Board separating method and board separating apparatus

Abstract

The application discloses a board splitting method and device. The board splitting method comprises the following steps: acquiring an edge feature image of a to-be-processed circuit board; comparing the edge feature image of the to-be-processed circuit board with a template in a visual library to determine an actual position coordinate deviation of the to-be-processed circuit board; generating a board splitting path based on the actual position coordinate deviation and a preset constraint condition, and performing a board splitting action according to the board splitting path; and after performing the action of cutting the to-be-processed circuit board, updating a processing parameter of the board splitting based on a comparison result of a cut image of the circuit board and a preset qualified board splitting image template in the visual library. The application aims to improve the accuracy and safety of the board splitting, improve the subjective difference of manual judgment in the detection of defective board splitting, improve the data reliability, and realize dynamic self-optimization of the parameter.

Description

Technical Field

[0001] This application relates to the field of board separation technology, specifically to a board separation method and a board separation device. Background Technology

[0002] Traditional PCB depaneling equipment relies on manual positioning and depaneling. The depaneling accuracy is easily limited by the accuracy of manual operation, and abnormal conditions require manual inspection, resulting in low efficiency and safety. Furthermore, the handling of defective products in traditional PCB depaneling depends on manual judgment, and defective conditions can only be recorded manually. This leads to large fluctuations in defect judgment results, affecting the consistency and reliability of product quality, and lacking reliable data support for subsequent quality analysis and root cause tracing. Summary of the Invention

[0003] The main purpose of this application is to propose a separation method and separation device, which aims to improve the accuracy and safety of separation, reduce the subjective differences in human judgment during defective product detection, improve data reliability, and achieve dynamic self-optimization of parameters.

[0004] This application proposes a plate-splitting method, including the following steps: Obtain the edge feature image of the circuit board to be processed; The edge feature image of the circuit board to be processed is compared with the template in the visual library to determine the actual position coordinate deviation of the circuit board to be processed. A board-splitting path is generated based on the actual position coordinate deviation and preset constraints, and the board-splitting action is performed according to the board-splitting path. After performing the cutting and separating action on the circuit board to be processed, the processing parameters for separating the circuit board are updated based on the comparison results between the cut image of the circuit board and the preset qualified separating image template in the vision library, and the separating action is performed according to the processing parameters.

[0005] Optionally, the step of acquiring the edge feature image of the circuit board to be processed includes: Obtain a feature image of the circuit board to be processed, the feature image including a reference point image of the circuit board to be processed; The reference point image of the circuit board to be processed is sequentially subjected to grayscale conversion, Gaussian filtering, binarization, and edge extraction to obtain the edge feature image.

[0006] Optionally, the step of sequentially performing grayscale conversion, Gaussian filtering, binarization, and edge extraction on the reference point image of the circuit board to be processed to obtain an edge feature image includes: The reference point image of the circuit board to be processed is converted from a color image to a grayscale image using the grayscale conversion formula fgray(x,y)=0.299R+0.587G+0.114B; where R, G, and B represent the values ​​of the red, green, and blue color channels of each pixel in the color image, respectively. The grayscale image is filtered using a 3×3 convolution kernel with a standard deviation σ=1.2. The grayscale image is converted into a binary image based on a preset grayscale threshold. Based on a preset edge threshold, the edge features of the reference point are extracted.

[0007] Optionally, the board separation method further includes obtaining the coordinate offset data and tilt angle of the circuit board to be processed; The step of comparing the edge feature image of the circuit board to be processed with the template in the visual library to determine the actual position coordinate deviation of the circuit board to be processed is as follows: The edge feature image is compared with the templates in the visual database using a feature matching algorithm; Based on the comparison results, the center coordinates and tilt angle of the reference point of the circuit board to be processed are determined; The actual position coordinate deviation of the circuit board to be processed is determined based on the reference point center coordinates and tilt angle of the circuit board to be processed, and the coordinate offset data and tilt angle of the circuit board to be processed.

[0008] Optionally, the step of comparing the edge feature image with the template in the visual database using a feature matching algorithm includes: A scale space is constructed based on the SIFT algorithm and according to the first preset scale value, the second preset scale value, and the third preset scale value. In scale space, extreme points are detected, and the direction of the extreme points is determined by neighborhood gradient statistics to generate SIFT feature vectors. Calculate the Euclidean distance between the SIFT feature vector and the feature vector of the preset reference point template in the visual library; If the ratio of the first Euclidean distance to the second Euclidean distance is less than a preset ratio, the center coordinates and tilt angle of the reference point in the visual coordinate system are determined, where the first Euclidean distance and the second Euclidean distance are the minimum and second minimum values ​​of the Euclidean distance, respectively.

[0009] Optionally, the preset constraint conditions include at least one of the following three: The minimum safe distance between the board separation path and the components shall not be less than a preset distance threshold. The cutting speed is within the preset speed range; The tool path is parallel to the edge or rib direction of the circuit board; Among them, when the preset constraint conditions include at least two of the three, the priority of the minimum safe distance between the board separation path and the component not being lower than the preset distance threshold is greater than the priority of the cutting speed being in the preset speed range, and the priority of the cutting speed being in the preset speed range is greater than the priority of the cutting direction being parallel to the board edge or rib direction of the circuit board.

[0010] Optionally, the step of generating the partitioning path based on the actual position coordinate deviation and preset constraints further includes: The contact pressure value between the cutting tool and the circuit board during the depaneling process is obtained, wherein the contact pressure value is related to the degree of tool wear; The plate separation path is dynamically adjusted based on the contact pressure value.

[0011] Optionally, the step of dynamically adjusting the separating path based on the contact pressure value includes: If the contact pressure value is less than the preset pressure value, the cutting is performed continuously according to a preset first length. When the contact pressure value is greater than or equal to the preset pressure value, the plate-splitting path is divided into sub-paths of no greater than the preset second length, and chip removal is performed during the segmentation gap.

[0012] Optionally, the step of generating a partitioning path based on the actual position coordinate deviation and preset constraints, and performing partitioning actions according to the partitioning path, further includes: If the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation, it is determined that the actual position coordinate deviation of the circuit board meets the cutting conditions, and the board separation action is performed according to the board separation path. If the actual position coordinate deviation of the circuit board to be processed is greater than the first preset deviation but less than the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, the board separation is paused and an alarm is triggered. If the actual position coordinate deviation of the circuit board to be processed is greater than or equal to the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, and an emergency stop action is directly executed.

[0013] Optionally, the step of determining that the actual position coordinate deviation of the circuit board meets the cutting conditions and performing the board separation action according to the board separation path when the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation further includes: Obtain the contact pressure value between the cutting tool and the circuit board during the board separation process; The speed of separating the boards and the contact pressure value are adjusted according to the thickness of the circuit board; During the separation process, the contact pressure value is adjusted by adjusting the depth of cut, and the fluctuation value of the contact pressure value relative to the preset target contact pressure value is kept within a first preset range.

[0014] Optionally, the plate-splitting method further includes: Acquire multi-dimensional operating status data, which includes at least two of the following: the actual position coordinate deviation of the circuit board, the contact pressure value between the tool and the circuit board, the vibration amplitude of the equipment during operation, and the operating temperature of the equipment. The multi-dimensional operating status data is compared with a preset threshold to determine the anomaly level, and the corresponding response action is executed based on the anomaly level.

[0015] Optionally, the anomaly level includes Level 1 anomaly, Level 2 anomaly, and Level 3 anomaly; the response actions include recording the anomaly and outputting a prompt signal, pausing the sub-board and issuing an alarm, and performing software emergency stop and hardware reset power-off within a preset time. The steps of comparing the multi-dimensional operational status data with preset thresholds to determine the anomaly level, and executing corresponding response actions based on the anomaly level include: If the vibration amplitude of the equipment is lower than the first preset vibration amplitude threshold or the operating temperature is lower than the preset temperature value, it is determined to be a level one anomaly. The response action based on the level one anomaly is to record the anomaly and output a prompt signal. If the actual position coordinate deviation of the circuit board is greater than the second preset deviation and less than the third preset deviation, or if the fluctuation value of the contact pressure value relative to the preset target contact pressure value is within the second preset range, it is determined to be a level two abnormality. The response action based on the level two abnormality is to suspend board separation and issue an alarm. If the vibration amplitude exceeds the second preset vibration amplitude threshold or the tool breaks / the plate gets stuck, it is determined to be a level three anomaly. The response action based on the level three anomaly is to perform a software emergency stop and a hardware reset power-off within a preset time.

[0016] Optionally, the multi-dimensional operating status data may also include: the offset of the board cutting trajectory; The steps of comparing the multi-dimensional operational status data with preset thresholds to determine the anomaly level and executing corresponding response actions based on the anomaly level further include: If the offset of the plate cutting trajectory is greater than the preset offset but does not reach the preset scrap offset threshold, a supplementary cutting action is triggered. The finishing move includes: Reacquire the edge feature image of the circuit board to be processed, compare the reacquired edge feature image of the circuit board with the template in the visual library, and determine the actual position coordinate deviation of the reacquired circuit board. The tooling path is generated based on the actual position coordinate deviation and preset constraints. The positional error between the tooling path and the board splitting path at the break point is less than the preset error. Perform the finishing move according to the stated finishing move path.

[0017] Optionally, the step of updating the processing parameters for the board separation after performing the cutting and separating action on the circuit board to be processed, based on the comparison result between the cut image of the circuit board and the preset qualified separation image template in the vision library, includes: Obtain an image of the panel cutting edge, and compare the image of the panel cutting edge with a preset qualified panel image template in the visual library; If the current board-making result is determined to be a defective product, the defect features are extracted from the cut image of the defective product, the defect features are compared with the existing defect features in the visual library, and new types of defect features are identified. The processing parameters of the partition are updated based on the type of defective feature.

[0018] Optionally, the step of comparing the image of the plate cutting with a preset qualified plate image template in the visual library includes: If the burr length at the depanel cut exceeds a preset length value, the depaneled circuit board is determined to be defective; or, If the offset value of the board separation cut is greater than the preset offset value, the board after separation is determined to be a defective product.

[0019] Optionally, the step of extracting defect features from the cut image of the defective product when the current panelization result is determined to be a defective product, comparing the defect features with existing defect features in the visual library, and determining new types of defect features includes: For cut images identified as defective products, extract the contour features, morphological features, and positional features of the cuts, and generate corresponding defect feature vectors; The similarity is calculated based on the defective feature vector and the known types of defective feature vectors stored in the visual database; If the similarity is less than a preset similarity threshold, the defective feature is determined to be a new type. If the similarity is greater than a preset similarity threshold, the defective feature is determined to be of a known type.

[0020] Optionally, the processing parameters for the plate separation include tool pressure, positioning error threshold for tool entry determination, and cutting speed. The step of updating the processing parameters for the plate separation based on the type of defect features includes: If the type of defect is tool wear, reduce the tool pressure; When the defective feature is of the plate position offset type, the positioning error threshold for determining the cutting edge is reduced from the first error threshold to the second error threshold; If the defective feature is of the burr type, reduce the cutting speed.

[0021] Optionally, the step of updating the processing parameters of the partition plate based on the type of defective feature further includes: If the yield rate meets the steady-state condition, maintain the updated board separation parameters; If the yield rate decreases, restore the board separation parameters to their original values ​​before optimization.

[0022] Optionally, the step of updating the processing parameters of the partition plate based on the type of defective feature further includes: Add the new types of defective features to the visual library and update the defective product judgment rules; The step of adding the new type of defective features to the visual library and updating the defective product judgment rules includes: If a new feature is detected in a cut image, and the same new type of defective feature is detected in cut images for a consecutive preset number of frames, it is stored in the cache pool. When the preset production batch is reached, the defective features in the cache pool are clustered, denoised, and checked for repetition before being added to the visual library and the defective product judgment rules are updated.

[0023] This application also proposes a board separation device, which includes a control module; the control module is used to perform the board separation method as described in any one of the above statements.

