Hanging rail type photovoltaic panel cleaning and detecting integrated device and method

Abstract

The invention relates to the technical field of photovoltaic panel cleaning and detection, in particular to a rail hanging type photovoltaic panel cleaning and detection integrated device and method. The device comprises a rack as well as a power supply module, a walking assembly, a cleaning assembly, an infrared detection assembly, an approach detection module, a communication module and a main controller which are arranged on the rack, the device can complete photovoltaic panel cleaning and infrared detection in one stroke, the infrared detection speed can be automatically adjusted according to the walking speed of the device, manual intervention is not needed, and the working efficiency is improved. A detection result can be wirelessly sent to a remote control terminal through the communication module, the infrared detection method determines a hot spot gray threshold value by calculating a gray probability density distribution relative change ratio of two images shot by the infrared camera successively at different moments, and performs hot spot segmentation and recognition on the images according to the threshold value, so that the detection accuracy is improved. According to the device, the operation and maintenance cost of the photovoltaic panel can be reduced, the defects of the photovoltaic panel can be found in time, and the intelligent level of operation and maintenance of the photovoltaic panel is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic panel cleaning and detection, and in particular to a rail-mounted photovoltaic panel cleaning and detection integrated device and method. Background Art

[0002] Traditional fossil energy (coal, oil) faces a series of problems such as environmental pollution and raw material depletion, while emerging energy such as solar energy will occupy an increasingly larger proportion in the energy structure due to its green, environmentally friendly and renewable advantages. my country has a vast territory, and some areas have long sunshine hours and high light intensity, which have unique natural conditions for the development of solar energy. Therefore, solar energy has developed rapidly in my country in recent years.

[0003] However, in addition to the performance of the photovoltaic modules themselves, the natural environment will also have a great impact on the power generation efficiency of the photovoltaic power generation system during its operation. The most important are the dust and particulate matter in the air, which accumulate on the surface of the photovoltaic modules and affect the light absorption of the photovoltaic modules, reducing the output power of the photovoltaic modules. At the same time, during the operation and maintenance of the photovoltaic power generation system, various forms and degrees of failure problems may occur due to differences in product quality, construction standards and environmental factors. For example, the mismatch of individual photovoltaic cells, cracks, local dirt on the surface of the module (such as bird droppings) or shading will cause hot spots on the photovoltaic modules, resulting in a decrease in the power generation of the photovoltaic modules and even damage to the photovoltaic panels. The efficiency improvement and intelligence of the current photovoltaic system are the main difficulties and challenges facing the safe, reliable and economical operation of the photovoltaic power generation system.

[0004] The patent application number is 202411250641.2, which discloses a fully automated photovoltaic cleaning robot and a method of using the same. Through the coordinated use of cleaning components, processing components, and changing components, the robot can work according to different cleaning conditions and effectively clean two types of photovoltaic panels at the same time. The patent application number is 202410674090.6, which discloses an intelligent photovoltaic inspection device and method based on drone inspection. By planning the inspection path and using drones to collect images of photovoltaic components, the defects of the prior art such as easy omission of photovoltaic components, repeated inspections, and low accuracy of photovoltaic fault detection can be effectively overcome. The patent application number is 202310833601.X, which discloses an infrared detection device for photovoltaic panel monitoring. By using infrared imaging as the first detection component, the state of the surface of the photovoltaic panel components is monitored in real time. When an abnormal situation is found, the main component can be driven to move by controlling the first driving component, and the photovoltaic panel components can be cleaned in time by the flushing component to achieve the characteristics of automatic cleaning.

