Chip mounting cooperative control system for flexible production

Abstract

The invention discloses a flexible production-oriented chip mounting cooperative control system, which comprises a sensing module, a decision module and an execution module, and is characterized in that improved Radon transformation and polar coordinate screening are adopted, and through dual-threshold segmentation and background compression, noise interference is reduced, and feature extraction robustness is enhanced. According to multi-resolution chip positioning, precise positioning of a chip in a machine coordinate system is achieved, an APF-RRT algorithm is improved, an attractive force-repulsive force field of an artificial potential field is combined with RRT random search, a collision-free path is dynamically planned, the path length, time and safety are optimized, master-slave cooperative control keeps the cooperative distance between devices, the flexible production requirement is met, and the flexible production requirement is met. A smooth acceleration curve is generated by adopting quintic polynomial interpolation, mechanical vibration is reduced, patch stability is improved, a closed loop is formed through visual positioning, path planning and dynamic control, self-adaptive adjustment of flexible production is achieved, and chip patch production closed loop management which is higher in precision, faster in response and higher in adaptability compared with a traditional method is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of chip placement collaborative control, and particularly to a chip placement collaborative control system for flexible production. Background Art

[0002] Chip placement is one of the key process links in electronic manufacturing, requiring the equipment to operate with extremely high precision and speed. The placement position of each chip and electronic component must be accurate, otherwise it will cause problems such as circuit short circuits and open circuits, affecting the reliability and performance of the product. The single assembly line mode often cannot handle complex production tasks under flexible production, and more intelligent control methods and coordination mechanisms need to be introduced. In a flexible production environment, the collaborative control system can make dynamic adjustments based on real-time data to cope with uncertain factors such as production plans and product model changes.

[0003] Currently, the Chinese invention patent with the application number CN202010183406.3 discloses a method and system for controlling the movement of a chip mounter based on a mathematical model, obtaining the static friction curve of the work head; determining the non-linear friction force according to the static friction curve; constructing a mathematical model for the movement of the chip mounter according to the dynamic equation and the non-linear friction force; and controlling the movement of the chip mounter according to the mathematical model. By controlling the movement of the chip mounter through the above method of the present invention, the accuracy of chip mounting by the chip mounter can be improved. However, the prior art cannot achieve sub-pixel level fiducial point center positioning by combining parabola fitting, cannot achieve multi-resolution chip positioning to accurately position the chip in the machine coordinate system, cannot dynamically plan a collision-free path while optimizing the path length, time, and safety, cannot maintain the collaborative distance between devices through master-slave collaborative control, and there is no three-dimensional dynamic operation coding to generate a three-dimensional trajectory path with Z-axis height, and cannot form a dynamic closed-loop of visual positioning (perception) → path planning (decision-making) → dynamic control (execution) to achieve adaptive adjustment of flexible production. Summary of the Invention

[0004] The technical problem solved by the present invention is that the prior art cannot achieve sub-pixel level fiducial point center positioning by combining parabola fitting, cannot achieve multi-resolution chip positioning to accurately position the chip in the machine coordinate system, cannot dynamically plan a collision-free path while optimizing the path length, time, and safety, cannot maintain the collaborative distance between devices through master-slave collaborative control, and there is no three-dimensional dynamic operation coding to generate a three-dimensional trajectory path with Z-axis height, and cannot form a dynamic closed-loop of visual positioning (perception) → path planning (decision-making) → dynamic control (execution) to achieve adaptive adjustment of flexible production.

[0005] To solve the above technical problem, the present invention provides the following technical solution: A chip placement collaborative control system for flexible production, including a sensing module, a decision-making module, and an execution module: The perception module is used to collect PCB and chip images through an industrial camera, and machine vision is used to locate the chip reference point position; The decision-making module is used for path planning and cooperative control; The execution module is used for multi-dimensional cooperative operations.