[0024] This application proposes a board separation method, comprising: acquiring an edge feature image of the circuit board to be processed; comparing the edge feature image of the circuit board to be processed with a template in a vision library to determine the actual position coordinate deviation of the circuit board to be processed; generating a separation path based on the actual position coordinate deviation and preset constraints, and performing separation actions according to the separation path; after performing the action of cutting and separating the circuit board to be processed, updating the processing parameters of the separation based on the comparison result of the cut image of the circuit board and the preset qualified separation image template in the vision library, and performing separation actions according to the processing parameters. Through the above settings, this application introduces a vision recognition and positioning mechanism based on a vision library, which improves the positioning deviation caused by manual loading, fixture errors, etc., provides a reliable data foundation for high-precision separation, improves the separation accuracy, and is superior to manual operation. By setting constraints, this application enables the device to intelligently plan an optimal path that avoids sensitive components, reduces the burr rate of the cutting edge, and reduces the internal stress of the board. This not only reduces the problem of component damage or circuit board scrap caused by improper cutting, but also improves the yield and reliability of the final product. In addition, this application achieves automated and standardized detection of plate-making quality by quantitatively comparing the plate-making cut images with qualified plate-making image templates in the visual library and comparing the similarity of defective features with existing defective features in the visual library, thus providing an accurate classification basis for subsequent targeted optimization. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0026] Figure 1 This is a schematic flowchart of an embodiment of the board separation method of this application; Figure 2 This is a schematic flowchart illustrating an embodiment of the board separation method of this application for obtaining edge feature images of the circuit board to be processed. Figure 3 In the plate-splitting method of this application Figure 2 A schematic flowchart of a refinement method for obtaining an edge feature image of a circuit board to be processed is shown in the example. Figure 4 This is a schematic flowchart illustrating an embodiment of the method for determining the actual position coordinate deviation of the circuit board to be processed in the board separation method of this application; Figure 5 This is a schematic flowchart illustrating an embodiment of the plate-splitting method in this application that compares edge feature images with templates in a visual library. Figure 6 This is a schematic flowchart of an embodiment of the board splitting method in this application, which generates a board splitting path based on actual position coordinate deviation and preset constraints. Figure 7 This is a schematic flowchart of an embodiment of the plate separation method in this application that dynamically adjusts the plate separation path based on the contact pressure value; Figure 8 This is a schematic flowchart illustrating an embodiment of the board splitting method in this application, which executes board splitting actions according to the board splitting path. Figure 9 This is a schematic flowchart of an embodiment of the plate-separation method in this application for determining whether the cutting conditions are met; Figure 10 This is a schematic flowchart of an embodiment of the abnormal state detection method in the board separation method of this application; Figure 11 This is a schematic flowchart illustrating an embodiment of the board-separating method of this application that performs the finishing cut action; Figure 12 This is a schematic flowchart of an embodiment of the board separation method of this application, which updates the processing parameters of the board separation and performs the board separation action according to the processing parameters. Figure 13 This is a schematic flowchart of an embodiment of the plate-separation method of this application, which compares the image of the plate-separation cut with a preset qualified plate-separation image template in a visual library. Figure 14 This is a schematic flowchart illustrating an embodiment of the board-separation method for identifying new types of defective features in this application. Figure 15 This is a schematic flowchart illustrating an embodiment of the board-separation method of this application that updates the processing parameters of the board based on the type of defective features; Figure 16 This is a schematic diagram of the process flow for updating the board separation parameters based on the yield rate in the board separation method of this application. Figure 17 This is a schematic flowchart illustrating an embodiment of the method for updating the defective product determination rules in the board separation method of this application.

[0027] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0029] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a specific posture. If the specific posture changes, the directional indicators will also change accordingly.

[0030] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the use of "and / or" or "and / or" throughout the text includes three parallel solutions. For example, "A and / or B" includes solution A, solution B, or a solution that simultaneously satisfies A and B. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0031] Currently, traditional PCB depaneling equipment relies on manual positioning, the creation of dedicated depaneling programs, and manual depaneling, which has the following shortcomings: First, the depaneling accuracy is limited by the accuracy of manual operation, easily leading to depaneling misalignment, copper leakage, and burrs. Second, abnormal conditions are difficult to monitor in a timely manner, such as equipment jamming, board deformation, and tool collision, requiring manual inspection, which is inefficient and has low safety. Furthermore, the handling of defective products in traditional PCB depaneling relies on manual judgment, and defective conditions can only be recorded manually, failing to form a structured defect database. This results in large fluctuations in defect judgment results, affecting the consistency and reliability of product quality, and lacking reliable data support for subsequent quality analysis and root cause tracing.

[0032] Therefore, this application proposes a board separation method that can be applied to a board separation device, which includes, but is not limited to: an image acquisition and sensing module, such as an industrial camera; a control module for running vision algorithms; and an execution module for performing cutting actions, such as a cutting tool or a transmission module.

[0033] The control module can be implemented using a main controller, such as an MCU, PLC, digital signal processing chip, device-level chip, or programmable gate array chip. The control module can establish bidirectional communication connections with both the sensing and execution modules via industrial Ethernet, forming a closed-loop control architecture of "sensing-decision-execution".

[0034] refer to Figure 1In one embodiment of this application, the board separation method includes: S100: Obtain the edge feature image of the circuit board to be processed; S200. Compare the edge feature image of the circuit board to be processed with the template in the visual library to determine the actual position coordinate deviation of the circuit board to be processed. S300: Generate a panel splitting path based on the actual position coordinate deviation and preset constraints, and execute panel splitting actions according to the panel splitting path; S500: After performing the action of cutting and separating the circuit board to be processed, based on the comparison results between the cut image of the circuit board and the preset qualified separation image template in the vision library, update the separation processing parameters and perform the separation action according to the processing parameters.

[0035] In this embodiment, optionally, the control module acquires a surface image of the circuit board to be processed through the sensing module. For example, an industrial camera with a global shutter and 2 megapixels, a resolution of 1600×1200, and a frame rate of 30fps is used to acquire images of at least two sets of reference points on the PCB board. After acquiring the original image, the device sequentially performs grayscale conversion, Gaussian filtering, binarization, and edge extraction on the image to extract clear edge feature images. Optionally, the control module can also directly acquire the edge feature image processed by the image processing module.

[0036] Optionally, the circuit board to be processed can be a PCB board or a PCBA board, wherein the PCBA board is a finished circuit board with all electronic components already soldered on.

[0037] Optionally, after obtaining the edge feature image of the circuit board to be processed, the control module calls the vision library module, which pre-stores standard reference point template features for the corresponding board type. The control module uses the SIFT feature matching algorithm to compare the acquired edge feature image with the preset reference point feature image of the board type in the vision library, thereby determining the center coordinates and tilt angle of the reference point in the visual coordinate system.

[0038] It is important to understand that, to further improve the robustness of positioning, the device can also acquire coordinate offset data and tilt angle collected by the displacement sensor as positioning redundancy data. Finally, by combining the center coordinates and tilt angle of the reference point with the coordinate offset data and tilt angle collected by the displacement sensor, the actual position coordinate deviation of the circuit board to be processed relative to the preset reference coordinate system is determined.

[0039] For example, consider a PCB board to be processed. Its theoretical reference position relative to a preset reference coordinate system is as follows: theoretical reference point coordinates are (100mm, 100mm), and the theoretical tilt angle is 0°, meaning it is horizontally placed. An industrial camera captures an image of the reference point on the board. After feature matching, the visual library module identifies the center coordinates as (100.3, 100.4) and the tilt angle as +0.6°, meaning the visual library module considers the board to be offset 0.3mm to the right, 0.4mm upwards, and rotated 0.6° counterclockwise. However, the displacement sensor acquires an initial X-axis offset of +0.25mm, an initial Y-axis offset of +0.45mm, and a tilt angle of +0.55°, meaning the sensor considers the board to be offset 0.25mm to the right, 0.45mm upwards, and rotated 0.55° counterclockwise. The control module can weightedly fuse the results identified by the visual library module with the data acquired by the displacement sensor to eliminate errors from a single source. Thus, by fusing the visual positioning data and the displacement sensor data as described above, the positioning redundancy data of the displacement sensor can effectively improve the problem that single visual positioning is easily affected by lighting and board texture.

[0040] Optionally, after obtaining the actual position coordinate deviation, the control module will combine it with the theoretical board splitting path and, in conjunction with preset constraints, generate the optimal actual board splitting path.

[0041] For example, the device needs to cut along a straight line on a circuit board to separate two areas. The theoretical path based on the CAD design is a horizontal straight line from point A (10mm, 10mm) to point B (110mm, 10mm). After visual recognition and comparison with a template, the device detects that the current circuit board has shifted relative to its standard position, for example, by +2mm in the X direction and +1mm in the Y direction. Therefore, the control module does not physically move the circuit board, but rather modifies the tool's trajectory to adapt to the actual position of the circuit board. The theoretical path A (10mm, 10mm) → B (110mm, 10mm), combined with the actual position coordinate deviation, becomes: A'(10+2,10+1)=A'(12mm,11mm)→B'(110+2,10+1)=B'(112mm,11mm).

[0042] Optionally, the preset constraints can be at least one of the following: the minimum safe distance between the separation path and the components on the circuit board is not less than a preset distance threshold to avoid cutting damage to the components; the cutting speed is within a preset speed range to balance cutting efficiency and quality; the cutting direction is parallel to the edge of the circuit board or the direction of the ribs to reduce internal stress during cutting and prevent board deformation. After the path is generated, the execution module performs the separation action according to the separation path.

[0043] In one example, the preset spacing threshold is 1.5mm, and the preset speed range is 20-80mm / s.

[0044] Optionally, after performing the cutting and separating action on the target circuit board, the control module acquires an image of the separating cut and compares it with a preset qualified separating image template in the vision library. If the current separating result is determined to be a defective product, the control module extracts defect features from the defective product's cut image, compares these features with existing defect features in the vision library, identifies new types of defect features, updates the separating processing parameters based on the type of defect features, and executes the separating action according to the processing parameters.

[0045] Through the above-described setup, this application introduces a vision-based visual recognition and positioning mechanism, improving positioning deviations caused by manual loading and fixture errors. This provides a reliable data foundation for high-precision board separation, enhancing separation accuracy compared to manual operation. By setting constraints, the device can plan an optimal path that avoids sensitive components, reduces burr rates at the cutting edges, and minimizes internal stress in the board. This not only reduces component damage or circuit board scrap caused by improper cutting but also improves the yield and reliability of the final product. Furthermore, this application achieves automated and standardized detection of board separation quality by quantitatively comparing the board separation cut image with qualified board separation image templates in the vision library and comparing the similarity of defective features with existing defective features in the vision library. This provides accurate classification criteria for subsequent targeted optimization.

[0046] According to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 2 The steps for obtaining the edge feature image of the circuit board to be processed include: S110. Obtain the feature image of the circuit board to be processed, including the reference point image of the circuit board to be processed; S120. The reference point image of the circuit board to be processed is sequentially converted to grayscale, Gaussian filtered, binarized and edge extracted to obtain the edge feature image.

[0047] In this embodiment, optionally, the control module of the board separation device controls the sensing module to acquire images of the circuit board to be processed. To ensure positioning accuracy, the acquired images contain at least two sets of reference points on the circuit board. Optionally, the reference points are positioning markers with standard geometric shapes preset in the circuit board design, such as circles or squares. The reference points are usually located at the corners or edges of the board.

[0048] In one example, the sensing module could be a 2-megapixel industrial camera with a global shutter, a resolution of 1600×1200, and a frame rate of 30fps, suitable for production line inspection or positioning tasks.

[0049] It is important to understand that the original image cannot be directly used for high-precision feature matching due to factors such as color information, environmental noise, and uneven lighting. Therefore, the control module sequentially performs grayscale conversion, Gaussian filtering, binarization, and edge extraction on the reference point image of the circuit board to be processed, resulting in a clear, continuous, and interference-free reference point edge feature image.

[0050] Through the above settings, this application constructs a complete preprocessing chain for circuit boards by sequentially performing grayscale conversion, Gaussian filtering, binarization, and edge extraction. This ensures that the final extracted edge feature image is clear, continuous, and realistic, thereby significantly improving the success rate and anti-interference capability of subsequent matching with visual library templates.

[0051] It is important to understand that, in the face of specific interferences in PCB depaneling scenarios, such as dust, reflections, scratches, and silkscreen printing, we need to extract reference point edge features suitable for high-precision matching in order to improve the accuracy of subsequent depaneling.

[0052] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 3 The steps for obtaining the edge feature image by sequentially performing grayscale conversion, Gaussian filtering, binarization, and edge extraction on the reference point image of the circuit board to be processed include: S121. The reference point image of the circuit board to be processed is converted from a color image to a grayscale image using the grayscale conversion formula fgray(x,y)=0.299R+0.587G+0.114B; where R, G, and B represent the values ​​of the red, green, and blue color channels of each pixel in the color image, respectively. S122. Use a 3×3 convolution kernel with a standard deviation σ=1.2 to filter the grayscale image; S123. Convert the grayscale image into a binary image based on a preset grayscale threshold; S124. Extract the edge features of the reference point based on the preset edge threshold.