[0005] It is true that a lot of work has been done in the field of photovoltaic panel cleaning and inspection at home and abroad, and great progress and remarkable practical effects have been achieved. However, there are few integrated devices that combine cleaning and inspection at the same time. The main difficulty is that the cleaning speed is not compatible with the inspection speed, and there is a lack of efficient and fast detection algorithms. This has caused the dilemma of time-sharing and equipment-sharing operation of photovoltaic cleaning and inspection, which has invisibly increased the energy consumption and other maintenance costs of the maintenance process. In view of this, we proposed a rail-mounted photovoltaic panel cleaning and inspection integrated device and method. Summary of the invention

[0006] The present invention provides a rail-mounted photovoltaic panel cleaning and detection integrated device and method, which aims to complete the cleaning and infrared detection of the photovoltaic panel body in one trip, reduce the operation and maintenance cost of the photovoltaic panel body, timely discover the defects of the photovoltaic panel body, and improve the intelligence level of the operation and maintenance of the photovoltaic panel body.

[0007] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0008] A rail-mounted photovoltaic panel cleaning and detection integrated device comprises a frame, a control box is fixedly mounted on the frame, and a power supply module, a walking component, a cleaning component, an infrared detection component, a proximity detection module, a communication module and a main controller are arranged on the frame, wherein:

[0009] The power supply module comprises a power supply photovoltaic panel, a battery controller and a lithium battery connected in sequence, and the power supply photovoltaic panel stores electric energy in the lithium battery through the battery controller;

[0010] The walking assembly includes a stepper motor, a driving wheel, a driven wheel and a limiting wheel connected to the frame, and is used to drive the entire device to move on the surface of the photovoltaic panel body;

[0011] The cleaning assembly includes a DC motor and a roller brush connected in sequence, wherein the DC motor is used to drive the roller brush to rotate, thereby cleaning the surface of the photovoltaic panel body;

[0012] The infrared detection component includes an infrared pulse light source, an infrared camera and a signal processor, which are used to capture an infrared image of the photovoltaic panel body and process the image to identify defects such as hot spots in the photovoltaic panel body.

[0013] Preferably, the infrared pulse light source of the infrared detection component is a long strip-shaped controllable infrared pulse flash light source, which is arranged inside the frame, and its two ends are connected to the inner end surfaces of the frame.

[0014] Preferably, the infrared camera of the infrared detection assembly is one or more cameras fixed to the outside of the frame, and the number of the cameras is determined by the width of the photovoltaic panel body and the field of view that the infrared camera can capture.

[0015] Preferably, the image shooting speed of infrared detection can be automatically adjusted according to the walking speed of the device, and the shooting interval of the infrared detection image is calculated according to the following formula:

[0016]

[0017] Wherein, L is the shooting width of a single image of the photovoltaic panel body in the walking direction of the device, v is the walking speed of the device, r is the shooting time factor, and 0≤r≤1.

[0018] An infrared detection method for a rail-mounted photovoltaic panel cleaning and detection integrated device, which completes a single photovoltaic panel body infrared detection, specifically includes the following steps:

[0019] S1, stimulating the pulse light source according to the shooting interval, and taking an infrared image of the photovoltaic panel body each time the light source is stimulated;

[0020] S2, obtaining the two most recent images at different times, namely, the image Img_a taken at the current time and the image Img_b taken at the previous time;

[0021] S3, performing image preprocessing on the two images;

[0022] S4, obtaining the grayscale probability distribution of the two images;

[0023] S5, calculating the relative change ratio of grayscale probability distribution and determining the hot spot segmentation threshold;

[0024] S6. Segment and mark the hot spots of the image according to the hot spot segmentation threshold;

[0025] S7. Determine whether there is a hot spot, and if so, record the hot spot.

[0026] Preferably, the image preprocessing comprises the following steps:

[0027] S31, image segmentation: segment the overlapping area of ​​the images taken at the current moment and the previous moment, and the overlapping area image taken at the previous moment is M b , the overlapping area image taken at the current moment is M a ;

[0028] S32, image grayscale: assign different weights to the three colors of red (R), green (G), and blue (B) in the two overlapping color images, and then calculate the weighted average as the grayscale value of the pixel point. Calculate the grayscale value of the two images according to the following formula:

[0029] Gray(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y)

[0030] After the image is grayed out, the image M a The corresponding grayscale image is G a , image M b The corresponding grayscale image is G b ;

[0031] S33, grayscale image filtering: set a 3×3 template K, mark the template coefficients from the center to the periphery as K0 to K8, and then overlap the grayscale value distribution map G of the gray image with the corresponding pixel points of the template K.

