A small groove distance main roller groove type roller cleaning method and system based on automation control

Abstract

The application discloses a kind of small groove distance main roller groove shape roller cleaning method and system based on automation control, comprising the following steps: copper brush that meets parameter requirement is fixed on copper brush tooling;With the head of main roller as origin, establish a right-hand Cartesian coordinate system, and obtain copper brush bristle width;According to right-hand Cartesian coordinate system and copper brush bristle width, the path of copper brush brush main roller is designed, and a main roller cleaning path is obtained;According to the preset parameter, start main roller, and according to main roller cleaning path, call copper brush to clean main roller groove shape roller;Video frame acquisition is carried out to main roller by acquisition device, and main roller groove shape roller image is obtained, and whether there is still foreign matter in main roller groove shape roller is identified by main roller groove shape roller image, if not, the cleaning of main roller groove shape roller is completed, if yes, according to the position of foreign matter, call copper brush to clean main roller groove shape roller again.The application achieves the purpose of reducing the slice failure rate, and prolonging the service life of main roller.

Description

Technical Field

[0001] This invention relates to the field of silicon wafer cutting technology, specifically to a method and system for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control. Background Technology

[0002] With the continuous expansion of my country's industrial scale and the increasingly tight energy supply, photovoltaic power generation, as a pollution-free and renewable resource, has developed exceptionally rapidly. The main raw materials for manufacturing solar cells include crystalline silicon, amorphous silicon, and crystalline silicon for composite solar cells. Solar-grade crystalline silicon wafers are generally manufactured using wire cutting technology. Multi-wire cutting (MWD), or wire sawing for short, is suitable for mass-producing large-size, ultra-thin silicon wafers.

[0003] The cutting principle of a multi-wire cutting machine is as follows: using a high-speed moving steel wire as a carrier and SiC particles (slurry) as the cutting blade, the silicon ingot is cut by the relative movement between the silicon ingot and the steel wire to obtain silicon wafers of a certain geometric thickness. Specifically, the cutting steel wire wound on the main roller moves horizontally at a certain cutting speed, carrying the slurry, while the silicon ingot moves vertically at a certain feed speed, thus achieving the cutting.

[0004] As silicon wafers move towards larger and thinner sizes, the groove pitch of the main roller trough conveyor is also decreasing. However, small-pitch main roller trough conveyors present the following problems: (1) The chip removal process of the trough roller with small groove pitch is more difficult. Foreign objects are easy to remain at the bottom of the groove, affecting the wiring success rate and the flatness of the wire mesh. This leads to the easy occurrence of wire skipping and wire breakage during the silicon wafer cutting process, which increases the slicing defect rate and also shortens the service life of the main roller.

[0005] (2) Currently, the main roller is cleaned by manual rinsing for foreign objects (including long strips of rubber filaments and small particles) in the trough-shaped roller table with small groove pitch. However, manual rinsing has problems such as low rinsing efficiency, high requirements for operation techniques, waste of water resources, waste of working hours and excessive humidity, and poor cleaning effect. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention provides a method and system for cleaning small-pitch main roller troughs based on automated control. This method addresses the technical problems of existing methods for cleaning small-pitch main roller troughs, such as poor cleaning effect, waste of water resources and labor time, and excessively humid environment. The goal is to reduce the defect rate of slices and extend the service life of the main roller.

[0007] To solve the above problems, the technical solution adopted by the present invention is as follows: A method for cleaning a small-pitch main roller trough conveyor based on automated control includes the following steps: Provide a copper brush that meets the parameter requirements, and fix the copper brush that meets the parameter requirements on a copper brush fixture; Establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and obtain the width of the copper brush bristles; Based on the right-hand Cartesian coordinate system and the width of the copper brush bristles, the path of the copper brush on the main roller is designed to obtain a main roller cleaning path. The main roller is started according to preset parameters, and the copper brush is called to clean the trough-shaped roller track of the main roller according to the main roller cleaning path; The main roller is captured by a data acquisition device to obtain a trough-shaped roller conveyor image. The image is then used to identify whether there are any foreign objects on the surface of the trough-shaped roller conveyor. If no foreign objects are found, the cleaning of the trough-shaped roller conveyor is completed. If foreign objects are found, the copper brush is used to clean the trough-shaped roller conveyor again based on the location of the foreign objects.