[0006] Preferably, the perception module includes a PCB positioning unit and a chip positioning unit: The PCB positioning unit includes positioning the cross-shaped reference points on the PCB board, establishing a global coordinate system. Specifically, establishing the global coordinate system includes: enhancing the contrast of the collected images during the chip soldering process by using piecewise linear transformation, suppressing image noise and retaining edge details by using Gaussian filtering, extracting the reference point contour through Canny edge detection, detecting the straight-line features of the cross-shaped reference points by improving the Radon transform, screening candidate straight lines through polar coordinate transformation, and fitting the center of the reference point by combining the least squares method; The chip positioning unit includes: quickly positioning the chip area in the case of low resolution, refining the angle matching in the case of high resolution, based on the point set registration algorithm, extracting the chip pin features through Harris corner detection, constructing the transformation matrix between the ideal model and the actual point set, and combining rotation, translation and scaling parameters to calculate the pose matrix of the chip relative to the machine coordinate system.

[0007] Preferably, the PCB positioning unit specifically includes: Dynamically dividing the gray level interval, linearly stretching the reference point area, compressing the background gray level interval, and optimizing the edge reference points through double thresholds. The optimization process includes: Setting the gray level threshold interval, the edge gray points higher than the gray level threshold interval are strong edges, and the edge gray points lower than the gray level threshold interval are weak edges. Only the weak edges connected to the strong edges are retained, all the reference point contours are extracted, the discrete edge points detected by Canny are parabola-fitted, the sub-pixel level edge position is calculated, and the sub-pixel level edge image is obtained; Project the sub-pixel level edge image onto the polar coordinate system, calculate the gray level integral projection at each angle in the polar coordinate system, detect the cross-shaped edge straight lines through Hough transform and delete them, perform gray level fitting on the remaining candidate straight lines after deleting the cross-edge straight lines, and use the calculated intersection point of the gray level fitting as the center of the reference point.

[0008] Preferably, the chip positioning unit specifically includes: Quickly screening candidate chip areas through a multi-resolution positioning strategy in the global image. When in the case of low resolution, downsample the image and use morphological operations to extract the approximate chip contour. The morphological operations include dilation and erosion; When in the high-resolution case, refine the candidate chip area of the approximate chip contour and perform local feature extraction. The local feature extraction includes: extracting chip pin feature points within the candidate area using Harris corner detection.

[0009] Preferably, extracting the chip pin feature points specifically includes: Calculate the gradients of each reference pixel point in the refined candidate chip area, and construct the gradient covariance matrix of each reference pixel point based on the gradient information. The mathematical expression of the covariance matrix is: ; Where, and are the gradients of the image in the x and y directions respectively, and x and y are the reference pixel points in the image; Calculate the Harris corner response value. The mathematical expression of the Harris corner response value is: ; Where, k is an empirical constant with a value in the range of [0.04, 0.06]; Screen the reference pixel points with corner response values higher than the preset response threshold, and retain the point with the maximum gradient. Perform eigenvalue decomposition of the covariance matrix based on basic matrix calculations to obtain the eigenvalues and , construct an ideal chip model, and output the chip pin feature points.

[0010] Preferably, constructing the ideal chip model includes: Perform ICP iterative closest point matching based on the chip pin feature points and the pre-extracted actual point set, calculate the rotation matrix, translation vector, and scaling factor, and generate the pose matrix; The mathematical expression of the pose matrix is: ; Where, N is the pose matrix, R is the rotation matrix, T is the translation vector, and S is the scaling factor; Convert the chip local coordinate system to the machine coordinate system to obtain the actual position and orientation of the chip, and output the reference point coordinates in the machine coordinate system, the rotation angle of the chip relative to the machine coordinate system, and the chip pose matrix. The reference point coordinates are the chip pin feature points.