[0032]

[0033] The output of the template is calculated as follows:

[0034]

[0035] The corresponding output F value of the template is assigned to the corresponding image pixel in the image to replace the gray value of the pixel at the center point (x, y). This calculation is performed on each pixel in the original image, and the resulting image is the filtered image. After the grayscale image is filtered, the image G a The corresponding filtered image is F a , image G b The corresponding filtered image is F b .

[0036] Preferably, the step of obtaining the grayscale probability distribution of the two images comprises the following steps:

[0037] S41, obtaining the grayscale histogram of the image pixels: taking the direction of increasing change of the pixel grayscale value as the positive direction of the horizontal axis and marking the frequency of occurrence of the pixel grayscale value with H in the vertical direction, a grayscale histogram is established:

[0038] H=p(i)

[0039] Where i represents the grayscale value in the 8-bit grayscale image, H represents the number of pixels corresponding to each grayscale, and the grayscale histogram of the image pixels is obtained. Image F a The corresponding grayscale histogram H a , image F b The corresponding grayscale histogram is H b ;

[0040] S42, grayscale histogram fitting: performing curve fitting on the grayscale histogram to obtain a fitted grayscale histogram curve, and grayscale histogram H a The fitted curve is S a , grayscale histogram H b The fitted curve is S b .

[0041] Preferably, the step of calculating the relative change ratio of grayscale probability distribution and determining the hot spot segmentation threshold comprises the following steps:

[0042] S51. Calculate the relative change ratio of grayscale probability distribution: Calculate the relative change ratio of grayscale probability distribution S according to the following formula: r :

[0043]

[0044] S52, determine the hot spot segmentation threshold: according to the relative change ratio S of the gray probability distribution r Curve, select the gray value corresponding to the trough to the left of the rightmost peak as the segmentation threshold.

[0045] Preferably, the hot spot segmentation and judgment method is to mark pixels in the grayscale image that are higher than the segmentation threshold as hot spot pixels, and further judge the connectivity of the hot spot pixels, and mark the corresponding area where the number of connected hot spot pixels exceeds a certain value Q as a hot spot.

[0046] Preferably, the hot spot recording records the characteristics of the area marked as a hot spot, including the size of the hot spot, the location of the device, and the detection time information, and sends them to the main controller to complete the hot spot recording.

[0047] Compared with the prior art, the present invention provides a rail-mounted photovoltaic panel cleaning and detection integrated device and method, which has the following beneficial effects:

[0048] 1. The integrated cleaning and detection device and method for the rail-mounted photovoltaic panel, in order to achieve efficient energy saving and intelligent monitoring of the operation and maintenance of the photovoltaic panel, coordinates the travel component, cleaning component, infrared detection component and proximity detection module in the device to simultaneously complete the cleaning and infrared detection work in the same trip, which not only reduces the number of round trips of the equipment and reduces the energy consumption during the operation and maintenance process, but also can detect the defects of the photovoltaic panel in time, effectively reducing the operation and maintenance cost of the photovoltaic panel, greatly improving the intelligent level of the operation and maintenance of the photovoltaic panel, and ensuring the stable operation and power generation efficiency of the photovoltaic panel.

[0049] 2. The integrated cleaning and detection device and method for the hanging rail photovoltaic panel, in order to optimize the economy and practicality of the operation and maintenance device, integrates the cleaning and detection functions of the photovoltaic panel with the help of the integrated device, and the various components are reasonably arranged on the rack, such as the power supply module to ensure energy supply, the walking component drives the overall movement, the cleaning component is responsible for surface cleaning, and the infrared detection component accurately detects defects such as hot spots. This integrated design reduces the redundancy of parts and reduces the manufacturing cost of the operation and maintenance device, making it simple, efficient, reliable and practical, and convenient for large-scale promotion and application.