[0053] In this embodiment, optionally, the control module first converts the color reference point image acquired by the sensing module into a grayscale image to reduce the amount of data in subsequent calculations. The standard brightness-weighted conversion formula, fgray(x,y)=0.299R+0.587G+0.114B, is used to calculate the value for each pixel in the image. Here, R, G, and B represent the values ​​of the red, green, and blue color channels of each pixel in the color image, respectively. The weighting coefficients in the formula—0.299, 0.587, and 0.114—conform to the physiological visual characteristics of the human eye, which is most sensitive to the green channel and least sensitive to the blue channel. This ensures that a grayscale image with the highest fidelity of brightness information is generated, guaranteeing that the contour information of the reference points is not lost during the conversion process.

[0054] Optionally, to suppress dust, lens salt-and-pepper noise, fine copper foil texture noise on the PCB board surface, and slight reflective interference present in the workshop environment, the control module performs Gaussian filtering on the grayscale image. A 3×3 convolution kernel with a standard deviation σ=1.2 is used to filter the grayscale image and perform convolution operation with the input image to suppress image noise.

[0055] The above settings effectively filter out high-frequency noise interference. Furthermore, the use of a 3×3 small convolution kernel provides only mild smoothing, avoiding blurring of the fine contour edges of the reference points, thus achieving a balance between noise reduction and edge preservation. The standard deviation σ = 1.2 ensures a moderate weight distribution for the Gaussian kernel, preventing it from being too concentrated, resulting in insufficient noise reduction, or too dispersed, leading to blurred edges.

[0056] Optionally, to clearly separate the reference point region from the background region, the control module converts the filtered grayscale image into a binary image based on a preset grayscale threshold. The conversion rule is as follows: the grayscale threshold T is set to 80, where the grayscale value ranges from 0 to 255. For each pixel in the image, if its grayscale value is greater than or equal to 80, the value of that pixel is set to 255, representing a white foreground; if its grayscale value is less than 80, the value of that pixel is set to 0, representing a black background. After this binarization process, the reference point region is clearly highlighted as a white foreground, while the background areas such as the circuit board substrate background color and silkscreen are classified as a black background, creating ideal regional conditions for subsequent edge extraction.

[0057] Optionally, the control module employs the Canny edge extraction algorithm, extracting edge features from reference points based on preset edge thresholds. The low threshold for the Canny operator is set to 50, and the high threshold to 150. The logic of the edge extraction algorithm is as follows: For each pixel in the image, its gradient magnitude and direction are first calculated; pixels with gradient magnitudes greater than the high threshold are directly identified as strong edge points; pixels with gradient magnitudes less than the low threshold are directly suppressed; pixels with gradient magnitudes between the low and high thresholds are retained as edge points only if they are connected to strong edge points.

[0058] It is important to understand that these high and low thresholds are parameters specifically designed for the regular geometric edges of PCBs: the high threshold of 150 is used to retain only the strong principal edges of the reference points, effectively filtering out false edges caused by minor scratches, silkscreen fragments, dust, etc. on the PCB board surface; the low threshold of 50 is used to connect the reference point contour edges that are broken due to uneven lighting or slight wear, ensuring that the extracted circular or square reference point edges are closed and complete, providing continuous geometric features for subsequent SIFT feature matching.

[0059] With the above settings, the processed edge feature image exhibits high signal-to-noise ratio, continuous edges, and a clean background. This allows subsequent feature matching algorithms such as SIFT to perform calculations based on higher-quality input data when constructing scale space, detecting extreme points, and generating feature vectors. This reduces the number of invalid feature points, thereby improving matching speed and success rate, ultimately ensuring the accuracy and reliability of the plate localization.

[0060] It is important to understand that, to further improve the robustness of positioning, the device can also acquire coordinate offset data and tilt angle collected by the displacement sensor as positioning redundancy data. Finally, by combining the center coordinates and tilt angle of the reference point with the coordinate offset data and tilt angle collected by the displacement sensor, the actual position coordinate deviation of the circuit board to be processed relative to the preset reference coordinate system is determined.

[0061] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 4 The plate-splitting method also includes: Obtain the coordinate offset data and tilt angle of the circuit board to be processed; The specific steps for determining the actual position coordinate deviation of the circuit board to be processed by comparing the edge feature image with the template in the visual library are as follows: S210. Use a feature matching algorithm to compare the edge feature image with the template in the visual database; S220. Determine the center coordinates and tilt angle of the reference point of the circuit board to be processed based on the comparison results; S230. Determine the actual position coordinate deviation of the circuit board to be processed based on the reference point center coordinates and tilt angle of the circuit board to be processed and the coordinate offset data and tilt angle of the circuit board to be processed.

[0062] In this embodiment, optionally, while acquiring the reference point image, the control module also obtains the coordinate offset data and tilt angle of the circuit board to be processed through a displacement sensor as positioning redundancy data for subsequent fusion calibration with the visual matching results. The board separation device is equipped with a displacement sensor, such as a laser displacement meter, a grating ruler, or a high-precision contact displacement sensor. This sensor acquires the coordinate offset data of the circuit board to be processed in real time at a preset sampling frequency. The acquired data includes: the initial offset of the X-axis, which characterizes the offset value of the actual position of the circuit board in the X direction relative to the reference position; the initial offset of the Y-axis, which characterizes the offset value of the actual position of the circuit board in the Y direction relative to the reference position; and the tilt angle, which characterizes the rotation angle of the circuit board relative to the reference position in its actual placement state.

[0063] Optionally, the control module calls the vision library module and uses a feature matching algorithm to compare the edge feature image with the pre-stored reference point templates in the vision library to obtain the positioning result based on the vision library module. After obtaining the center coordinates and tilt angle of the reference point and the coordinate offset data and tilt angle collected by the displacement sensor, the control module performs fusion processing on these two types of data to finally determine the actual position coordinate deviation of the circuit board to be processed relative to the preset reference coordinate system.

[0064] For example, the control module uses a weighted average to fuse the positioning results from the vision library module with the coordinate offset data and tilt angle collected by the displacement sensor: Final direction deviation = Weight of the positioning result from the vision library module × Deviation of the positioning result from the vision library module relative to the theoretical benchmark + Weight of the data collected by the displacement sensor × Coordinate offset data collected by the displacement sensor. The specific values ​​of the weighting coefficients can be adjusted according to the actual application scenario. For example, the visual weight can be appropriately increased in environments with good lighting conditions, and the sensor weight can be appropriately increased in environments with significant vibration interference.

[0065] By introducing coordinate offset data and tilt angle of the circuit board to be processed collected by displacement sensors as redundant positioning data, a dual positioning mechanism of "vision + physical" is formed. When the positioning of the vision library module is disturbed, the displacement sensor data can serve as an effective supplement and verification; when there is mechanical backlash error in the displacement sensor, the positioning result of the vision library module can provide high-precision correction. The redundancy and fusion of the two data sources improve the overall positioning accuracy.

[0066] According to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 5The steps of comparing edge feature images with templates in a visual database using a feature matching algorithm include: S211. Construct a scale space based on the SIFT algorithm and according to the first preset scale value, the second preset scale value, and the third preset scale value; S212. Detect extreme points in scale space, determine the direction of extreme points through neighborhood gradient statistics, and generate SIFT feature vectors. S213. Calculate the Euclidean distance between the SIFT feature vector and the feature vector of the preset reference point template in the visual library. S214. If the ratio of the first Euclidean distance to the second Euclidean distance is less than a preset ratio, determine the center coordinates and tilt angle of the reference point in the visual coordinate system. The first Euclidean distance and the second Euclidean distance are the minimum and second minimum Euclidean distances, respectively.

[0067] In this embodiment, optionally, the control module uses the SIFT algorithm to extract features from the preprocessed edge feature image. Unlike the traditional SIFT algorithm, which traverses the entire scale space, this application considers that the reference points on the circuit board are mostly fixed-size circular or square reference markers with minimal physical size fluctuations and relatively stable pixel sizes after imaging, eliminating the need for indiscriminate traversal across the entire scale. Therefore, this embodiment selects a first preset scale value σ1=0.8, a second preset scale value σ2=1.6, and a third preset scale value σ3=3.2, constructing a scale space only at these three discrete scales. Through the above settings, the pixel size range of the circuit board reference points at typical imaging distances is accurately covered, and extreme value detection is performed only within the effective scale, eliminating invalid scale traversal operations, significantly reducing computational load, improving matching speed, and adapting to the high efficiency requirements of the board separation production line.

[0068] Optionally, in the constructed three-layer scale space, the control module detects extreme points in each scale space. The control module compares each pixel with its eight neighboring pixels at the same scale and nine pixels in each of the adjacent scales above and below. If the value of a pixel is greater than or less than that of all its neighboring pixels, it is identified as a candidate extreme point. For each detected extreme point, the control module calculates the gradient magnitude and direction of each pixel in the neighborhood of that extreme point within its scale space, generates a gradient direction histogram, and determines the peak direction of the histogram as the principal direction of that extreme point. After assigning the direction, the control module divides the neighborhood of the extreme point into multiple sub-regions, calculates the gradient information within each sub-region, and generates a 128-dimensional SIFT feature vector V. This feature vector describes the local texture and gradient distribution information within the neighborhood of the extreme point and possesses scale invariance and rotation invariance.

[0069] Optionally, the control module calculates the similarity between the SIFT feature vector V and the feature vector V1 of the pre-stored reference point template in the visual library, using Euclidean distance as the similarity metric. The smaller the Euclidean distance, the more similar the two feature vectors are. For each SIFT feature point in the input image, the control module calculates its Euclidean distance with all feature points in the visual library template, and selects the first Euclidean distance D_min1 and the second Euclidean distance D_min2, where the first Euclidean distance is the minimum and the second Euclidean distance is the second minimum. The template feature point corresponding to the first Euclidean distance D_min1 has the highest similarity to the input feature point, and the template feature point corresponding to the second Euclidean distance D_min2 has the second highest similarity to the input feature point.

[0070] Optionally, to ensure the validity of the matching and avoid false matches, the control module sets a matching judgment threshold, which is the ratio of the first Euclidean distance D_min1 to the second Euclidean distance D_min2. In a specific embodiment, a preset ratio of 0.75 is set. When D_min1 / D_min2 < 0.75, it indicates that the similarity of the nearest neighbor matching point is significantly better than that of the second nearest neighbor matching point, the matching result has a high confidence level, and the control module judges the pair of matches as valid matches.

[0071] After acquiring a sufficient number of valid matching points, the control module calculates the center coordinates and tilt angle of the reference point in the visual coordinate system by solving the spatial transformation relationship between the matching point pairs. This result will be used as the comparison result for the localization of the visual library module and will be fused with the displacement sensor data in subsequent steps.

[0072] In one example, after obtaining at least three pairs of valid matching points, the control module calculates the center coordinates and tilt angle of the reference point in the visual coordinate system by solving the affine transformation or similarity transformation relationship between the matching point pairs.

[0073] Conversely, if D_min1 / D_min2≥0.75, it indicates that the similarity difference between the nearest neighbor matching point and the second nearest neighbor matching point is not significant enough, the reliability of the matching result is low, the control module determines that the matching is invalid, and can trigger the re-acquisition of images or output a positioning failure alarm.

[0074] By using the above settings, and considering the extremely small fluctuation range of the physical dimensions of the circuit board reference points, only three preset scale values ​​are selected to construct the scale space, accurately covering the imaging pixel size range of the circuit board reference points and eliminating invalid scale traversal calculations. This optimization significantly reduces the computational load of scale space construction and extreme point detection, and significantly improves the speed of feature matching while maintaining matching accuracy, achieving good adaptation to production line cycle time.

[0075] According to some embodiments of this application, optionally, in one embodiment of this application, the preset constraint conditions include at least one of the following three: The minimum safe distance between the board separation path and the components shall not be less than the preset distance threshold. The cutting speed is within the preset speed range; The tool path is parallel to the edge of the circuit board or the direction of the ribs; Among them, when the preset constraints include at least two of the three, the priority of the minimum safe distance between the board separation path and the components not being lower than the preset distance threshold is greater than the priority of the cutting speed being within the preset speed range, and the priority of the cutting speed being within the preset speed range is greater than the priority of the cutting direction being parallel to the board edge or rib direction of the circuit board.

[0076] In this embodiment, optionally, when planning the board separation path, the control module needs to obtain the component layout information on the circuit board. This layout information can come from board templates stored in the vision library or preset CAD design data. The module then calculates the shortest distance between the planned cutting path and the outline of each component, which is defined as the minimum safety clearance. For example, a preset clearance threshold of 1.5mm is set, meaning that the distance between each point on the planned board separation path and any component outline on the circuit board must be greater than or equal to 1.5mm. If a potential path has a risk of having a distance of less than 1.5mm from a component, the control module will automatically adjust the path until the safety clearance requirement is met. This constraint has the highest priority among all constraints because component damage will directly lead to circuit board malfunction or scrap, the consequences of which are more serious than issues of cutting quality and efficiency.