[0008] In a preferred embodiment of the present invention, when providing a copper brush that meets the parameter requirements, the following are included: We provide copper brushes with bristle diameters of 0.076 mm. When the main roller is started according to preset parameters, the following is included: The main roller is started at a speed of 240 revolutions per minute.

[0009] In a preferred embodiment of the present invention, obtaining a main roller cleaning path includes: Obtain the coordinates of the head and tail of the main roller in the right-hand Cartesian coordinate system; The first approach position, the second approach position, and the third approach position are respectively set according to the coordinates of the head of the main roller; Based on the coordinates of the head and tail of the main roller, the distance the copper brush moves to the left after brushing the main roller for one revolution is set. The first approach position, the second approach position, and the third approach position are gradually approaching the head of the main roller; the overall direction of brushing the main roller is from head to tail and from right to left.

[0010] In a preferred embodiment of the present invention, when using copper brushes to clean the trough-shaped roller conveyor of the main roller, the method further includes: The copper brush is gradually brought closer to the head of the main roller, and when the copper brush is detected to have reached the first approach position, it pauses for the first time. When the first dwell time reaches the preset pause time, the copper brush continues to be gradually moved closer to the head of the main roller. When the copper brush is detected to have reached the second approach position, a second dwell is performed. When the second dwell time reaches the preset pause time, the copper brush continues to be gradually moved closer to the head of the main roller. When the copper brush is detected to have reached the third approach position, a third dwell is performed. When the third dwell time reaches the preset pause time, the copper brush continues to be used until it reaches the coordinates of the head of the main roller, and then the trough roller channel of the main roller is cleaned according to the preset parameters.

[0011] In a preferred embodiment of the present invention, cleaning the main roller trough-shaped roller conveyor according to preset parameters includes: According to the preset elasticity of the downward pressure copper brush, the copper brush is used to clean the trough-shaped roller channel of the main roller; Among them, when presetting the elasticity of the downward-pressing copper brush, the following are included: Obtain the depth of the trough-shaped roller channel of the main roller, and determine the preset amount of elasticity of the downward pressure copper brush based on the depth of the trough-shaped roller channel of the main roller.

[0012] In a preferred embodiment of the present invention, obtaining the image of the main roller trough includes: The main roller is captured by a camera. The main roller image frame data contained in the video frame is preprocessed to obtain a preprocessed main roller image frame. The preprocessed main roller image frame is converted into a grayscale image of the main roller. The main roller trough roller track is extracted from the grayscale image of the main roller to obtain the image of the main roller trough roller track. The preprocessing process includes mean filtering and pixel-based gradient-based image enhancement. Mean filtering includes: Let the target pixel be in the main roller image frame. The pixel value With the target pixel Select one as the center The window is used to calculate the sum of each pixel in the main roller image frame and all pixels in the surrounding template window, and an average value is obtained. The average value is used as the target pixel. The processed grayscale values ​​are shown in Formula 1: (1); In the formula, For the target pixel point Smoothed grayscale values The size of the template window. It is an odd number; When performing pixel-based gradient-based image enhancement, the following are included: Let the image after mean filtering be And the image size is 4096*2048. This indicates that the image after mean filtering is at point... The specific enhancement process for the pixel value at that location is as follows: The mean-filtered image is divided into 256*512 8*8 sub-grids, and the average value of each grid is obtained. As shown in Formula 2: (2); The difference between the midpoint pixel value and the grid mean in the mean-filtered image is obtained to determine the threshold for gradient enhancement. As shown in Formula 3: (3); In each grid cell, the pixel values ​​of the mean-filtered image are compared with the gradient enhancement threshold. Perform summation and limit the image pixel values ​​to 10 ... Between these, an enhanced image is obtained. As shown in Formula 4: (4); For an 8x8 image grid, move it up and to the right three times, moving two pixels each time. Calculate the mean value again within each different grid cell. Repeat this process to obtain the mean value for each grid cell. and gradient enhancement threshold Steps; Each move changes the number of grid cells. The final augmentation result is obtained by combining the seven images acquired from the grid cell movement.