[0050] 3. The integrated device and method for cleaning and detecting rail-mounted photovoltaic panels, in order to accurately identify hot spot defects of photovoltaic panels and improve the scientific nature of detection, uses the infrared pulse light source, infrared camera and signal processor of the infrared detection component to carry out work, automatically adjusts the image shooting speed according to the walking speed of the device, calculates the shooting interval and shoots images according to a specific formula, and then obtains the grayscale probability distribution of the image and calculates the relative change ratio through preprocessing steps such as image segmentation, graying, and filtering, and then determines the hot spot segmentation threshold, and performs hot spot segmentation, marking and judgment based on this threshold. Finally, the relevant characteristic information of the hot spot is recorded. This series of processes ensures the accuracy and efficiency of hot spot detection, provides reliable data support for the maintenance and fault prevention of photovoltaic panels, and helps to extend the service life of photovoltaic panels and improve the stability of the overall power generation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of the structure of the rail-mounted photovoltaic panel cleaning and detection integrated device of the present invention;

[0052] Figure 2 This is a schematic diagram of the external structure of the rail-mounted photovoltaic panel cleaning and detection integrated device of the present invention;

[0053] Figure 3 It is a schematic diagram of the internal structure of the rail-mounted photovoltaic panel cleaning and detection integrated device of the present invention;

[0054] Figure 4 It is an infrared detection flow chart of the integrated cleaning and detection device for the rail-mounted photovoltaic panel of the present invention;

[0055] Figure 5 It is a schematic diagram of the overlapping area of ​​two infrared images of the present invention;

[0056] Figure 6 It is a schematic diagram of the infrared detection grayscale probability distribution signal and relative change ratio signal of the present invention.

[0057] In the figure: 10, frame; 20, main controller; 30, power supply module; 31, power supply photovoltaic panel; 32, battery controller; 33, lithium battery; 40, walking component; 41, stepping motor; 42, driving wheel; 43, driven wheel; 44, limiting wheel; 50, cleaning component; 51, DC motor; 52, roller brush; 60, infrared detection component; 61, infrared pulse light source; 62, infrared camera; 63, signal processor; 70, proximity detection module; 80, communication module; 90, control box. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the invention more clear, the invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described here are only used to explain the invention and do not constitute a limitation of the invention.

[0059] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0060] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0061] like Figure 1 As shown, a rail-mounted photovoltaic panel cleaning and detection integrated device includes a frame 10, a control box 90 is fixedly installed on the frame 10, and a power supply module 30, a walking component 40, a cleaning component 50, an infrared detection component 60, a proximity detection module 70, a communication module 80 and a main controller 20 are arranged on the frame 10, wherein:

[0062] The power supply module 30 includes a power supply photovoltaic panel 31, a battery controller 32 and a lithium battery 33 which are connected in sequence. The power supply photovoltaic panel 31 stores electric energy in the lithium battery 33 via the battery controller 32;

[0063] The walking assembly 40 includes a stepper motor 41 fixedly connected to the frame 10, a driving wheel 42 is fixedly installed on the outer side of the output end of the stepper motor 41, the driving wheel 42 is connected to the driven wheel 43 through a roller shaft, the driving wheel 42 and the driven wheel 43 are rolled and fitted above the photovoltaic panel body, wherein a limiting wheel 44 is rotatably installed on the frame 10, and the limiting wheel 44 is rolled and fitted on the outer side of the photovoltaic panel body, thereby driving the frame 10 to move on the surface of the photovoltaic panel body;

[0064] The cleaning assembly 50 includes a DC motor 51 and a roller brush 52 connected in sequence. The DC motor 51 is fixedly installed on the inner side of the frame 10, and the roller brush 52 is clamped on the outer side of the output shaft of the DC motor 51. The DC motor 51 is started to drive the roller brush 52 to rotate, thereby cleaning the surface of the photovoltaic panel body;

[0065] The infrared detection component 60 includes an infrared pulse light source 61, an infrared camera 62 and a signal processor 63. The infrared camera 62 is fixedly installed on the outside of the frame 10, and is used to capture infrared images of the photovoltaic panel body and process the image to identify defects such as hot spots in the photovoltaic panel body.