[0077] Optionally, the preset speed range can be 20mm / s to 80mm / s. That is, when generating the board separation path, the control module must ensure that the corresponding cutting speed falls within the range of 20mm / s to 80mm / s. The specific speed value can be dynamically adjusted within this range based on factors such as the thickness and material of the circuit board, and the tool condition. For example, for thicker circuit boards, a lower speed can be selected to improve cutting quality, such as 30mm / s; for thinner circuit boards, a higher speed can be selected to improve efficiency, such as 70mm / s. This constraint has lower priority than the safety clearance constraint, but higher than the tool path constraint. That is, under the premise of ensuring the safety clearance, the cutting speed is controlled within a reasonable range first, and then the tool path is optimized.

[0078] Optionally, when planning the cutting path, the control module prioritizes the cutting direction with the smallest angle to the edge of the circuit board or the direction of the reinforcing ribs. For example, for rectangular circuit boards, it prioritizes horizontal or vertical cutting paths parallel to the edge; for circuit boards with reinforcing ribs, it prioritizes cutting paths aligned with the extension direction of the ribs. This constraint has the lowest priority and is only considered as an optimization objective if both safety clearance constraints and cutting speed constraints are met simultaneously. This means that the device will not sacrifice component safety or cutting quality in pursuit of parallel cutting directions.

[0079] Optionally, if the preset constraints include at least two of the above three, the R&D personnel will preset the priority order of the constraints and store it in the control module. For example, if the preset conditions are that the minimum safe distance between the board separation path and the component is not lower than a preset distance threshold and the cutting speed is within a preset speed range, the priority of "the minimum safe distance between the board separation path and the component is not lower than the preset distance threshold" is greater than the priority of "the cutting speed is within a preset speed range".

[0080] By incorporating the three key constraints—"component safety spacing," "cutting speed," and "tool path direction"—into the path generation framework, the control module can comprehensively consider multiple dimensions such as safety, cutting quality, cutting efficiency, and board deformation to generate a superior board separation path. This offers significant advantages over existing technologies that plan based on only a single factor. Furthermore, by explicitly setting priority orders, a clear logical basis is provided for decision-making in cases of constraint conflicts. This makes the path generation process predictable and deterministic, reducing the problem of arbitrary and uncontrollable path generation results due to a lack of prioritization.

[0081] It is important to understand that in traditional plate-separating methods, the wear of cutting tools usually relies on manual experience or periodic inspections, which makes it difficult to grasp the situation in a timely and accurate manner. Problems are often only discovered when obvious abnormalities such as chipping or tool breakage occur.

[0082] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 6 The step of generating the board splitting path based on the actual position coordinate deviation and preset constraints also includes: S310. Obtain the contact pressure value between the tool and the circuit board during the separation process, wherein the contact pressure value is related to the degree of tool wear; S320: Dynamically adjust the plate separation path based on the contact pressure value.

[0083] In this embodiment, optionally, during the board separation process, the control module collects the contact pressure value between the tool and the circuit board in real time through a pressure sensor configured on the execution module. This pressure sensor can be installed at the tool spindle or tool holder, and can continuously monitor the axial or radial contact force experienced by the tool during cutting at a high sampling frequency. It should be noted that there is a clear correlation between the contact pressure value and the degree of tool wear. Under the condition that the cutting speed and depth of cut remain constant, as the tool's usage time increases, the cutting edge gradually wears and becomes dull, and the contact resistance between the tool and the circuit board increases accordingly, leading to an increase in the contact pressure value measured by the pressure sensor. Therefore, the contact pressure value can serve as an indirect indicator of the current wear state of the tool. When the contact pressure value is below a certain threshold, it indicates that the tool wear is relatively mild; when the contact pressure value reaches or exceeds this threshold, it indicates that the tool wear has reached a level requiring attention.

[0084] Optionally, the control module compares the real-time contact pressure value with the preset pressure threshold and dynamically adjusts the cutting method of the plate-splitting path according to the comparison result: switching between continuous cutting mode and segmented cutting mode, and adjusting the length of a single continuous cut accordingly.

[0085] Optionally, refer to Figure 7 The steps for dynamically adjusting the separation path based on the contact pressure value include: S321. When the contact pressure value is less than the preset pressure value, cut continuously according to the preset first length; S322. When the contact pressure value is greater than or equal to the preset pressure value, the plate-splitting path is divided into sub-paths with a length not greater than the preset second length, and the tool retraction and chip removal are performed during the segmentation gap.

[0086] For example, when the real-time contact pressure value measured by the pressure sensor is lower than the wear threshold, the control module determines that the current tool wear is relatively light, the cutting edge is still sharp, and it has the ability to cut continuously for a longer period of time. In this case, the control module plans a continuous cutting path according to a preset first length, allowing the tool to cut continuously for the preset first length without interruption. This continuous cutting mode helps reduce the number of tool advances and retractions, improving the efficiency of plate separation. When the real-time contact pressure value measured by the pressure sensor reaches or exceeds the wear threshold, the control module determines that the current tool wear has reached a level that requires attention, the cutting edge becomes dull, and continuous long-term cutting may lead to excessive cutting resistance, resulting in adverse consequences such as edge chipping and tool breakage. In this case, the control module automatically segments the original continuous plate separation path, dividing the entire plate separation path into multiple sub-paths, each with a cutting length not exceeding a preset second length. At the same time, a retraction and chip removal command is inserted between the segments of adjacent sub-paths, causing the tool to briefly lift and remove chips after completing a section of cutting, thereby reducing the impact of chip accumulation on the subsequent cutting quality and reducing the instantaneous load on the tool.

[0087] In one example, the wear threshold is 18N, the preset first length is 60mm, the preset second length is 25mm, and the retraction chip removal command can be a retraction time of 0.1 seconds.

[0088] By dynamically adjusting the cutting method of the board separation path based on the contact pressure value, the path achieves real-time adaptation to the tool state, effectively reducing processing abnormalities caused by excessive tool wear. It also reduces the load time and accumulated cutting heat during a single continuous cut, decreasing the stress on the tool under high wear conditions, thereby lowering the probability of edge chipping and tool breakage, and ensuring circuit board yield and tool life.

[0089] It is important to understand that in traditional plate-separating operations, whether or not to make a cut often depends on the operator's visual observation and experience, which is highly subjective and lacks consistent standards.

[0090] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 8 The steps of generating a panelization path based on the actual position coordinate deviation and preset constraints, and performing panelization actions according to the panelization path, also include: S330. If the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation, determine that the actual position coordinate deviation of the circuit board meets the cutting conditions, and perform the board separation action according to the board separation path. S340. If the actual position coordinate deviation of the circuit board to be processed is greater than the first preset deviation and less than the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, the board separation is paused and an alarm is triggered. S350: If the actual position coordinate deviation of the circuit board to be processed is greater than or equal to the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, and an emergency stop action is directly executed.

[0091] In this embodiment, optionally, the first preset deviation is used to distinguish between normal deviations that allow for cutting and moderate deviations that require manual intervention. For example, the first preset deviation can be exemplarily set to 0.008 mm. The second preset deviation is used to distinguish between moderate deviations that require manual intervention and severe deviations that require emergency stopping. For example, the second preset deviation can be exemplarily set to 0.01 mm. It should be noted that the first preset deviation is smaller than the second preset deviation.

[0092] Optionally, the control module compares the actual position coordinate deviation of the circuit board with a first preset deviation. If the absolute value of the actual position coordinate deviation is less than or equal to the first preset deviation, it indicates that the offset between the current placement position of the circuit board and the theoretical reference position is within an acceptable range. Even if cutting is performed according to the current deviation, the board separation path will not exceed the allowable error tolerance. In this case, the control module determines that the actual position coordinate deviation of the circuit board meets the cutting conditions. The device generates the board separation path and executes the board separation action directly without any manual intervention.

[0093] Optionally, when the actual position coordinate deviation is greater than the first preset deviation but less than the second preset deviation, it indicates that the placement position of the circuit board has shifted to a certain extent. Although this shift has not reached the level that would cause a serious accident, it has exceeded the acceptable safety range for automatic board separation. If the automatic cutting continues, it may cause a significant deviation in the board separation path, affecting the board separation quality. In this case, the control module determines that the actual position coordinate deviation of the circuit board does not meet the cutting conditions. The device will not automatically execute the board separation action, but will take the following response measures: First, suspend the board separation process to prevent the cutter from continuing to move; second, trigger the audible and visual alarm device to alert the on-site operator. At this time, manual intervention is required to confirm the current placement status of the circuit board and determine whether it can be manually adjusted or repositioned before continuing the board separation. Only after manual confirmation and handling can the device continue to execute the board separation action.

[0094] Optionally, when the actual position coordinate deviation is greater than or equal to the second preset deviation, it indicates a serious offset in the placement of the circuit board. This level of deviation may stem from reasons such as a malfunction in the feeding mechanism, fixture failure, or severe deformation of the circuit board. If the cutting is forced under these conditions, the tool may directly collide with components, fixtures, or other obstacles on the circuit board, leading to serious consequences such as tool breakage, circuit board scrapping, or even equipment damage. In this case, the control module determines that the actual position coordinate deviation of the circuit board does not meet the cutting conditions and is classified as the highest level of anomaly. The device will not wait for manual confirmation but will immediately execute an emergency stop: sending a software emergency stop command to the execution module within a short time to stop the tool's movement and reset it to a safe position. Simultaneously, the power supply can be cut off to ensure the safety of the equipment and operators.

[0095] With the above settings, the control module automatically and consistently makes a judgment based on the comparison between the actual position coordinate deviation obtained from the fusion positioning and the preset threshold, which improves the uncertainty of manual judgment and enhances the uniformity and reproducibility of the judgment standard.

[0096] It is important to understand that the cutting speed and cutting force requirements differ significantly for circuit boards of different thicknesses.

[0097] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 9 If the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation, determining that the actual position coordinate deviation of the circuit board meets the cutting conditions and performing the board separation action according to the board separation path also includes: S331. Obtain the contact pressure value between the tool and the circuit board during the board separation process; S332. Adjust the separation speed and contact pressure value according to the thickness of the circuit board; During the separation process, the contact pressure value is adjusted by adjusting the depth of cut, and the fluctuation value of the contact pressure value relative to the preset target contact pressure value is kept within the first preset range.

[0098] In this embodiment, optionally, during the board separation process, the control module collects the contact pressure value between the tool and the circuit board in real time through a pressure sensor configured on the execution module. This pressure sensor can be installed at the tool spindle or tool holder, and can continuously monitor the axial or radial contact force experienced by the tool during cutting at a high sampling frequency. The contact pressure value is an important parameter characterizing the current cutting state. Insufficient pressure may indicate insufficient cutting depth, leading to incomplete cutting; excessive pressure may cause accelerated tool wear, burrs on the cutting edge, or even chipping.

[0099] Optionally, different thicknesses of circuit boards require different separation parameters. Thinner circuit boards have lower rigidity, and excessive contact pressure or cutting speed can easily lead to board deformation, edge burrs, or even breakage. Thicker circuit boards require greater cutting force and a more reasonable cutting speed to achieve smooth edges. Therefore, before initiating the separation process or during the separation operation, the control module first obtains the thickness information of the circuit board to be processed. This thickness information can be obtained in the following ways: reading preset board type parameters, measuring in real time through a thickness detection sensor, or manually input by the operator.

[0100] For example, for a thin plate with a thickness of 0.6 mm, the initial cutting speed can be set to 70 mm / s; for a plate with a thickness of 1.0 mm, the cutting speed can be adjusted to 50 mm / s; and for a thick plate with a thickness of 1.6 mm, the cutting speed can be further reduced to 30 mm / s. The purpose of this adjustment is that thicker plates require more sufficient cutting time, and appropriately reducing the speed can improve the cutting quality and reduce problems such as burrs or incomplete cutting caused by excessive speed.

[0101] Optionally, during the board separation process, even if the cutting speed and target contact pressure value have been preset according to the board thickness, fluctuations in the contact pressure value may still occur during the actual cutting process. The reasons for these fluctuations may include: slight non-uniformity in the board thickness, gradual changes in tool wear, and increased instantaneous resistance due to chip accumulation. To maintain stable cutting quality, when the contact pressure value is too high, the control module controls the execution module to appropriately reduce the cutting depth; when the contact pressure value is too low, the control module controls the execution module to appropriately increase the cutting depth to increase the contact force and ensure a smooth cutting edge.