[0013] In a preferred embodiment of the present invention, the process of extracting the trough-shaped roller track from the grayscale image of the main roller includes: use The operator obtains the gradient magnitude and direction of the grayscale image of the main roller, and applies non-maximum suppression to the gradient magnitude for edge refinement; A dual-threshold algorithm is used to detect and connect edges to obtain the image contour; The straight line segments present in the grayscale image of the main roller are detected by the probabilistic Hough transform algorithm and filtered to finally segment the trough-shaped roller track of the main roller.

[0014] In a preferred embodiment of the present invention, when identifying whether there are still foreign objects in the main roller trough-shaped roller conveyor through the image of the main roller trough-shaped roller conveyor, the following steps are included: Foreign objects in the trough-shaped roller conveyor of the main roller are detected using a background difference method with automatic background updating. The differential image of the main roll trough is shown in Equations 5 and 6: (5); (6); In the formula, Indicates the main roller image frame number. This indicates the update rate of the main roller image frames. Represents the pixels in the horizontal direction of the main roller image frame. Represents the pixels in the vertical direction of the main roller image frame. Function representing the background frame image. The function representing the current frame image. This represents the difference image function.

[0015] In a preferred embodiment of the present invention, after obtaining the differential image of the main roll trough, the method includes: Set the threshold for binarization processing Based on the threshold The differential image of the main roller trough is binarized to obtain the binarized image of the main roller trough, as shown in Formula 7: (7); The binarized image of the main roller grooved roller track is ANDed with the main roller grooved roller track extracted from the grayscale image of the main roller to obtain a preliminary image of the foreign object in the main roller grooved roller track. The preliminary image is processed using morphological filtering to obtain the final image of the foreign object in the main roller trough. Based on the final image of the foreign object in the main roller trough, the center distance of the foreign object in the main roller trough is extracted, and the coordinates of the centroid of the foreign object in the image are obtained according to the center distance. Furthermore, combined with the position of the acquisition device, the actual position of the foreign object in the main roller trough is obtained.

[0016] A small-pitch main roller trough cleaning system based on automated control includes: Parameter setting unit: used to establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and to obtain the width of the copper brush bristles; Path planning unit: used to design the path of the copper brush on the main roller according to the right-hand Cartesian coordinate system and the width of the copper brush bristles, so as to obtain a main roller cleaning path; Cleaning unit: used to start the main roller according to preset parameters, and call the copper brush to clean the trough roller track of the main roller according to the main roller cleaning path; Confirmation unit: Used to identify whether there are still foreign objects on the trough-shaped roller surface of the main roller through the image of the main roller. If not, the cleaning of the trough-shaped roller surface of the main roller is completed. If so, the copper brush is called to clean the trough-shaped roller surface of the main roller again according to the location of the foreign object.

[0017] The process involves providing a copper brush that meets the parameter requirements and fixing the copper brush on a copper brush fixture; and acquiring 2D video frames of the main roller using a data acquisition device to obtain an image of the main roller.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention establishes a right-handed Cartesian coordinate system by fixing a copper brush that meets the parameter requirements onto a copper brush fixture, and designs a path in this coordinate system based on the width of the copper brush bristles to obtain a main roller cleaning path. Based on this main roller cleaning path, the copper brush is automatically invoked to clean the trough-shaped roller track of the main roller, thereby effectively removing foreign matter (long strips of rubber fibers and small particles) remaining at the bottom of the trough-shaped roller track, solving the problem of smoothness of the main roller with small groove spacing. The long strips of rubber fibers, such as… Figure 1 As shown; (2) After the copper brush cleaning is completed, the present invention further confirms whether there are still foreign objects in the grooved roller channel of the main roller through image recognition. When foreign objects are identified, the actual position of the foreign objects is obtained, and the copper brush is called to clean the grooved roller channel of the main roller again, thereby further ensuring the smoothness of the small grooved main roller, avoiding the easy occurrence of skipping and breaking during the silicon wafer cutting process, effectively improving the slicing yield, and extending the service life of the main roller; (3) This invention effectively improves the problem of generating a large amount of water mist during the manual rinsing process, which affects the humidity of the grooving workshop, avoids the waste of action in the manual brushing of the main roller, and improves the efficiency of cleaning the main roller.