[0011] Preferably, the decision module includes a planning unit and a control unit: The planning unit includes: Generate a multi-dimensional collaborative operation path based on the chip pin feature points, perform non-repetitive coding on the chip placement operation machinery and equipment and obtain the operation code, and construct the mechanical equipment link coordinate system based on the operation code; Generating a multi-dimensional collaborative operation path based on the chip pin feature points specifically includes: Randomly extract chip pin feature points as path nodes, generate a collision-free path through the RRT algorithm, and perform multi-objective optimization on the operation path of chip soldering. The optimization objectives include path length, time, and safety; Obtaining the operation code includes: Generate time-series path points according to the beat period of the mounter, superimpose the Z-axis height coordinate information of the chip pin feature point machine coordinate system on the path of the discrete time-series path points to generate a three-dimensional trajectory path, and allocate the corresponding coding prefix for each process stage according to the process stage. The process stages include loading, soldering, and detection, and output a three-dimensional dynamic operation code; Constructing the mechanical equipment link coordinate system includes: Construct an adaptive mechanical equipment coordinate system through forward kinematics and inverse kinematics, control the position and posture of the end effector of the mechanical equipment, and calibrate and note the position and posture of the end effector in the mechanical equipment coordinate system. Calculate the position and posture of the end effector of the mechanical equipment through the chain multiplication of the link transformation matrix, optimize the joint angles of the end effector of the mechanical equipment through the particle swarm algorithm, and perform precise control in combination with the Jacobian matrix constraint.

[0012] Preferably, the control unit is used to perform master-slave collaborative formation control on the multi-dimensional collaborative operation path: Master-slave collaborative formation control includes master device global path planning and slave device tracking feedback; Master device global path planning includes: inputting the chip soldering task sequence and the global map, using the improved APF-RRT algorithm for path planning, constructing the gravitational field of the target point and the repulsive field of the obstacle, and generating a collision-free and smooth optimized path; Using the improved APF-RRT algorithm for path planning includes: S01: Initialize the optimization environment, set the global path parameters. The global path parameters include the joint point coordinates, angles of the initial end effector of the mechanical equipment, and obstacle information. The obstacle information includes the coordinate set of the mechanical equipment link coordinate system that the end effector of the mechanical equipment is prohibited from reaching, and set the optimization target as the expected joint point coordinates; S02: Perform initialization global path planning according to the multi-dimensional collaborative operation path; S03: Calculate the resultant force received by the current end effector of the mechanical equipment. The resultant force is the resultant force of all forces obtained by performing a force analysis according to Newton's laws of motion. Set the resultant force threshold F_min. When the resultant force is greater than the resultant force threshold F_min, it indicates that the current path planning is effective. Continue to detect the resultant force in real time and execute step S05; When the resultant force is less than the resultant force threshold F_min, it indicates that the current path planning is invalid, and step S04 is executed; S04: Conduct a limited number of searches through an improved RRT algorithm, select the node on the RRT growth tree that is closest to the current critical angle as the temporary target joint angle, and continue to execute step S02; S05: Determine whether the joint angle of the end effector of the current mechanical equipment has reached the expected coordinates of the target joint point: When the joint angle of the end effector of the current mechanical equipment has reached the expected coordinates of the target joint point, the path planning optimization task is completed and the loop is exited; When the joint angle of the end effector of the current mechanical equipment has not reached the expected coordinates of the target joint point, return to step S02.

[0013] Device tracking feedback includes: Using a feedback linearization control architecture, error compensation is performed based on the Jacobian matrix, a six-dimensional state equation is established through an LQR controller and the optimal control law is solved to achieve precise trajectory tracking, a virtual spring-damping model is used to maintain the desired distance between the master and slave devices, and the device states are synchronized in the communication network through a distributed consensus algorithm.

[0014] Preferably, the execution module includes: Based on the PID control theory, a three-loop controller for position, speed, and acceleration is designed, an error model is constructed according to the multi-dimensional collaborative operation path output by the decision module, the acceleration and deceleration process of the motion trajectory is optimized by an S-curve, the position deviation of the chip is continuously monitored through a vision system, the attitude of the placement head is adjusted using fuzzy adaptive control, and the drift of the vision system caused by environmental changes is compensated through an online calibration algorithm.

[0015] Constructing the error model includes: Define the errors of position, speed, and acceleration, and based on the link coordinate system of the mechanical equipment output by the decision module, obtain the real-time actual pose of the chip accordingly, and compare the coordinates with the preset target pose to obtain the error vector.

[0016] Preferably, the S-curve acceleration and deceleration optimization includes: The acceleration profile is generated by a fifth-order polynomial interpolation method to generate an acceleration curve, constraining the continuity of the initial / terminal velocity, acceleration, and chip displacement.

