal, the processing terminal transmits the fault signal to an electronic device.

Images