[0102] In one example, when the contact pressure value is greater than the preset target contact pressure value, and the difference reaches or exceeds the upper limit of the preset first range, the control module controls the execution module to appropriately reduce the cutting depth; when the contact pressure value is less than the preset target contact pressure value, and the difference reaches or exceeds the upper limit of the preset first range, the control module controls the execution module to appropriately increase the cutting depth to increase the contact force and ensure a smooth cutting edge.

[0103] Optionally, the first preset range is exemplarily set to ±0.1N. The control module adjusts the depth of cut to ensure that the contact pressure value remains within the ±0.1N fluctuation range of the target contact pressure value.

[0104] By adjusting the separation speed and contact pressure based on the thickness, the separation parameters are adaptively matched to the board thickness. For example, the speed decreases by 10 mm / s for every 0.2 mm increase in board thickness. This allows the same separation device to automatically adapt to circuit boards of different thicknesses without manual parameter adjustment, improving the device's versatility and ease of use. Furthermore, it effectively reduces edge burrs caused by sudden pressure changes, resulting in smoother and flatter separation edges and improved separation quality.

[0105] According to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 10 The plate-splitting method also includes: S400. Acquire multi-dimensional operating status data, including at least two of the following: actual position coordinate deviation of the circuit board, contact pressure value between the tool and the circuit board, vibration amplitude of the equipment during operation, and operating temperature of the equipment.

[0106] In this embodiment, optionally, during the separation process, the control module collects various types of operational status data through sensing modules configured at various locations on the separation device. These data originate from different sensors and device modules, collectively forming a multi-dimensional description of the entire separation process. The multi-dimensional operational status data includes the following four categories: the actual position coordinate deviation of the circuit board, used to characterize the degree of positional offset of the current circuit board relative to a preset reference coordinate system; the contact pressure value between the cutter and the circuit board, used to characterize the real-time contact force between the cutter and the circuit board during the cutting process; the vibration amplitude of the equipment operation, derived from vibration sensors installed on the separation device, used to characterize the mechanical vibration intensity of the equipment during operation; and the operating temperature of the equipment, derived from temperature sensors installed on key heat-generating components such as the control module, drive motor, and spindle, used to characterize the thermal state of the equipment. Optionally, each operational status data point is preset with a corresponding risk threshold, and the control module determines different risk levels based on the number and type of status data exceeding the risk threshold.

[0107] S410. Compare the multi-dimensional operating status data with the preset threshold to determine the anomaly level, and execute the corresponding response action based on the anomaly level.

[0108] In this embodiment, optionally, the anomaly level includes Level 1 anomaly, Level 2 anomaly, and Level 3 anomaly; the response actions include recording the anomaly and outputting a prompt signal, pausing the sub-board and setting an alarm, and performing software emergency stop and hardware reset power-off within a preset time.

[0109] In one embodiment, the steps of comparing multi-dimensional operational status data with preset thresholds to determine the anomaly level, and executing corresponding response actions based on the anomaly level, include: S411. If the vibration amplitude of the equipment is lower than the first preset vibration amplitude threshold or the operating temperature is lower than the preset temperature value, it is determined to be a level one abnormality. The response action based on the level one abnormality is to record the abnormality and output a prompt signal. S412. If the actual position coordinate deviation of the circuit board is greater than the second preset deviation and less than the third preset deviation, or the fluctuation value of the contact pressure value relative to the preset target contact pressure value is within the second preset range, it is determined to be a level two abnormality. The response action based on the level two abnormality is to suspend board separation and alarm. S413. If the vibration amplitude is greater than the second preset vibration amplitude threshold or the tool breaks / the plate gets stuck, it is determined to be a level three abnormality. The response action based on the level three abnormality is to perform software emergency stop and hardware reset power-off within a preset time.

[0110] In this embodiment, optionally, after acquiring the aforementioned multi-dimensional operating status data, the control module compares each type of data with its corresponding preset threshold to comprehensively determine the current anomaly level of the device. The anomaly level is used to characterize the severity of the current anomaly, and different levels correspond to different response actions.

[0111] Optionally, the control module monitors the vibration amplitude and operating temperature of the equipment. A level one anomaly is determined when one of the following conditions is met: First, the vibration amplitude of the equipment is lower than a first preset vibration amplitude threshold.

[0112] For example, the first preset vibration amplitude threshold is 0.5g. When the real-time vibration amplitude value measured by the vibration sensor is lower than 0.5g, it indicates that the equipment is operating smoothly. However, if this value fluctuates abnormally compared to historical normal values ​​or is in a critical state, the device classifies it as a Level 1 anomaly. It should be noted that "lower than" here means that the vibration amplitude is within the normal range but may have a slow upward trend, or although it has not exceeded the safety upper limit, it has deviated from the benchmark value. As an alternative implementation, it can also be set to a situation where the vibration amplitude exceeds a certain normal fluctuation upper limit but has not yet reached the danger threshold. Those skilled in the art can adjust it according to the actual equipment characteristics. Second, the operating temperature is lower than the preset temperature value. For example, the preset temperature value is set to 85℃, which is determined according to the operating temperature specification of the embedded processor of the control module. When the operating temperature measured by the temperature sensor is lower than 85℃, it indicates that the thermal state of the equipment is normal; however, if the temperature shows an abnormally rapid rise or a trend close to this threshold, the device classifies it as a Level 1 anomaly and records it.

[0113] Optionally, in the event of a Level 1 anomaly, the control module's response is to record the anomaly and output a warning signal. For example, the device stores information such as the time, type, and relevant data values ​​of the anomaly in the data storage module for subsequent production review and trend analysis. Simultaneously, the device outputs a warning signal, such as displaying a yellow warning message on the operating interface, illuminating a warning light, or sending a reminder message to the remote monitoring platform. This level of anomaly will not interrupt the ongoing board separation operation; the board separation process continues normally.

[0114] Optionally, the control module monitors the actual position coordinate deviation of the circuit board and the fluctuation of the contact pressure value between the tool and the circuit board. A level two anomaly is determined when one of the following conditions is met: First, the actual position coordinate deviation of the circuit board is greater than a second preset deviation but less than a third preset deviation. For example, the second preset deviation is set to 0.008 mm, and the third preset deviation is set to 0.01 mm. When the actual position coordinate deviation is greater than 0.008 mm but less than 0.01 mm, it indicates that the positioning deviation of the circuit board exceeds the safe range for automatic board separation, but has not yet reached the level that would cause a serious accident. Second, the fluctuation value of the contact pressure value relative to the preset target contact pressure value is within a second preset range. This preset target contact pressure value is preset in advance by the R&D personnel within the control module, or a first generation logic is pre-stored within the control module so that the control module determines the target contact pressure value based on the first generation logic and relevant parameters such as the board type. In one example, the second preset range is set to ±0.5 N. When the fluctuation value of the contact pressure measured by the pressure sensor relative to the preset target contact pressure value exceeds the normal fluctuation range but is still within ±0.5N, it indicates that there are unstable factors in the cutting process, such as uneven plate thickness or local hard spots, but not yet reaching a seriously abnormal level. The normal fluctuation range can be a first preset range exceeding ±0.1N.

[0115] Optionally, in the event of a Level 2 anomaly, the control module's response is to pause the board splitting process and trigger an alarm. For example, the device immediately sends a pause command to the execution module, causing the cutter to stop cutting and remain in its current position or reset to a safe position. Simultaneously, the device triggers an audible and visual alarm, such as a buzzer and flashing alarm lights, to alert the on-site operator. The operator needs to intervene to confirm the cause of the anomaly and take appropriate action, such as repositioning, inspecting the cutter, and cleaning up chips. Only after manual confirmation and deactivation of the alarm can the device resume board splitting operations.

[0116] Optionally, the control module continuously monitors the vibration amplitude of the equipment and detects the presence of sudden serious events. A Level 3 anomaly is determined when one of the following conditions is met: First, the vibration amplitude exceeds a second preset vibration amplitude threshold. For example, the second preset vibration amplitude threshold is set to 2g, where g is the acceleration due to gravity. When the real-time vibration amplitude value measured by the vibration sensor exceeds 2g, it indicates that the equipment has experienced severe vibration, which may be caused by tool breakage, board jamming, collision, or other serious mechanical failures. If severe vibration persists, it will cause irreversible damage to the equipment's accuracy and service life, and may even lead to safety accidents. Second, the criteria for determining a sudden serious event: detection of tool breakage or board jamming. These events can be detected in various ways, such as: sudden changes in vibration amplitude, abnormal drops or spikes in contact pressure values, and abnormal feedback from position sensors. When the control module confirms through comprehensive judgment that a tool breakage or board jamming event has occurred, it immediately determines it as a Level 3 anomaly, regardless of whether the current vibration amplitude exceeds 2g.

[0117] Optionally, in the event of a Level 3 anomaly, the control module's response is to execute a software emergency stop and a hardware reset power-off within a preset time. For example, within the preset time after detecting a Level 3 anomaly, the control module sends a software emergency stop command to the execution module, immediately halting all movement of the tool and transmission module. Simultaneously, the device triggers a hardware reset power-off mechanism, cutting off the power supply to the execution module and resetting the tool, ensuring the equipment does not unexpectedly restart. This dual protection mechanism of software command and hardware power-off can bring the equipment to a safe state in the shortest possible time, minimizing accident losses. In one example, the preset time can be 200ms.

[0118] Through the above settings, this multi-dimensional detection mechanism can more comprehensively and accurately reflect the health status of the equipment, effectively reducing the problem of missing important abnormal information due to monitoring a single parameter. Furthermore, by setting differentiated response actions corresponding to different levels, a tiered handling strategy of "minor abnormalities not interrupting production, moderate abnormalities requiring manual intervention, and severe abnormalities requiring emergency shutdown" is achieved, striking a good balance between ensuring production continuity and guaranteeing safety.

[0119] It is important to understand that traditional panel separation methods typically involve quality inspection of the finished product only after panel separation is completed. By the time problems are discovered, they are often irreparable, and the panel must be scrapped.

[0120] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 11 The multi-dimensional operational status data also includes: the offset of the board cutting trajectory; The steps of comparing multi-dimensional operational status data with preset thresholds to determine the anomaly level and executing corresponding response actions based on the anomaly level also include: S414. If the offset of the plate cutting trajectory is greater than the preset offset but has not reached the preset scrap offset threshold, trigger the finishing action. The finishing move includes: Reacquire the edge feature image of the circuit board to be processed, compare the reacquired edge feature image of the circuit board with the template in the visual library, and determine the actual position coordinate deviation of the reacquired circuit board. The tooling path is generated based on the actual position coordinate deviation and preset constraints. The positional error between the tooling path and the board separation path at the break point is less than the preset error. Perform the last-hit action according to the last-hit path.

[0121] In this embodiment, optionally, during the separation process, the control module controls the sensing module to acquire dynamic images of the separation area. These dynamic images record the state of the cutting tool during the cutting process, including the actual position of the cutting trajectory and the morphological characteristics of the cut. Unlike acquiring reference point images before separation, the purpose of acquiring dynamic images of the separation process is to monitor whether the cutting trajectory is accurately executed according to the planned path. For example, an industrial camera continuously captures images of the cutting area at a preset frame rate of 30fps and transmits the images to the control module for analysis in real time.

[0122] Optionally, the control module analyzes the acquired dynamic images of the board separation process, extracts the position information of the actual cutting trajectory, and compares it with the theoretical board separation path or the generated board separation path to calculate the board separation cutting trajectory offset. For example, the control module can use image matching or edge detection algorithms to extract the actual position of the tool cut from the dynamic image. Then, it compares the actual position with the theoretical position of the corresponding point in the planned path and calculates the spatial distance difference between the two, which is the board separation cutting trajectory offset. This offset can be expressed as an X-direction offset, a Y-direction offset, or a combined offset distance. It should be noted that the cutting trajectory offset may be caused by various reasons, such as: slight displacement of the circuit board due to loosening of the fixture during the board separation process, uneven force on the tool causing path deviation, or residual errors in the initial positioning that have not been fully compensated.

[0123] Optionally, the control module compares the calculated board-separation cutting trajectory offset with two thresholds: a preset offset, which is the minimum offset to trigger corrective cutting, and a preset scrap offset threshold, which is the critical offset for determining scrap. The preset offset is used to determine whether the offset has reached a level requiring remediation. Small offsets may not have a substantial impact on the board-separation quality and do not require corrective cutting. For example, an exemplary preset offset value of 0.01mm is used. When the board-separation cutting trajectory offset is greater than 0.01mm, it indicates that the cutting has produced a significant deviation and requires remediation.