[0017] Update the PCB reference point and the chip pin feature point according to the preset period, calculate and obtain the offset of the link coordinate system of the mechanical equipment, use the RANSAC algorithm to fit the calibration error model, calculate the influence of the device running time and temperature on the vision system in real time, feedback the calibration error as an error function to the multi-dimensional collaborative operation path, and correct the pose of the end effector through master-slave collaborative formation control.

[0018] Advantages of the present invention: By adopting improved Radon transform and polar coordinate screening, combined with parabolic fitting to achieve sub-pixel level center positioning of fiducial points, the accuracy is increased by 10 times compared with traditional methods. Through double-threshold segmentation and background compression, noise interference is reduced and the robustness of feature extraction is enhanced. Multi-resolution chip positioning realizes accurate positioning of the chip in the machine coordinate system. The improved APF-RRT algorithm combines the gravitational-repulsive force field of the artificial potential field with the random search of RRT to dynamically plan a collision-free path, optimizing the path length, time and safety. The master-slave cooperative control maintains the cooperative distance between devices. The three-dimensional dynamic operation coding generates a three-dimensional trajectory path with Z-axis height to adapt to the flexible production requirements. The quintic polynomial interpolation is used to generate a smooth acceleration curve to reduce mechanical vibration and improve the patch stability. The calibration error is fitted by the RANSAC algorithm to compensate for the influence of environmental factors such as temperature and mechanical drift in real time. Visual positioning (perception) → path planning (decision-making) → dynamic control (execution) forms a closed loop to achieve adaptive adjustment of flexible production. Combining technologies in multiple fields such as machine vision, robotics, optimization algorithms, and control theory, it breaks through the limitations of single methods, meets the requirements of micron-level chip patching, supports multi-variety and small-batch production, adapts to the trend of intelligent manufacturing, and the dynamic error compensation mechanism can cope with environmental interference and improve the yield rate. Through the three major innovations of visual enhanced positioning, intelligent cooperative control, and dynamic error compensation, the system realizes a closed-loop management of chip patching production with higher accuracy, faster response, and stronger adaptability than traditional methods. Brief Description of the Drawings

[0019] Figure 1 FIG. is a schematic diagram of the basic process of a chip patching cooperative control system for flexible production provided by an embodiment of the present invention. Detailed Embodiments

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed embodiments of the present invention in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them.

[0021] Refer to Figure 1 , which is an embodiment of the present invention, providing a chip patching cooperative control system for flexible production, including a sensing module, a decision-making module, and an execution module: The sensing module is used to collect images of the PCB and the chip through an industrial camera, and machine vision is used to locate the position of the chip fiducial point; The decision-making module is used for path planning and cooperative control; The execution module is used for multi-dimensional cooperative operations.

[0022] The sensing module includes a PCB positioning unit and a chip positioning unit: The PCB positioning unit includes locating the cross-shaped fiducial points on the PCB board and establishing a global coordinate system. Specifically, establishing the global coordinate system includes: enhancing the contrast of the acquired image during the chip placement process using piecewise linear transformation, suppressing image noise and preserving edge details using Gaussian filtering, extracting the fiducial point contour through Canny edge detection, detecting the linear features of the cross-shaped fiducial points using improved Radon transform, screening candidate lines through polar coordinate transformation, and fitting the center of the fiducial point using the least squares method. The chip positioning unit includes: quickly locating the chip area at low resolution, refining the angle matching at high resolution, based on the point set registration algorithm, extracting the chip pin features through Harris corner detection, constructing the transformation matrix between the ideal model and the actual point set, and calculating the pose matrix of the chip relative to the machine coordinate system by combining rotation, translation, and scaling parameters.