[0066] like Figure 3 As shown, the infrared pulse light source 61 of the infrared detection component 60 is a long strip-shaped controllable infrared pulse flash light source, which is arranged inside the frame 10 and has two ends connected to the inner end surfaces of the frame 10 .

[0067] like Figure 2 As shown, the infrared camera 62 of the infrared detection assembly 60 is one or more cameras fixed to the outside of the frame 10 , and the number of cameras is determined by the width of the photovoltaic panel body and the field of view that the infrared camera 62 can capture.

[0068] The image shooting speed of infrared detection can be automatically adjusted according to the walking speed of the device. The shooting interval of infrared detection images is calculated according to the following formula:

[0069]

[0070] Wherein, L is the shooting width of a single image of the photovoltaic panel body in the walking direction of the device, v is the walking speed of the device, r is the shooting time factor, and 0≤r≤1.

[0071] like Figure 4 As shown, an infrared detection method of a rail-mounted photovoltaic panel cleaning and detection integrated device is used to complete a single infrared detection of the photovoltaic panel body, specifically including the following steps:

[0072] S1, stimulating the pulse light source according to the shooting interval, and taking an infrared image of the photovoltaic panel body each time the light source is stimulated;

[0073] S2, obtaining the two most recent images at different times, namely, the image Img_a taken at the current time and the image Img_b taken at the previous time;

[0074] S3, performing image preprocessing on the two images;

[0075] S4, obtaining the grayscale probability distribution of the two images;

[0076] S5, calculating the relative change ratio of grayscale probability distribution and determining the hot spot segmentation threshold;

[0077] S6. Segment and mark the hot spots of the image according to the hot spot segmentation threshold;

[0078] S7. Determine whether there is a hot spot, and if so, record the hot spot.

[0079] Furthermore, the image preprocessing S3 comprises the following steps:

[0080] S31, image segmentation: segment the overlapping area of ​​the images taken at the current moment and the previous moment, and the overlapping area image taken at the previous moment is M b , the overlapping area image taken at the current moment is M a In one embodiment, preferably, r = 0.5, then 50% of the overlapping area, such as Figure 5 As shown, at this time, the overlapping area image M b is the first image taken after the infrared pulse light source 61 is irradiated, and the overlapping area image Mb The second image is taken after being irradiated by the infrared pulse light source 61;

[0081] S32, image grayscale: assign different weights to the three colors of red (R), green (G), and blue (B) in the two overlapping color images, and then calculate the weighted average as the grayscale value of the pixel point. Calculate the grayscale value of the two images according to the following formula:

[0082] Gray(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y)

[0083] After the image is grayed out, the image M a The corresponding grayscale image is G a , image M b The corresponding grayscale image is G b ;

[0084] S33, grayscale image filtering: set a 3×3 template K, mark the template coefficients from the center to the periphery as K0 to K8, and then overlap the grayscale value distribution map G of the gray image with the corresponding pixel points of the template K.

[0085]

[0086] The output of the template is calculated as follows:

[0087]

[0088] The corresponding output F value of the template is assigned to the corresponding image pixel in the image to replace the gray value of the pixel at the center point (x, y). This calculation is performed on each pixel in the original image, and the resulting image is the filtered image. After the grayscale image is filtered, the image G a The corresponding filtered image is F a , image G b The corresponding filtered image is F b , in one embodiment, Preferably, K0=0.46, K1=K3=K5=K7=0.09, K2=K4=K6=K8=0.045.