[0019] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the long strip-shaped adhesive filaments described in the section on the beneficial effects of this invention; Figure 2 This is a step diagram of the cleaning method for the small-pitch main roller trough roller conveyor based on automated control provided by the present invention; Figure 3 This is a photograph of a copper brush with a bristle diameter of 0.076 mm provided by the present invention. Figure 4 This is a photograph of a copper brush that meets the parameter requirements being fixed on a copper brush fixture in this invention. Figure 5 This is the main roller trough-shaped roller conveyor image segmented during the image extraction process of this invention. Detailed Implementation

[0021] A method for cleaning the trough-shaped roller conveyor with small trough pitch main rollers based on automated control, such as... Figure 2 As shown, it includes the following steps: Step S1: Provide a copper brush that meets the parameter requirements and fix the copper brush that meets the parameter requirements on the copper brush fixture; Step S2: Establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and obtain the width of the copper brush bristles; Step S3: Based on the right-hand Cartesian coordinate system and the width of the copper brush bristles, design the path of the copper brush on the main roller to obtain a main roller cleaning path. Step S4: Start the main roller according to the preset parameters, and call the copper brush to clean the trough roller track of the main roller according to the main roller cleaning path; Step S5: The main roller is captured by the acquisition device to obtain the image of the trough roller conveyor. The image of the trough roller conveyor is used to identify whether there are still foreign objects on the surface of the trough roller conveyor. If not, the cleaning of the trough roller conveyor is completed. If so, the copper brush is used to clean the trough roller conveyor again according to the location of the foreign object.

[0022] In step S1 above, providing a copper brush that meets the parameter requirements includes: We provide copper brushes with bristle diameters of 0.076 mm. When starting the main roller according to preset parameters, the following steps are included: Start the main roller at a speed of 240 revolutions per minute.

[0023] Specifically, a copper brush with bristle diameter of 0.076mm, such as... Figure 3 As shown. Fix the copper brush that meets the parameter requirements onto the copper brush fixture, as follows. Figure 4 As shown.

[0024] In step S3 above, when obtaining a main roller cleaning path, the following is included: Obtain the coordinates of the head and tail of the main roller in the right-hand Cartesian coordinate system; The first approach position, the second approach position, and the third approach position are set according to the coordinates of the head of the main roller. Based on the coordinates of the head and tail of the main roller, set the distance to the left after the copper brush has brushed the main roller for one revolution; Among them, the first approach position, the second approach position and the third approach position are gradually approaching the head of the main roller; the overall direction of the brush main roller is from head to tail, from right to left.

[0025] In step S4 above, when using copper brushes to clean the trough-shaped roller conveyor of the main roller, the following is also included: The copper brush is gradually brought closer to the head of the main roller. When the copper brush is detected to have reached the first approach position, it pauses for the first time. When the first dwell time reaches the preset pause time, the copper brush continues to gradually approach the head of the main roller. When the copper brush is detected to have reached the second approach position, a second dwell is performed. When the second dwell time reaches the preset pause time, the copper brush continues to gradually approach the head of the main roller. When the copper brush is detected to have reached the third approach position, a third dwell is performed. When the third dwell time reaches the preset pause time, the copper brush continues to be used until it reaches the coordinates of the head of the main roller, and then the trough roller conveyor of the main roller is cleaned according to the preset parameters.

[0026] Specifically, the three dwell times in this invention are set to check whether the main roller is operating normally multiple times during the process of the copper brush approaching, so as to avoid the copper brush colliding with the main roller due to abnormal operation of the main roller, thereby causing damage to the main roller.