[0124] Optionally, a preset scrap offset threshold is used to determine whether the offset is so severe as to be irreparable. If the offset is too large, even with additional cutting, the quality of the separated board cannot be restored. In this case, the board should be directly deemed scrapped, and no resources should be wasted on additional cutting. For example, the preset scrap offset threshold is 0.03mm. When the offset of the board cutting trajectory is greater than 0.03mm, the device directly determines that the board is an unrepairable defective product, triggering the scrapping process, such as marking and sorting, and no longer performing additional cutting.

[0125] Optionally, when the condition of preset offset < board cutting trajectory offset < preset scrap offset threshold is met, the control module determines that the current cut has a significant deviation but has not yet reached the scrap level, and there is a possibility of remediation, triggering a supplementary cutting action. In one example, the board cutting trajectory offset is 0.015mm, which is between 0.01mm and 0.03mm, and the control module determines that the current cut has a significant deviation but has not yet reached the scrap level.

[0126] Optionally, before performing the finishing cut, the device needs to re-acquire the current accurate position of the circuit board. This is because the circuit board may have shifted slightly during the initial separation process, or the original positioning information may no longer be sufficient to guide the high-precision finishing cut operation. Therefore, the control module controls the industrial camera to re-acquire the reference point image of the circuit board and, following the method described above, redetermines the actual position coordinate deviation of the circuit board at the current moment. The accuracy requirement for this repositioning process is usually higher than that for the initial positioning; for example, the calibration error for repositioning should not exceed ±0.005 mm.

[0127] Optionally, based on the actual position coordinate deviation obtained from the repositioning, the control module generates a tooling path according to the above constraints. It should be noted that the tooling path needs to achieve a smooth and precise connection with the already completed panel separation path at the junction.

[0128] For example, the starting point of the finishing cut path is the point where the initial cutting on the panel separation path is interrupted, i.e., the end point of the current actual cut. The finishing cut path needs to start from this interruption point and continue to complete the remaining uncut or incompletely cut portion. To ensure that the overall cut after finishing cut is continuous and without steps, the positional error between the finishing cut path and the original panel separation path at the point of connection needs to be less than a preset error, which is exemplarily set at ±0.003mm. This results in a smooth and continuous cut after finishing cut, reducing the problem of new steps or misalignment defects caused by finishing cut.

[0129] Optionally, after generating the supplementary cutting path, the control module sends a supplementary cutting command to the execution module, and the execution module performs a secondary cutting action according to the supplementary cutting path. The supplementary cutting action needs to be completed in a short time to avoid affecting the overall production cycle. For example, the response time of the supplementary cutting action should not exceed 500ms.

[0130] With the above setup, the device can automatically correct cutting deviations without interrupting production or relying on manual intervention. This mechanism allows panels that might otherwise be scrapped due to cutting misalignment to be salvaged, reducing the scrap rate of panels during the separation process and saving production costs.

[0131] It is important to understand that traditional PCB depaneling defective product handling relies on manual judgment. Defective situations can only be recorded manually, and a structured defective product database cannot be formed. This leads to large fluctuations in defective product judgment results, affecting the consistency and reliability of product quality, and lacking reliable data support for subsequent quality analysis and cause tracing.

[0132] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 12 After performing the cutting and separating action on the circuit board to be processed, the method of updating the separation processing parameters based on the comparison results between the circuit board's cut image and the preset qualified separation image templates in the vision library, and then performing the separation action according to the processing parameters, includes: S510. Obtain the image of the plate cutting edge and compare the image of the plate cutting edge with the preset qualified plate image template in the vision library; S520. If the current board-making result is determined to be a defective product, extract the defective features from the cut image of the defective product, compare the defective features with the existing defective features in the visual library, and determine the new type of defective features. S530: Update the processing parameters of the board based on the type of defect features, and perform the board separation action according to the processing parameters.

[0133] In this embodiment, optionally, after completing the positioning calibration and path planning of the target circuit board, the control module controls the execution module to perform the board separation. After the board separation is completed, the device enters the defective product detection stage.

[0134] Optionally, the control module controls the sensing module to acquire images of the cut edges of the target circuit board after the separation process. The images of the cut edges should clearly show the complete shape of the separation cut, including details such as the edge contours, burrs, and degree of offset.

[0135] Optionally, after acquiring the slab cutting image, the control module compares the acquired slab cutting image with a preset qualified slab cutting image template in the vision library. The qualified slab cutting image template is a standard cutting image of the slab type after successful slab cutting, representing the quality benchmark for qualified slab cutting. The comparison methods may include: calculating the positional deviation between the cutting edge and the standard template edge, detecting the burr length in the cutting area, and analyzing the continuity and integrity of the cutting.

[0136] Optionally, if the current panelization result is determined to be a defective product, the control module analyzes the cut image of the defective product and extracts defective features that can characterize the defect type of the defective product. The extracted feature types include, but are not limited to: contour features, morphological features, and positional features.

[0137] In one example, the contour features can be geometric attributes such as the shape, edge curvature, and area size of the defect area; the morphological features can be the direction of burr extension and the shape of the chipped edge; and the location features can be the location of the defect on the circuit board, such as the edge, corner, or rib connection.

[0138] Optionally, the control module compares the extracted defective features with the feature vectors of known defect types already stored in the visual database. This comparison method can perform similarity calculations. The similarity can be measured using methods such as Euclidean distance or cosine similarity; a higher similarity indicates a higher degree of matching between the current defective feature and the defective features of known defect types.

[0139] Optionally, if the similarity is greater than or equal to a preset similarity threshold, it indicates that the current defective feature highly matches a certain defect type already existing in the visual library, belonging to a known type of defective feature. In this case, the device can directly identify the type of the defective product, such as "burr defect," "offset defect," or "tool wear defect," without needing to register a new type. If the similarity is less than the preset similarity threshold, it indicates that the current defective feature differs significantly from all known defect types in the visual library, belonging to a new type of defective feature. In this case, the control module identifies the defective feature as a new type and prepares to add it to the visual library to expand the coverage of the device's visual library.

[0140] It is important to understand that regardless of whether the current defect is a new type or a known type, the control module will update and optimize the processing parameters of subsequent board separations based on the type of defect. By adjusting the processing parameters, the probability of the same type of defect occurring is reduced.

[0141] By using the above settings, the panel cutting images are quantitatively compared with qualified panel image templates in the visual library, achieving automated and standardized inspection of panel quality. This improves both inspection efficiency and the consistency and objectivity of the judgment criteria. Furthermore, by comparing the similarity of defective features with existing defective features in the visual library, the device can identify defect types or discover new types of defective features, providing accurate classification criteria for subsequent targeted optimization.

[0142] It is important to understand that traditional manual visual inspection for defective products relies on personal experience, and the judgment criteria may differ between different operators or even between the same operator at different times.

[0143] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 13 The steps of comparing the image of the panel cutting edge with the preset qualified panel image template in the visual library include: S511. If the burr length at the depanel cut exceeds a preset length value, the depaneled circuit board is determined to be defective; or, S512. If the offset value of the board separation cut is greater than the preset offset value, the board after separation is determined to be a defective product.

[0144] In this embodiment, it is important to understand that burrs are excess protrusions or rough parts generated at the cut edge during the board separation process, and are one of the important indicators for measuring the quality of board separation. Excessively long burrs may affect the risk of the circuit board detaching during subsequent use, leading to an electrical short circuit.

[0145] Optionally, the control module performs edge extraction on the image of the board separation cut to detect burr features at the cut edge. For example, the device identifies the edge line of the cut and then measures the maximum length of the burr extending outward from the cut edge. In one example, the control module compares the detected burr length with a preset length value of 0.02 mm: if the burr length detected in the board separation cut image is greater than 0.02 mm, it is determined that the burr length exceeds the standard, and the board after separation is determined to be defective; if the burr length is less than or equal to 0.02 mm, the burr index is determined to be qualified, and the judgment of other indicators continues.

[0146] It's important to understand that cut offset refers to the degree of deviation between the actual cutting position and the theoretical board separation path. Excessive cut offset may cause the cutting path to intrude into the component area, the cutting line to deviate from the predetermined position and affect the functional area of ​​the circuit board, or result in excessive deviation in the dimensions of the board edge after separation.

[0147] Therefore, in one example, the control module compares the detected cut offset value with the preset offset value of 0.01mm: if the cut offset value detected in the board separation cut image is greater than 0.01mm, it is determined that the offset exceeds the standard and the board after separation is determined to be a defective product; if the cut offset value is less than or equal to 0.01mm, the offset index is determined to be qualified.

[0148] It should be noted that this preset offset value is set based on the positioning accuracy requirements of the PCB depaneling industry, representing the upper limit of the allowable cut position deviation for qualified depaneling. Cut offset determination can be based on the maximum offset value, the average offset value, or the cumulative length proportion of offsets exceeding the preset value. For example, the device continuously monitors the offset distribution along the entire cut length; if any point has an offset value exceeding the preset offset value by 0.01mm, it is determined to be a defective product, ensuring the overall consistency of depaneling quality.

[0149] By setting specific preset length and offset values, the determination of defective products is transformed into an objective numerical comparison. This quantitative standard improves the consistency and reproducibility of the determination results and eliminates the uncertainty caused by subjective judgment.

[0150] According to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 14 If the current panelization result is determined to be a defective product, the steps of extracting defect features from the cut image of the defective product, comparing the defect features with existing defect features in the visual database, and identifying new types of defect features include: S521. For the cut image of the product identified as defective, extract the contour features, morphological features and position features of the cut, and generate the corresponding defect feature vector. S522. Calculate the similarity between the defective feature vector and the known types of defective feature vectors stored in the visual database; S523. If the similarity is less than the preset similarity threshold, determine the bad feature as a new type; S524. If the similarity is greater than the preset similarity threshold, determine the bad feature as a known type.

[0151] In this embodiment, optionally, the control module extracts contour features from the cut image to obtain contour information of the cut area. Contour features include, but are not limited to, the geometric shape of the cut area, such as the sharp shape of burrs, the arc shape of chipped edges, the triangle of missing corners, the curvature change of the contour, the area size and perimeter of the defect area, etc., reflecting the geometric shape of the cut defect.

[0152] Optionally, the control module extracts the morphological features of the cut. The morphological features include, but are not limited to: the extension direction of the burr, the length-to-width ratio of the burr, the distribution pattern of the offset, the depth of the chipped edge and the opening angle, etc., which reflect the generation mechanism of the cut defect. For example, uniform offset may be due to positioning deviation, while local offset may be due to local deformation of the plate or interference from foreign objects.

[0153] Optionally, the control module analyzes the location of the cut defect on the circuit board. Location characteristics include, but are not limited to: the relative coordinates of the cut defect from the circuit board reference point, whether the cut defect is located at the edge of the board or inside the board, whether the cut defect is located on a straight cut segment or a curved / corner cut segment, and whether the cut defect is located at a rib connection. Location characteristics help determine whether the cut defect is related to a specific structure; for example, burrs are more likely to occur at corners, and chipping is more likely to occur at rib connections.

[0154] It is important to understand that after completing the feature extraction described above, the control module integrates the extracted features to generate one or more defective feature vectors. Optionally, the defective feature vector is a numerical representation of the features, for example, it can be an N-dimensional vector, where each dimension represents a specific feature value, such as burr length, offset distance, curvature value, or position coordinates. This vectorized representation facilitates subsequent similarity calculation and comparison.

[0155] Optionally, the control module calculates the similarity between the generated defective feature vector and the feature vectors of each known defect type stored in the visual library. The known defect types in the visual library are confirmed and classified defective feature templates accumulated during historical segmentation processes, with each defect type corresponding to one or more feature vectors. Optionally, the similarity can be measured using various standard algorithms; for example, Euclidean distance can be used as a similarity metric, where a smaller Euclidean distance indicates greater similarity between the two feature vectors.

[0156] Optionally, if the similarity is greater than or equal to a preset similarity threshold, it indicates that the current defective feature vector highly matches the feature vector of a known defect type in the visual library, and the similarity between the two reaches a level that can be identified as belonging to the same type. In this case, the control module determines that the current defective feature is of a known type and classifies it into the defect type with the highest similarity. If the similarity is less than the preset similarity threshold, it indicates that the current defective feature vector does not meet the identification criteria for similarity with the feature vectors of all known defect types in the visual library, and the difference between the two is significant enough that it cannot be classified into any existing category. In this case, the control module determines that the current defective feature is a new type of defective feature.