[0023] Specifically, the PCB positioning unit includes: Dynamically dividing the gray level interval, performing linear stretching on the fiducial point area, for example, mapping the gray value to [0, 255], compressing the background gray level interval, for example, setting the fiducial point gray range to [80, 150], and setting the background to [0, 79] and [151, 255], and optimizing the edge fiducial points through double thresholds. The optimization process includes: Setting the gray threshold interval, the edge gray points above the gray threshold interval are strong edges, and the edge gray points below the gray threshold interval are weak edges. Only retain the weak edges connected to the strong edges to avoid interference from isolated noise points, extract all fiducial point contours, perform parabolic fitting on the discrete edge points detected by Canny, calculate the sub-pixel level edge position, obtain the sub-pixel level edge image, and reduce the fiducial point center positioning error from ±1 pixel to ±0.1 pixel. Project the sub-pixel level edge image into the polar coordinate system, calculate the gray integral projection at each angle in the polar coordinate system, detect and delete the cross-shaped edge lines through Hough transform, perform gray fitting on the remaining candidate lines after deleting the cross-shaped edge lines, and use the calculated intersection point of the gray fitting as the center of the fiducial point, with stronger anti-interference ability.

[0024] Specifically, the chip positioning unit includes: Quickly screening candidate chip areas through a multi-resolution positioning strategy in the global image. When at low resolution, downsample the image and use morphological operations to extract the approximate chip contour. The morphological operations include dilation and erosion. When at high resolution, refine the candidate chip area for the approximate chip contour and perform local feature extraction. The local feature extraction includes: extracting chip pin feature points using Harris corner detection within the candidate area.

[0025] Specifically, extracting the chip pin feature points includes: Calculate the gradients of each reference pixel point in the refined candidate chip region to describe the local changes in the image. Based on the gradient information, construct the gradient covariance matrix for each reference pixel point. The covariance matrix captures the structural features of the local image, and the mathematical expression of the covariance matrix is: ; Among them, and are the gradients in the x and y directions of the image respectively, and x and y are the reference pixel points in the image; Calculate the Harris corner response value. The mathematical expression of the Harris corner response value is: ; Among them, k is an empirical constant, and its value range is [0.04, 0.06]; Screen the reference pixel points with corner response values higher than the preset response threshold, and retain the points with the maximum gradient value. Perform eigenvalue decomposition on the covariance matrix based on basic matrix calculations to obtain the eigenvalues and . The eigenvalues indicate the change direction and degree of the image at this point. Construct an ideal chip model and output the chip pin feature points.

[0026] Constructing the ideal chip model includes: Based on the chip pin feature points and the pre-extracted actual point set, perform ICP iterative closest point matching, calculate the rotation matrix, translation vector, and scaling factor, and generate the pose matrix; The mathematical expression of the pose matrix is: ; Among them, N is the pose matrix, R is the rotation matrix, T is the translation vector, and S is the scaling factor; Convert the chip local coordinate system to the machine coordinate system to obtain the actual position and orientation of the chip, and output the reference point coordinates in the machine coordinate system, the rotation angle of the chip relative to the machine coordinate system, and the chip pose matrix. The reference point coordinates are the chip pin feature points.

[0027] The decision-making module includes a planning unit and a control unit: The planning unit includes: Generate a multi-dimensional collaborative operation path according to the chip pin feature points, perform non-repetitive coding on the chip placement operation machine equipment and obtain the operation code, and construct the mechanical equipment link coordinate system based on the operation code; Generating a multi-dimensional collaborative operation path according to the chip pin feature points specifically includes: Randomly extract the feature points of the chip pins as path nodes, generate a collision-free path through the RRT algorithm, and perform multi-objective optimization on the operation path of chip placement considering the chip size constraint. The optimization objectives include path length, time, and safety; Obtaining the operation code includes: According to the beat cycle of the mounter, generate time-series path points, superimpose the Z-axis height coordinate information of the chip pin feature points in the machine coordinate system on the path of the discrete time-series path points to generate a three-dimensional trajectory path, and allocate the corresponding coding prefix for each process stage according to the process stage. The process stages include loading, placement, and detection, and output a three-dimensional dynamic operation code; Constructing the link coordinate system of the mechanical equipment includes: Construct an adaptive mechanical equipment coordinate system through forward kinematics and inverse kinematics, control the position and posture of the end effector of the mechanical equipment, and calibrate and note the position and posture of the end effector in the mechanical equipment coordinate system. Calculate the position and posture of the end effector of the mechanical equipment through the chained multiplication of the link transformation matrix, optimize the joint angles of the end effector of the mechanical equipment through the particle swarm algorithm, and perform precise control in combination with the Jacobian matrix constraint.