[0027] Specifically, the automated programming logic of the present invention includes: The overall direction of the brush main roller, from beginning to end, from right to left, is the main body of the cycle from N40 to R106; R104 = 85 (width of copper brush bristles) R105=0 (parameter variable definition) R106 = R105 (Assigning values ​​to parameter variables) R107 = 1.25 (pause time in seconds) R108 = 0.3 (elasticity of the downward-pressing copper brush) R109 = R60 - R61 - R104 (depth of the main roller trough) R110 = R60 - R104 (brush cleaning distance) IF (R106>R110) GOTO N60 (Safety Protection) G26 S240 (Maximum Main Roller Speed) T08D8 (Device calls copper brush) When using the G97 M04 S240 copper brush, the main roller rotates at 240 revolutions per minute (4 revolutions per second). G94 G90 G00 X320 Z10 M08 (In a right-handed Cartesian coordinate system, the brush is closer) N40G01 Z=-R106 F8000 (In a right-handed Cartesian coordinate system, the brush is close to the head of the main roller) G01 X=R51+10 F6000 (In the right-hand Cartesian coordinate system, the brush is closer to the head of the main roller) G01 X=R51+6 F5000 (In the right-hand Cartesian coordinate system, the brush is closer to the head of the main roller) G01 X=R51-2*R108 F4000 (The brush presses down on the head of the main roller, and the brushing action begins) G04 F=R107 (The width of the brush is equal to the width of the bristles in one revolution of the main roller) G01 X=R51+10 F6000 (In the right-hand Cartesian coordinate system, the brush leaves the head of the main roller) R106 = R106 + R104 (parameter calculation, the copper brush moves to the left by one bristle width) STOPRE IF ( R106<=R109 ) GOTO N40 (Macro parameter calculation: if the entire length of the main roller is brushed, the process jumps to N60 after brushing the main roller; if the entire length of the main roller is not brushed, the process jumps to N40 and the brushing action continues to cycle.) N60 G01 X320 F8000 M09 Z10 F8000 M05 M30 Furthermore, when cleaning the main roller trough conveyor according to preset parameters, the process includes: Based on the preset elasticity of the downward pressure copper brush, the copper brush is used to clean the trough-shaped roller track of the main roller; Among them, when presetting the elasticity of the downward-pressing copper brush, the following are included: Obtain the depth of the trough-shaped roller path of the main roller, and determine the preset amount of elasticity of the downward pressure copper brush based on the depth of the trough-shaped roller path of the main roller.

[0028] In step S5 above, obtaining the image of the main roller trough includes: The main roller video frames are captured by the camera. The main roller image frame data contained in the video frames are preprocessed to obtain the preprocessed main roller image frames. The preprocessed main roller image frames are then converted into a grayscale image of the main roller. The main roller trough roller track is extracted from the grayscale image of the main roller to obtain the image of the main roller trough roller track. The preprocessing process includes mean filtering and pixel-based gradient-based image enhancement. Mean filtering includes: Let the target pixel be in the main roller image frame. The pixel value , with target pixel Select one as the center The window is used to calculate the sum of each pixel in the main roller image frame and all pixels in the surrounding template window, and an average value is obtained. Use the average value as the target pixel. The processed grayscale values ​​are shown in Formula 1: (1); In the formula, For target pixel Smoothed grayscale values The size of the template window. It is an odd number; When performing pixel-based gradient-based image enhancement, the following are included: Let the image after mean filtering be And the image size is 4096*2048. This indicates that the image after mean filtering is at point The specific enhancement process for the pixel value at that location is as follows: The mean-filtered image was divided into 256*512 8*8 sub-grids, and the average value of each grid was obtained. As shown in Formula 2: (2); The difference between the midpoint pixel value and the grid mean in the mean-filtered image is used to obtain the threshold for gradient enhancement. As shown in Formula 3: (3); In each grid cell, the pixel values ​​of the mean-filtered image are compared with the gradient enhancement threshold. Perform summation and limit the image pixel values ​​to 10 ... Between these, an enhanced image is obtained. As shown in Formula 4: (4); For an 8x8 image grid, move it up and to the right three times, moving two pixels each time. Calculate the mean value again within each different grid cell. Repeat this process to obtain the mean value for each grid cell. and gradient enhancement threshold Steps; Each move changes the number of grid cells. The final augmentation result is obtained by combining the seven images acquired from the grid cell movement.