[0157] It should be noted that the preset similarity threshold is set based on engineering experience in classifying defect features in the PCB depaneling field. Generally, the higher the preset similarity threshold, the more likely the device is to classify the defect as a new type; the lower the preset similarity threshold, the more likely the device is to classify the defect type as a known type.

[0158] Through the above settings, this application can automatically determine whether the current defect feature is a recurrence of a known type or the first discovery of a new type of defect without manual intervention. For known types, the device can directly apply existing optimization strategies for a rapid response; for new types, the device can trigger a supplementary data entry process to expand the knowledge base. This distinguishing ability reduces the probability of incorrectly classifying new types of defects into known types, leading to optimization strategy mismatch, and also reduces the probability of visual library redundancy caused by repeatedly registering known types of defects as new types.

[0159] According to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 15 The machining parameters for the panel separation include tool pressure, positioning error threshold for tool entry determination, and cutting speed. The steps for updating the machining parameters for the panel separation based on the type of defect feature include: S531. When the type of defect is tool wear, reduce the tool pressure; S532. When the type of defective feature is plate position offset, reduce the positioning error threshold for tool entry determination from the first error threshold to the second error threshold. S533. If the type of defective feature is burrs, reduce the cutting speed.

[0160] In this embodiment, optionally, tool wear defects typically manifest as a gradual deterioration in the quality of the slit as the tool's usage time increases, resulting in increased burrs, rough cuts, and even chipping. This is because as the tool edge dulls, the cutting resistance increases under the same pressure, leading to a decrease in cut quality. When the control module determines that the current defect is of the tool wear type, the parameter update operation performed is to reduce the tool pressure.

[0161] For example, the reduction in tool pressure can be from 0.5N to 1N. For instance, if the current tool pressure is set to 18N and tool wear is detected, the control module will adjust the tool pressure to 17N or 17.5N. The purpose of reducing tool pressure is to reduce the cutting load on the worn tool during the cutting process, reduce the problem of chipping or tool breakage caused by excessive pressure, and improve the cut quality to some extent by reducing the pressure.

[0162] Optionally, board position offset defects typically manifest as a mechanical deviation of the board separation cut from the theoretical path, with a consistent direction and relatively stable offset amount. This is due to insufficient initial positioning accuracy of the circuit board. Under the existing cutting judgment positioning error threshold, boards with a certain degree of offset are allowed to enter the board separation process, leading to a deviation in the cutting position. When the control module determines that the current defect is of the board position offset type, the parameter update operation performed is to reduce the cutting judgment positioning error threshold from the first error threshold to the second error threshold.

[0163] For example, a first error threshold is set to 0.01mm, and a second error threshold is set to 0.008mm. The first error threshold is the initial cutting judgment standard; when the actual position coordinate deviation of the circuit board is less than or equal to 0.01mm, the device allows cutting. After detecting a board position offset defect, the control module reduces the cutting judgment positioning error threshold from 0.01mm to 0.008mm. This means that subsequent circuit boards can only be allowed to be cut if the actual position coordinate deviation is less than or equal to 0.008mm. By reducing the cutting judgment positioning error threshold, the device will require more stringent standards for the positioning accuracy of subsequent circuit boards, thereby reducing offset defects caused by positioning deviations.

[0164] Optionally, burr defects typically manifest as excess protrusions at the edges of the board separation cut, resulting in an uneven cut. This is caused by excessively high cutting speeds, leading to insufficient cutting action between the tool and the circuit board, causing the material to be torn instead of cleanly cut, forming burrs. If the control module determines that the current defect is of the burr type, the parameter update operation performed is to reduce the cutting speed.

[0165] For example, the cutting speed can be reduced by 10 mm / s to 20 mm / s. If the current cutting speed is set to 70 mm / s and burr-like defects are detected, the control module will adjust the cutting speed to 50 mm / s or 60 mm / s. The purpose of reducing the cutting speed is to increase the cutting time between the tool and the circuit board, allowing for more thorough cutting and resulting in smoother, cleaner cut edges, thus reducing burr formation.

[0166] Through the above settings, this application designs targeted parameter adjustment strategies for defects such as tool wear, plate misalignment, and burrs. This precise optimization method ensures that each parameter adjustment directly addresses the root cause of the problem, improving the effectiveness of the optimization measures. Furthermore, the quantification of parameters provides the control module with clear execution instructions, eliminating the need for manual judgment of the adjustment amount and automating parameter optimization.

[0167] It should be understood that simply detecting defective features and adjusting the depaneling parameters may fall into a local optimization trap. Therefore, in order to improve the accuracy of depaneling parameter optimization and contribute to improving and stabilizing the yield of the depaneled circuit board, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 16 After updating the processing parameters for the separation board and performing the separation action according to the processing parameters, the method further includes: S534. If the yield rate of the processed circuit board is greater than or equal to the preset yield rate, maintain the updated board separation parameters. S535. If the yield rate is less than the preset yield rate, restore the board separation parameters to the parameter values ​​before the update.

[0168] In this embodiment, optionally, after the control module updates the processing parameters according to the type of defective characteristics, the device enters an observation period to continuously monitor the yield rate of subsequent batches of circuit boards. The yield rate is calculated as the proportion of circuit boards judged as good within a certain production batch or quantity to the total number of boards separated. For example, if 100 circuit boards are produced continuously, and 95 are judged as good, the yield rate is 95%. Optionally, the separation device can determine the yield rate of the separated circuit boards based on the content of some of the above embodiments. Alternatively, the yield rate can be entered periodically by R&D personnel or testing personnel.

[0169] Optionally, the preset yield rate can be preset in the control device by the R&D personnel according to the requirements, such as 95%.

[0170] Optionally, if the yield rate of the processed circuit boards is greater than or equal to the preset yield rate, the control module determines that the current parameter adjustment is effective and the board separation quality of the device has reached a stable state. In this case, the control module maintains the updated board separation parameters and continues to use them as the benchmark parameters for subsequent board separation operations. Meanwhile, the device can continue to monitor the yield rate, but will no longer actively adjust the maintained parameters unless defective products are detected again.

[0171] Optionally, if the yield rate is lower than the preset yield rate, the control module determines that the current parameter adjustment may be excessive or have side effects, leading to a deterioration in the board separation quality. The criteria for judging a decline in yield rate may include: the yield rate falling below the preset target value, the yield rate showing a continuous downward trend, and the detection of frequent recurrence of defect types related to previous optimization targets.

[0172] In one example, a yield rate falling below the preset target value can be defined as a decrease from over 95% to below 90%, and a continuous downward trend in the yield rate can be defined as a decrease in the yield rate of three consecutive batches.

[0173] If the yield rate drops, the control module will restore the parameters of the sub-board to their pre-optimization values. The aim is to minimize the negative impact of this parameter adjustment, returning the equipment to a relatively stable state and reducing the likelihood of widespread quality problems caused by a single unsuccessful parameter optimization.

[0174] The above settings ensure that only parameter adjustments that have been verified in production and can reliably improve quality are retained, reducing the quality risks associated with blind adjustments. Furthermore, if the yield rate decreases after adjustment, the device can automatically revert to the parameter values ​​before optimization. This reversible design allows the device to boldly attempt parameter optimization without worrying about long-term quality problems caused by a single failed adjustment, thus reducing the risk and cost of optimization exploration.

[0175] It's important to understand that in industrial environments, factors such as momentary light flickering, brief dust obstruction, and occasional image acquisition anomalies can lead to false defect features in a single image. If every newly detected feature is immediately added to the database, the visual database may be contaminated with a large amount of invalid, noisy data, causing subsequent defect judgment rules to become ineffective.

[0176] Therefore, according to some embodiments of this application, optionally, in one embodiment of this application, reference is made to... Figure 17 The step of updating the processing parameters of the depanel based on the type of defect features also includes: S536. Add new types of defective features to the visual database and update the defective product judgment rules; The steps for adding new types of defective features to the visual database and updating the defective product judgment rules include: S537. If a new feature is detected in a cut image, and the same new type of defective feature is detected in cut images for a consecutive preset number of frames, it is stored in the cache pool. S538. When the preset production batch is reached, the defective features in the cache pool are clustered, denoised, and checked for repetition before being added to the visual library and the defective product judgment rules are updated.

[0177] In this embodiment, it is important to understand that when the control module determines that a certain defective feature is a new type, the device will not immediately add it to the visual library. Instead, it first enters a controlled caching mechanism to prevent false defective features caused by a single accidental event from polluting the visual library, such as a momentary flash of light, a brief obstruction by dust, or an abnormal image acquisition.

[0178] Optionally, the device acquires subsequent cut images and checks whether the same new type of defect feature can be detected within a preset number of consecutive frames. For example, the preset number of frames is set to 3 frames. That is, the device needs to detect the same new type of defect feature, such as the same morphology of burrs or the same type of offset pattern, in 3 consecutive different cut images to consider the defect feature to be stable rather than accidental. Only then will the control module store the new type of defect feature in the cache pool.

[0179] Optionally, the cache pool is a temporary storage area used to temporarily store new types of defect features that have passed preliminary verification but have not yet been formally added to the database. The data stored in the cache pool includes defect feature vectors, corresponding cut images, detection time, board type information, etc.

[0180] Optionally, defective features stored in the cache pool are not immediately incorporated into the visual library, but require further batch verification and screening. The device sets a batch threshold, triggering a batch entry operation only after production reaches a preset batch size. For example, if the preset production batch size is 100 times, the device performs centralized processing on all candidate new types of defective features accumulated in the cache pool after every 100 board separation processes or every 100 circuit boards processed. The centralized processing steps include: clustering noise reduction and repeatability verification.

[0181] Optionally, the clustering denoising step includes: the device performing clustering analysis on the bad feature vectors in the cache pool, and classifying candidate features with high feature similarity into the same potential category. During the clustering denoising process, isolated features that cannot be classified into any category are considered as possible noise or occasional anomalies and are filtered out in the denoising stage, not entering the subsequent data entry process.

[0182] Optionally, the duplication verification step includes: for each potential new category formed after clustering, the device further verifies the consistency of its internal features. If there are significant differences between the feature vectors contained in the same category, such as features with excessively different morphologies being mixed into the same "new type," the device will split the category or mark it for manual review. At the same time, the device will also perform a secondary comparison between the potential new category and the known defective types already in the main visual library. When the similarity is lower than a preset threshold, it is confirmed that it is indeed significantly different from all existing types, reducing the probability of duplicate entry into the library.

[0183] Optionally, after clustering noise reduction and repeatability verification, new defect types that are confirmed to be real and stable will be batch-integrated into the main visual library. The integration operation includes: adding the defect feature vector of the new type to the visual library, and updating the defect judgment rules so that subsequent board separation quality inspection can identify this new type of defect.

[0184] By using the above-mentioned clustering and noise reduction settings for defective features, only consistently occurring defective features can enter the cache pool, ensuring that the visual library remains effective during long-term operation and will not gradually become ineffective due to various interference factors in the industrial field, thus enhancing the long-term operational robustness of the device.

[0185] It is important to understand that the board separation method also includes loading the visual library corresponding to the circuit board to be processed. The visual library includes the reference point features of the circuit board to be processed, qualified board separation images, and known defective features. The step of loading the visual library corresponding to the target circuit board further includes: Read the circuit board's preset parameters; If the circuit board is a new type, import a preset number of standard images, automatically extract features, and generate a unique visual library template.

[0186] Optionally, before loading the vision library, the control module first reads the preset parameters of the circuit board to be processed. These preset parameters can be obtained in various ways: automatically identifying the board type by reading the identification code on the circuit board, obtaining the board type information of the current batch of products from the production management device, or manually input by the operator. The identification code can be a QR code or a barcode. The preset parameters of the circuit board include, but are not limited to: board type number, board thickness, material, number and theoretical coordinates of reference points, component layout information, preset cutting speed range, preset tool pressure range, and positioning accuracy requirements. These parameters form the basis for subsequent vision library generation and board separation parameter settings. For example, the preset parameters are stored in the control module in the form of configuration files, with different configuration files corresponding to different board types. The control module automatically loads the corresponding configuration file based on the read board type identifier.

[0187] With the above setup, this device can accurately compare the acquired circuit board image with a pre-stored standard template to calculate the actual position coordinate deviation of the circuit board. Without this standardized reference benchmark, visual positioning would lack a reference point and would be unable to achieve high-precision deviation calculation.

[0188] Optionally, when the control module determines that the board type of the current circuit board to be processed does not have a corresponding template in the vision library, it identifies the circuit board as a new board type. The system then enters an initialization mode, imports a preset number of standard images (e.g., 3-5 standard images), and automatically generates a dedicated vision library template for the new board type without retraining the algorithm model or performing complex manual annotation. After the initialization is completed, the newly generated dedicated vision library template is stored in the vision library for direct loading and use during subsequent board separation operations of the same board type. This process does not require retraining the algorithm model or manual feature annotation, achieving rapid adaptation to the new board type.