[0089] Furthermore, obtaining the grayscale probability distribution S4 of the two images includes the following steps:

[0090] S41, obtaining the grayscale histogram of the image pixels: taking the direction of increasing change of the pixel grayscale value as the positive direction of the horizontal axis and marking the frequency of occurrence of the pixel grayscale value with H in the vertical direction, a grayscale histogram is established:

[0091] H=p(i)

[0092] Where i represents the grayscale value in the 8-bit grayscale image, H represents the number of pixels corresponding to each grayscale, and the grayscale histogram of the image pixels is obtained. Image F a The corresponding grayscale histogram H a , image F b The corresponding grayscale histogram is H b ;

[0093] S42, grayscale histogram fitting: performing curve fitting on the grayscale histogram to obtain a fitted grayscale histogram curve, and grayscale histogram H a The fitted curve is S a , grayscale histogram H b The fitted curve is S b In the present invention, there are multiple schemes for histogram curve fitting. In one embodiment, the fitting method is the B-spline least squares fitting method; in another embodiment, the fitting method is the least squares cubic curve fitting method.

[0094] Further, calculating the relative change ratio of the grayscale probability distribution and determining the hot spot segmentation threshold S5 includes the following steps:

[0095] S51. Calculate the relative change ratio of grayscale probability distribution: Calculate the relative change ratio of grayscale probability distribution S according to the following formula: r :

[0096]

[0097] S52, determine the hot spot segmentation threshold: according to the relative change ratio S of the gray probability distribution r Curve, select the gray value corresponding to the trough to the left of the rightmost peak as the segmentation threshold.

[0098] Furthermore, if Figure 6 As shown, the hot spot segmentation and judgment method S6 is to mark the pixels in the grayscale image that are higher than the segmentation threshold as hot spot pixels, and further judge the connectivity of the hot spot pixels, and mark the corresponding area where the number of connected hot spot pixels exceeds a certain value Q as a hot spot. In one embodiment, the hot spot pixel threshold Q=5.

[0099] Furthermore, the hot spot recording S7 is to record the characteristics of the area marked as a hot spot, including the size of the hot spot, the location of the device, the detection time and other information, and send it to the main controller 20 to complete the hot spot recording.

[0100] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure of the present invention can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0101] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A rail-mounted photovoltaic panel cleaning and detection integrated device, characterized in that: The invention comprises a frame (10), a control box (90) being fixedly mounted on the frame (10), and a power supply module (30), a walking assembly (40), a cleaning assembly (50), an infrared detection assembly (60), a proximity detection module (70), a communication module (80) and a main controller (20) arranged on the frame (10), wherein: The power supply module (30) comprises a power supply photovoltaic panel (31), a battery controller (32) and a lithium battery (33) which are connected in sequence, and the power supply photovoltaic panel (31) stores electric energy in the lithium battery (33) via the battery controller (32); The walking assembly (40) comprises a stepping motor (41), a driving wheel (42), a driven wheel (43) and a limiting wheel (44) connected to the frame (10); The cleaning assembly (50) comprises a DC motor (51) and a roller brush (52) which are connected in sequence; The infrared detection component (60) comprises an infrared pulse light source (61), an infrared camera (62) and a signal processor (63), and is used to capture an infrared image of the photovoltaic panel body and process the image to identify hot spot defects existing in the photovoltaic panel body.

2. The rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 1, characterized in that: The infrared pulse light source (61) of the infrared detection component (60) is a long strip-shaped controllable infrared pulse flash light source, which is arranged inside the frame (10) and has two ends connected to the inner end surfaces of the frame (10).

3. The rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 1, characterized in that: The infrared camera (62) of the infrared detection assembly (60) is one or more cameras fixed to the outside of the frame (10), and the number of the cameras is determined by the width of the photovoltaic panel body and the field of view that the infrared camera (62) can capture.

4. The rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 1, characterized in that: The image shooting speed of infrared detection can be automatically adjusted according to the walking speed of the device. The shooting interval of infrared detection images is calculated according to the following formula: Wherein, L is the shooting width of a single image of the photovoltaic panel body in the walking direction of the device, v is the walking speed of the device, r is the shooting time factor, and 0≤r≤1.