[0029] Specifically, in this invention, mean filtering is used for denoising. Essentially, it uses neighboring pixels to distribute the impact of noise points, thereby relatively weakening the noise. The smoothing effect varies depending on the selected template. Generally, a larger template results in better denoising, but also greater blurring of image edges. In this invention, a 3×3 template is selected for filtering to achieve the best denoising effect.

[0030] Specifically, the image enhancement method described above can significantly improve the contrast of the image without increasing the noise present in the image, which is beneficial for the subsequent extraction of the main roller trough.

[0031] Furthermore, when extracting the trough-shaped roller track from the grayscale image of the main roller, the following steps are included: use The operator obtains the gradient magnitude and direction of the grayscale image of the main roller, and applies non-maximum suppression to the gradient magnitude for edge refinement; A dual-threshold algorithm is used to detect and connect edges to obtain the image contour; The probabilistic Hough transform algorithm is used to detect and filter straight line segments in the grayscale image of the main roller, ultimately segmenting the trough-shaped roller track of the main roller. Figure 3 As shown.

[0032] In step S5 above, when identifying whether there are still foreign objects inside the main roller trough roller conveyor through the image of the main roller trough roller conveyor, the following steps are included: Foreign objects in the trough-shaped roller conveyor of the main roll are detected using a background difference method with automatic background update, and the difference image of the trough-shaped roller conveyor of the main roll is obtained, as shown in Formulas 5 and 6: (5); (6); In the formula, Indicates the main roller image frame number. This indicates the update rate of the main roller image frames. Represents the pixels in the horizontal direction of the main roller image frame. Represents the pixels in the vertical direction of the main roller image frame. Function representing the background frame image. The function representing the current frame image. This represents the difference image function.

[0033] Specifically, when the update rate of the main roller image frame When the value is 0.3, the spatiotemporal correlation between pixels can be fully considered, and foreign objects that do not change significantly in the image sequence of the main roller trough can be effectively detected. At the same time, it can prevent moving targets from suddenly disappearing.

[0034] Further, after obtaining the differential image of the main roll trough, the process includes: Set the threshold for binarization processing Based on threshold Binarizing the differential image of the main roll trough is shown in Equation 7. (7); The binarized image of the main roller grooved roller track is ANDed with the main roller grooved roller track extracted from the grayscale image of the main roller to obtain a preliminary image of the foreign object in the main roller grooved roller track. The preliminary image was processed using morphological filtering to obtain the final image of the foreign object in the trough-shaped roller conveyor of the main roller. Based on the final image of the foreign object in the trough-shaped roller conveyor, the center distance of the foreign object in the trough-shaped roller conveyor is extracted, and the coordinates of the centroid of the foreign object in the image are obtained according to the center distance. Furthermore, combined with the position of the acquisition device, the actual position of the foreign object in the trough-shaped roller conveyor is obtained.

[0035] A small-pitch main roller trough cleaning system based on automated control includes: a parameter setting unit, a path planning unit, a cleaning unit, and a confirmation unit.

[0036] Parameter setting unit: used to establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and to obtain the width of the copper brush bristles.

[0037] Path planning unit: Used to design the path of the copper brush on the main roller based on the right-hand Cartesian coordinate system and the width of the copper brush bristles, so as to obtain a main roller cleaning path.

[0038] Cleaning unit: Used to start the main roller according to preset parameters and call copper brushes to clean the trough roller track of the main roller according to the main roller cleaning path.

[0039] Confirmation unit: Used to identify whether there are still foreign objects on the trough roller surface of the main roller through the image of the main roller. If not, the cleaning of the trough roller surface of the main roller is completed. If so, according to the location of the foreign object, the copper brush is called to clean the trough roller surface of the main roller again.