[0028] The link transformation matrix expresses the position and posture of each link relative to the previous link through matrix multiplication. These transformation matrices are gradually accumulated, and the position and posture of the end effector are obtained through chained multiplication.

[0029] The Jacobian matrix is used to describe the relationship between the motion of the end effector in the joint space and the workspace. Through the Jacobian matrix, the relationship between the speed and acceleration of the end effector and the joint angular velocity can be analyzed, and then precise control can be performed. When performing inverse kinematics calculations, by combining the Jacobian matrix constraint, it can be ensured that the solution of the joint angles conforms to physical limitations, such as the motion range and speed limit of the joints, so as to achieve precise control.

[0030] The control unit is used to perform master-slave cooperative formation control on the multi-dimensional collaborative operation path: The master-slave cooperative formation control includes global path planning of the master device and tracking feedback of the slave device; The global path planning of the master device includes: inputting the chip placement task sequence and the global map, using the improved APF-RRT algorithm for path planning, constructing the gravitational field of the target point and the repulsive field of the obstacle, and generating a collision-free and smooth optimized path; Using the improved APF-RRT algorithm for path planning includes: S01: Initialize the optimization environment, set the global path parameters. The global path parameters include the joint point coordinates, angles of the initial end - effector of the mechanical equipment, and obstacle information. The obstacle information includes the coordinate set of the mechanical equipment link coordinate system where the end - effector of the mechanical equipment is prohibited from reaching. Set the optimization goal as the expected joint point coordinates; S02: Perform global path planning initialization according to the multi - dimensional collaborative operation path; S03: Calculate the resultant force received by the current end - effector of the mechanical equipment. The resultant force is the resultant force of all forces obtained by force analysis according to Newton's laws of motion. Set the resultant force threshold F_min. When the resultant force is greater than the resultant force threshold F_min, it indicates that the current path planning is effective. Continue to detect the resultant force in real - time and execute step S05; When the resultant force is less than the resultant force threshold F_min, it indicates that the current path planning is invalid, and execute step S04; S04: Conduct a limited number of searches through the improved RRT algorithm. These searches attempt to search through different paths and select the node closest to the current joint angle as the temporary target joint angle, thereby providing a new path planning direction. Select the node on the RRT growth tree closest to the current critical angle as the temporary target joint angle, and continue to execute step S02; The improved RRT algorithm includes improving the node selection strategy and introducing more efficient sampling methods or pruning strategies to improve the path planning efficiency. The improved RRT algorithm is a prior art.

[0031] S05: Determine whether the joint angle of the current end - effector of the mechanical equipment has reached the expected coordinates of the target joint point: When the joint angle of the current end - effector of the mechanical equipment has reached the expected coordinates of the target joint point, the path planning optimization task is completed and the loop exits; When the joint angle of the current end - effector of the mechanical equipment has not reached the expected coordinates of the target joint point, return to step S02.

[0032] The slave device tracking feedback includes: Using the feedback linearization control architecture, perform error compensation based on the Jacobian matrix, establish a six - dimensional state equation through the LQR controller and solve the optimal control law to achieve precise trajectory tracking. Use the virtual spring - damping model to maintain the expected distance between the master and slave devices, and synchronize the device states in the communication network through the distributed consensus algorithm to ensure the collaborative work of the master and slave devices.

[0033] The execution module includes: Based on the PID control theory, a three-loop controller for position, speed, and acceleration is designed. An error model is constructed according to the multi-dimensional collaborative operation path output by the decision-making module. The S-curve is used to optimize the acceleration and deceleration process of the motion trajectory, reducing mechanical vibrations in the chip placement environment. The position deviation of the chip is continuously monitored through the vision system, and the attitude of the placement head is adjusted using fuzzy adaptive control. The drift of the vision system caused by environmental changes is compensated through an online calibration algorithm.

[0034] The construction of the error model includes: Define the errors of position, speed, and acceleration. Based on the link coordinate system of the mechanical equipment output by the decision-making module, obtain the real-time actual pose of the chip accordingly, and compare the coordinates with the preset target pose to obtain the error vector.

[0035] The S-curve acceleration and deceleration optimization includes: The acceleration profile is generated by a fifth-order polynomial interpolation method, constraining the continuity of the initial / terminal speed, acceleration, and chip displacement.