[0189] By importing a preset number of standard images and automatically extracting features through the above settings, a custom visual library template for the new pattern type can be quickly generated. This process requires no retraining of the algorithm model or manual annotation, significantly shortening the preparation time for launching the new pattern type and improving the production line's ability to respond quickly to multiple product varieties.

[0190] This application also proposes a board splitting device, including a control module; the control module is used to execute the board splitting method as described above.

[0191] Optionally, the control module can be implemented using a main controller, such as an MCU, PLC, digital signal processing chip, device-level chip, or programmable gate array chip. The control module can establish bidirectional communication connections with both the sensing module and the execution module via industrial Ethernet, forming a closed-loop control architecture of "sensing-decision-execution".

[0192] Optionally, since the technical solution and its effects of the plate separating device have been described in detail in the description of the plate separating method, those skilled in the art can understand the specific working principle of the device, so it will not be repeated here.

[0193] The above description is merely an exemplary embodiment of this application and does not limit the patent scope of this application. Any equivalent structural transformations made based on the technical concept of this application and the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included within the patent protection scope of this application.

Claims

1. A plate separation method, characterized in that, The plate separation method includes: Obtain the edge feature image of the circuit board to be processed; The edge feature image of the circuit board to be processed is compared with the template in the visual library to determine the actual position coordinate deviation of the circuit board to be processed. A board-splitting path is generated based on the actual position coordinate deviation and preset constraints, and the board-splitting action is performed according to the board-splitting path. After performing the cutting and separating action on the circuit board to be processed, the processing parameters for separating the circuit board are updated based on the comparison results between the cut image of the circuit board and the preset qualified separating image template in the vision library, and the separating action is performed according to the processing parameters.

2. The plate-separation method as described in claim 1, characterized in that, The step of acquiring the edge feature image of the circuit board to be processed includes: Obtain a feature image of the circuit board to be processed, the feature image including a reference point image of the circuit board to be processed; The reference point image of the circuit board to be processed is sequentially subjected to grayscale conversion, Gaussian filtering, binarization, and edge extraction to obtain the edge feature image.

3. The plate separation method as described in claim 2, characterized in that, The steps of sequentially performing grayscale conversion, Gaussian filtering, binarization, and edge extraction on the reference point image of the circuit board to be processed to obtain the edge feature image include: The reference point image of the circuit board to be processed is converted from a color image to a grayscale image using the grayscale conversion formula fgray(x,y)=0.299R+0.587G+0.114B; where R, G, and B represent the values ​​of the red, green, and blue color channels of each pixel in the color image, respectively. The grayscale image is filtered using a 3×3 convolution kernel with a standard deviation σ=1.

2. The grayscale image is converted into a binary image based on a preset grayscale threshold. Based on a preset edge threshold, the edge features of the reference point are extracted.

4. The plate-separating method as described in claim 1, characterized in that, The board separation method further includes obtaining the coordinate offset data and tilt angle of the circuit board to be processed; The step of comparing the edge feature image of the circuit board to be processed with the template in the visual library to determine the actual position coordinate deviation of the circuit board to be processed is as follows: The edge feature image is compared with the templates in the visual database using a feature matching algorithm; Based on the comparison results, the center coordinates and tilt angle of the reference point of the circuit board to be processed are determined; The actual position coordinate deviation of the circuit board to be processed is determined based on the reference point center coordinates and tilt angle of the circuit board to be processed, and the coordinate offset data and tilt angle of the circuit board to be processed.

5. The plate-separation method as described in claim 4, characterized in that, The step of comparing the edge feature image with the template in the visual database using a feature matching algorithm includes: A scale space is constructed based on the SIFT algorithm and according to the first preset scale value, the second preset scale value, and the third preset scale value. In scale space, extreme points are detected, and the direction of the extreme points is determined by neighborhood gradient statistics to generate SIFT feature vectors. Calculate the Euclidean distance between the SIFT feature vector and the feature vector of the preset reference point template in the visual library; If the ratio of the first Euclidean distance to the second Euclidean distance is less than a preset ratio, the center coordinates and tilt angle of the reference point in the visual coordinate system are determined, where the first Euclidean distance and the second Euclidean distance are the minimum and second minimum values ​​of the Euclidean distance, respectively.

6. The plate-separation method as described in claim 1, characterized in that, The preset constraints include at least one of the following three: The minimum safe distance between the board separation path and the components shall not be less than a preset distance threshold. The cutting speed is within the preset speed range; The tool path direction is parallel to the edge or rib direction of the circuit board; Among them, when the preset constraint conditions include at least two of the three, the priority of the minimum safe distance between the board separation path and the component not being lower than the preset distance threshold is greater than the priority of the cutting speed being in the preset speed range, and the priority of the cutting speed being in the preset speed range is greater than the priority of the cutting direction being parallel to the board edge or rib direction of the circuit board.

7. The plate separation method as described in claim 6, characterized in that, The step of generating the partition path based on the actual position coordinate deviation and preset constraints further includes: The contact pressure value between the cutting tool and the circuit board during the depaneling process is obtained, wherein the contact pressure value is related to the degree of tool wear; The plate separation path is dynamically adjusted based on the contact pressure value.

8. The plate separation method as described in claim 7, characterized in that, The step of dynamically adjusting the plate separating path based on the contact pressure value includes: If the contact pressure value is less than the preset pressure value, the cutting is performed continuously according to a preset first length. When the contact pressure value is greater than or equal to the preset pressure value, the plate-splitting path is divided into sub-paths of no more than a preset second length, and chip removal is performed during the segmentation gap.

9. The plate-separation method as described in claim 1, characterized in that, The step of generating a partitioning path based on the actual position coordinate deviation and preset constraints, and performing partitioning actions according to the partitioning path, further includes: If the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation, it is determined that the actual position coordinate deviation of the circuit board meets the cutting conditions, and the board separation action is performed according to the board separation path. If the actual position coordinate deviation of the circuit board to be processed is greater than the first preset deviation but less than the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, the board separation is paused and an alarm is triggered. If the actual position coordinate deviation of the circuit board to be processed is greater than or equal to the second preset deviation, it is determined that the actual position coordinate deviation of the circuit board does not meet the cutting conditions, and an emergency stop action is directly executed.

10. The plate-separating method as described in claim 9, characterized in that, The step of determining that the actual position coordinate deviation of the circuit board meets the cutting conditions and performing the board separation action according to the board separation path when the actual position coordinate deviation of the circuit board to be processed is less than or equal to the first preset deviation further includes: Obtain the contact pressure value between the cutting tool and the circuit board during the board separation process; The speed of separating the boards and the contact pressure value are adjusted according to the thickness of the circuit board; During the separation process, the contact pressure value is adjusted by adjusting the depth of cut, and the fluctuation value of the contact pressure value relative to the preset target contact pressure value is kept within a first preset range.

11. The plate-separating method as described in claim 1, characterized in that, The plate-splitting method further includes: Acquire multi-dimensional operating status data, which includes at least two of the following: the actual position coordinate deviation of the circuit board, the contact pressure value between the tool and the circuit board, the vibration amplitude of the equipment during operation, and the operating temperature of the equipment. The multi-dimensional operating status data is compared with a preset threshold to determine the anomaly level, and the corresponding response action is executed based on the anomaly level.

12. The plate-separating method as described in claim 11, characterized in that, The anomaly levels include Level 1, Level 2, and Level 3 anomalies; the response actions include recording the anomaly and outputting a prompt signal, pausing the sub-board and issuing an alarm, and performing a software emergency stop and hardware reset power-off within a preset time. The steps of comparing the multi-dimensional operational status data with preset thresholds to determine the anomaly level, and executing corresponding response actions based on the anomaly level include: The multi-dimensional operating status data includes the vibration amplitude of the equipment and the operating temperature of the equipment. If the vibration amplitude is lower than the first preset vibration amplitude threshold or the operating temperature is lower than the preset temperature value, it is determined as a level one anomaly. The response action based on the level one anomaly is to record the anomaly and output a prompt signal. The multi-dimensional operating status data also includes the actual position coordinate deviation of the circuit board and the contact pressure value between the tool and the circuit board. If the actual position coordinate deviation is greater than the second preset deviation and less than the third preset deviation, or if the fluctuation value of the contact pressure value relative to the preset target contact pressure value is within the second preset range, it is determined to be a level two abnormality. The response action based on the level two abnormality is to suspend board separation and issue an alarm. If the vibration amplitude exceeds the second preset vibration amplitude threshold or the tool breaks / the plate gets stuck, it is determined to be a level three anomaly. The response action based on the level three anomaly is to perform a software emergency stop and a hardware reset power-off within a preset time.

13. The plate-separation method as described in claim 11, characterized in that, The multi-dimensional operational status data also includes: the offset of the board cutting trajectory; The steps of comparing the multi-dimensional operational status data with preset thresholds to determine the anomaly level and executing corresponding response actions based on the anomaly level further include: If the offset of the plate cutting trajectory is greater than the preset offset but does not reach the preset scrap offset threshold, a supplementary cutting action is triggered. The finishing move includes: Reacquire the edge feature image of the circuit board to be processed, compare the reacquired edge feature image of the circuit board with the template in the visual library, and determine the actual position coordinate deviation of the reacquired circuit board. The tooling path is generated based on the actual position coordinate deviation and preset constraints. The positional error between the tooling path and the board splitting path at the break point is less than the preset error. Perform the finishing move according to the stated finishing move path.

14. The plate-separating method as described in claim 1, characterized in that, The step of updating the board separation processing parameters based on the comparison result between the board's cut image and the preset qualified board separation image template in the vision library after performing the cutting action on the circuit board to be processed includes: Obtain an image of the panel cutting edge, and compare the image of the panel cutting edge with a preset qualified panel image template in the visual library; If the current board-making result is determined to be a defective product, the defect features are extracted from the cut image of the defective product, the defect features are compared with the existing defect features in the visual library, and new types of defect features are identified. The processing parameters of the partition are updated based on the type of defective feature.

15. The plate-separation method as described in claim 14, characterized in that, The step of comparing the image of the plate cutting edge with a preset qualified plate image template in the visual library includes: If the burr length at the depanel cut exceeds a preset length value, the depaneled circuit board is determined to be defective; or, If the offset value of the board separation cut is greater than the preset offset value, the board after separation is determined to be a defective product.

16. The plate-separation method as described in claim 14, characterized in that, The step of extracting defect features from the cut image of the defective product when the current board-making result is determined to be a defective product, comparing the defect features with existing defect features in the visual library, and determining new types of defect features includes: For cut images identified as defective products, extract the contour features, morphological features, and positional features of the cuts, and generate corresponding defect feature vectors; The similarity is calculated based on the defective feature vector and the known types of defective feature vectors stored in the visual database; If the similarity is less than a preset similarity threshold, the defective feature is determined to be a new type. If the similarity is greater than a preset similarity threshold, the defective feature is determined to be of a known type.

17. The plate-separating method as described in claim 14, characterized in that, The processing parameters for the plate separation include tool pressure, positioning error threshold for tool entry determination, and cutting speed. The step of updating the processing parameters for the plate separation based on the type of defect features includes: If the type of defect is tool wear, reduce the tool pressure; When the defective feature is of the plate position offset type, the positioning error threshold for determining the cutting edge is reduced from the first error threshold to the second error threshold; If the defective feature is of the burr type, reduce the cutting speed.

18. The plate-separation method as described in claim 17, characterized in that, After the steps of updating the processing parameters of the board and performing the board separation action according to the processing parameters, the method further includes: If the yield rate of the processed circuit board is greater than or equal to the preset yield rate, maintain the updated board separation parameters; If the yield rate is less than the preset yield rate, restore the board separation parameters to the parameter values ​​before the update.

19. The plate-separation method as described in claim 14, characterized in that, The step of updating the processing parameters of the partition plate based on the type of defect features further includes: Add the new types of defective features to the visual library and update the defective product judgment rules; The step of adding the new type of defective features to the visual library and updating the defective product judgment rules includes: If a new feature is detected in a cut image, and the same new type of defective feature is detected in cut images for a consecutive preset number of frames, it is stored in the cache pool. When the preset production batch is reached, the defective features in the cache pool are clustered, denoised, and checked for repetition before being added to the visual library and the defective product judgment rules are updated.

20. A plate separating device, characterized in that, The board separation device includes a control module; the control module is used to execute the board separation method as described in any one of claims 1-19.

Images