[0040] The process involves providing a copper brush that meets the parameter requirements and fixing the copper brush on a copper brush fixture; and acquiring video frames of the main roller through a data acquisition device to obtain an image of the main roller.

[0041] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control, characterized in that, Includes the following steps: Provide a copper brush that meets the parameter requirements, and fix the copper brush that meets the parameter requirements on a copper brush fixture; Establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and obtain the width of the copper brush bristles; Based on the right-hand Cartesian coordinate system and the width of the copper brush bristles, the path of the copper brush on the main roller is designed to obtain a main roller cleaning path. Obtain the coordinates of the head and tail of the main roller in the right-hand Cartesian coordinate system; set a first approach position, a second approach position, and a third approach position according to the coordinates of the head of the main roller; set the distance to the left after the copper brush cleans the main roller for one revolution according to the coordinates of the head and tail of the main roller; wherein, the first approach position, the second approach position, and the third approach position are gradually approaching the head of the main roller; the overall direction of brushing the main roller is from head to tail, from right to left; The main roller is started according to preset parameters, and the copper brush is called to clean the trough-shaped roller track of the main roller according to the main roller cleaning path; According to the preset elasticity of the downward pressure copper brush, the copper brush is used to clean the trough-shaped roller channel of the main roller; wherein, when the elasticity of the downward pressure copper brush is preset, it includes: obtaining the depth of the trough-shaped roller channel of the main roller, and determining the elasticity of the downward pressure copper brush to be preset according to the depth of the trough-shaped roller channel of the main roller. The copper brush is gradually brought closer to the head of the main roller. When the copper brush reaches the first approach position, it pauses for the first time. When the first pause time reaches a preset pause time, the copper brush is brought closer to the head of the main roller again. When the copper brush reaches the second approach position, it pauses for the second time. When the second pause time reaches a preset pause time, the copper brush is brought closer to the head of the main roller again. When the copper brush reaches the third approach position, it pauses for the third time. When the third pause time reaches a preset pause time, the copper brush is brought closer to the head of the main roller until it reaches the coordinates of the head of the main roller. Then, the trough-shaped roller conveyor of the main roller is cleaned according to preset parameters. The acquisition device captures video frames of the main roller trough-shaped roller conveyor to obtain an image of the main roller trough-shaped roller conveyor. The image of the main roller trough-shaped roller conveyor is used to identify whether there are still foreign objects on the surface of the main roller trough-shaped roller conveyor. If not, the cleaning of the surface of the main roller trough-shaped roller conveyor is completed. If so, the copper brush is used to clean the surface of the main roller trough-shaped roller conveyor again according to the location of the foreign object.

2. The method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control according to claim 1, characterized in that, When providing a copper brush that meets the parameter requirements, it includes: We provide copper brushes with bristle diameters of 0.076 mm. When the main roller is started according to preset parameters, the following is included: The main roller is started at a speed of 240 revolutions per minute.

3. The method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control according to claim 1, characterized in that, When obtaining the image of the main roller trough, the following steps are included: The main roller is captured by a camera. The main roller image frame data contained in the video frame is preprocessed to obtain a preprocessed main roller image frame. The preprocessed main roller image frame is converted into a grayscale image of the main roller. The main roller trough roller track is extracted from the grayscale image of the main roller to obtain the image of the main roller trough roller track. The preprocessing process includes mean filtering and pixel-based gradient-based image enhancement. Mean filtering includes: Let the target pixel be in the main roller image frame. The pixel value With the target pixel Select one as the center The window is used to calculate the sum of each pixel in the main roller image frame and all pixels in the surrounding template window, and an average value is obtained. The average value is used as the target pixel. The processed grayscale values ​​are shown in Formula 1: （1）； In the formula, For the target pixel point Smoothed grayscale values The size of the template window. It is an odd number; When performing pixel-based gradient-based image enhancement, the following are included: Let the image after mean filtering be And the image size is 4096*2048. This indicates that the image after mean filtering is at point... The specific enhancement process for the pixel value at that location is as follows: The mean-filtered image is divided into 256*512 8*8 sub-grids, and the average value of each grid is obtained. As shown in Formula 2: （2）； The difference between the midpoint pixel value and the grid mean in the mean-filtered image is obtained to determine the threshold for gradient enhancement. As shown in Formula 3: （3）； In each grid cell, the pixel values ​​of the mean-filtered image are compared with the gradient enhancement threshold. Perform summation and limit the image pixel values ​​to 10 ... Between these, an enhanced image is obtained. As shown in Formula 4: （4）； For an 8x8 image grid, move it up and to the right three times, moving two pixels each time. Calculate the mean value again within each different grid cell. Repeat this process to obtain the mean value for each grid cell. and gradient enhancement threshold Steps; Each move changes the number of grid cells. The final augmentation result is obtained by combining the seven images acquired from the grid cell movement.