5. An infrared detection method for a rail-mounted photovoltaic panel cleaning and detection integrated device, applicable to the rail-mounted photovoltaic panel cleaning and detection integrated device as claimed in any one of claims 1 to 4, characterized in that: Complete a single infrared detection of the photovoltaic panel body, including the following steps: S1, stimulating the pulse light source according to the shooting interval, and taking an infrared image of the photovoltaic panel body each time the light source is stimulated; S2, obtaining the two most recent images at different times, namely, the image Img_a taken at the current time and the image Img_b taken at the previous time; S3, performing image preprocessing on the two images; S4, obtaining the grayscale probability distribution of the two images; S5, calculating the relative change ratio of grayscale probability distribution and determining the hot spot segmentation threshold; S6. Segment and mark the hot spots of the image according to the hot spot segmentation threshold; S7. Determine whether there is a hot spot, and if so, record the hot spot.

6. The infrared detection method of the rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 5, characterized in that: The image preprocessing comprises the following steps: S31, image segmentation: segment the overlapping area of ​​the images taken at the current moment and the previous moment, and the overlapping area image taken at the previous moment is M b , the overlapping area image taken at the current moment is M a ; S32, image grayscale: assign different weights to the three colors of red (R), green (G), and blue (B) in the two overlapping color images, and then calculate the weighted average as the grayscale value of the pixel point. Calculate the grayscale value of the two images according to the following formula: Gray(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y) After the image is grayed out, the image M a The corresponding grayscale image is G a , image M b The corresponding grayscale image is G b ; S33, grayscale image filtering: set a 3×3 template K, mark the template coefficients from the center to the periphery as K0 to K8, and then overlap the grayscale value distribution map G of the gray image with the corresponding pixel points of the template K. The output of the template is calculated as follows: The corresponding output F value of the template is assigned to the corresponding image pixel in the image to replace the gray value of the pixel at the center point (x, y). This calculation is performed on each pixel in the original image, and the resulting image is the filtered image. After the grayscale image is filtered, the image G a The corresponding filtered image is F a , image G b The corresponding filtered image is F b .

7. The infrared detection method of the rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 5, characterized in that: The step of obtaining the grayscale probability distribution of the two images comprises the following steps: S41, obtaining the grayscale histogram of the image pixels: taking the direction of increasing change of the pixel grayscale value as the positive direction of the horizontal axis and marking the frequency of occurrence of the pixel grayscale value with H in the vertical direction, a grayscale histogram is established: H=p(i) Where i represents the grayscale value in the 8-bit grayscale image, H represents the number of pixels corresponding to each grayscale, and the grayscale histogram of the image pixels is obtained. Image F a The corresponding grayscale histogram H a , image F b The corresponding grayscale histogram is H b ; S42, grayscale histogram fitting: performing curve fitting on the grayscale histogram to obtain a fitted grayscale histogram curve, and grayscale histogram H a The fitted curve is S a , grayscale histogram H b The fitted curve is S b .

8. The infrared detection method of the rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 5, characterized in that: The step of calculating the relative change ratio of grayscale probability distribution and determining the hot spot segmentation threshold comprises the following steps: S51. Calculate the relative change ratio of grayscale probability distribution: Calculate the relative change ratio of grayscale probability distribution S according to the following formula: r : S52, determine the hot spot segmentation threshold: according to the relative change ratio S of the gray probability distribution r Curve, select the gray value corresponding to the trough to the left of the rightmost peak as the segmentation threshold.

9. The infrared detection method of the rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 5, characterized in that: The hot spot segmentation and judgment method is to mark pixels in the grayscale image that are higher than the segmentation threshold as hot spot pixels, and further judge the connectivity of the hot spot pixels, and mark the corresponding area where the number of connected hot spot pixels exceeds a certain value Q as a hot spot.

10. The infrared detection method of the rail-mounted photovoltaic panel cleaning and detection integrated device according to claim 5, characterized in that: The hot spot record records the characteristics of the area marked as a hot spot, including the size of the hot spot, the location of the device, and the detection time information, and sends them to the main controller (20) to complete the hot spot record.

Images