4. The method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control according to claim 3, characterized in that, Extracting the main roller grooved track from the grayscale image of the main roller includes: use The operator obtains the gradient magnitude and direction of the grayscale image of the main roller, and applies non-maximum suppression to the gradient magnitude for edge refinement; A dual-threshold algorithm is used to detect and connect edges to obtain the image contour; The straight line segments present in the grayscale image of the main roller are detected by the probabilistic Hough transform algorithm and filtered to finally segment the trough-shaped roller track of the main roller.

5. The method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control according to claim 3, characterized in that, When identifying whether there are still foreign objects inside the main roller trough-shaped roller conveyor through the image of the main roller trough-shaped roller conveyor, the following steps are included: Foreign objects in the trough-shaped roller conveyor of the main roller are detected using a background difference method with automatic background updating. The differential image of the main roll trough is shown in Equations 5 and 6: （5）； （6）； In the formula, Indicates the main roller image frame number. This indicates the update rate of the main roller image frames. Represents the pixels in the horizontal direction of the main roller image frame. Represents the pixels in the vertical direction of the main roller image frame. Function representing the background frame image. The function representing the current frame image. This represents the difference image function.

6. The method for cleaning a small-pitch main roller trough-shaped roller conveyor based on automated control according to claim 4, characterized in that, After obtaining the differential image of the main roll trough, the process includes: Set the threshold for binarization processing Based on the threshold The differential image of the main roller trough is binarized to obtain the binarized image of the main roller trough, as shown in Formula 7: （7）； The binarized image of the main roller grooved roller track is ANDed with the main roller grooved roller track extracted from the grayscale image of the main roller to obtain a preliminary image of the foreign object in the main roller grooved roller track. The preliminary image is processed using morphological filtering to obtain the final image of the foreign object in the main roller trough. Based on the final image of the foreign object in the main roller trough, the center distance of the foreign object in the main roller trough is extracted, and the coordinates of the centroid of the foreign object in the image are obtained according to the center distance. Furthermore, combined with the position of the acquisition device, the actual position of the foreign object in the main roller trough is obtained.

7. A small-pitch main roller trough-shaped roller conveyor cleaning system based on automated control, characterized in that, The cleaning method according to any one of claims 1-6 includes: Parameter setting unit: used to establish a right-handed Cartesian coordinate system with the head of the main roller as the origin, and to obtain the width of the copper brush bristles; Path planning unit: used to design the path of the copper brush main roller according to the right-hand Cartesian coordinate system and the width of the copper brush bristles, so as to obtain a clear logical flow path of one brush action; Cleaning unit: used to start the main roller according to preset parameters, and call the copper brush to clean the trough roller track of the main roller according to the main roller cleaning path; Confirmation unit: used to identify whether there are still foreign objects on the trough roller surface of the main roller through the image of the main roller. If not, the cleaning of the trough roller surface of the main roller is completed. If so, the copper brush is called to clean the trough roller surface of the main roller again according to the location of the foreign object. The process involves providing a copper brush that meets the parameter requirements and fixing the copper brush on a copper brush fixture; and acquiring video frames of the main roller trough-shaped roller conveyor using a data acquisition device to obtain an image of the main roller trough-shaped roller conveyor.

Images