[0036] Update the PCB reference point and the chip pin feature point according to the preset period, calculate the offset of the link coordinate system of the mechanical equipment, fit the calibration error model using the RANSAC algorithm, calculate the influence of the equipment running time and temperature on the vision system in real time, feedback the calibration error as an error function to the multi-dimensional collaborative operation path, and correct the pose of the end effector through master-slave collaborative formation control to maintain an accurate operation path.

[0037] The present invention uses improved Radon transform and polar coordinate screening, combined with parabolic fitting to achieve sub-pixel level benchmark point center positioning, with the accuracy improved by 10 times compared with traditional methods. Through double-threshold segmentation and background compression, noise interference is reduced and the robustness of feature extraction is enhanced. Multi-resolution chip positioning realizes the precise positioning of the chip in the machine coordinate system. The improved APF-RRT algorithm combines the gravitational-repulsive force field of the artificial potential field and the random search of RRT to dynamically plan a collision-free path, optimizing the path length, time, and safety. The master-slave cooperative control maintains the cooperative spacing between devices. The three-dimensional dynamic operation coding generates a three-dimensional trajectory path with the Z-axis height to adapt to the flexible production requirements. The quintic polynomial interpolation is used to generate a smooth acceleration curve to reduce mechanical vibration and improve the placement stability. The calibration error is fitted by the RANSAC algorithm to compensate in real time for the influence of environmental factors such as temperature and mechanical drift. Vision positioning (perception) → path planning (decision-making) → dynamic control (execution) forms a closed loop to achieve the adaptive adjustment of flexible production. Combining technologies in multiple fields such as machine vision, robotics, optimization algorithms, and control theory, it breaks through the limitations of single methods, meets the requirements of micron-level chip placement, supports multi-variety and small-batch production, adapts to the trend of intelligent manufacturing, and the dynamic error compensation mechanism copes with environmental interference to improve the yield. Through the three major innovations of visual enhanced positioning, intelligent cooperative control, and dynamic error compensation, the system realizes a closed-loop management of chip placement production with higher accuracy, faster response, and stronger adaptability than traditional methods.

[0038] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk. These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the functions specified in one process Figure 1 one process or multiple processes and / or boxes Figure 1 specified in one box or multiple boxes.

[0039] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. Chip placement collaborative control system for flexible production, characterized in that It includes a sensing module, a decision-making module, and an execution module: The sensing module is used to collect PCB and chip images through an industrial camera and locate the reference point position of the chip by machine vision; The decision-making module is used for path planning and collaborative control; The execution module is used for multi-dimensional collaborative operations.

2. The chip placement collaborative control system for flexible production according to claim 1, wherein The sensing module includes a PCB positioning unit and a chip positioning unit: The PCB positioning unit includes locating the cross-shaped reference point on the PCB board, establishing a global coordinate system. The specific process of establishing the global coordinate system includes: enhancing the contrast of the collected images during the chip soldering process using piecewise linear transformation, suppressing image noise and retaining edge details using Gaussian filtering, extracting the reference point contour through Canny edge detection, detecting the linear features of the cross-shaped reference point using improved Radon transform, screening candidate lines through polar coordinate transformation, and fitting the center of the reference point using the least squares method; The chip positioning unit includes: quickly locating the chip area in the case of low resolution, refining the angle matching in the case of high resolution. Based on the point set registration algorithm, extracting the chip pin features through Harris corner detection, constructing the transformation matrix between the ideal model and the actual point set, and calculating the pose matrix of the chip relative to the machine coordinate system by combining rotation, translation, and scaling parameters.

3. The chip mounter collaborative control system for flexible production according to claim 2, characterized in that, The PCB positioning unit specifically includes: Dynamically dividing the gray level interval, linearly stretching the reference point area, compressing the background gray level interval, and optimizing the edge reference points through double thresholds. The optimization process includes: Setting the gray level threshold interval, the edge gray points above the gray level threshold interval are strong edges, and the edge gray points below the gray level threshold interval are weak edges. Only retain the weak edges connected to the strong edges, extract all the reference point contours, perform parabolic fitting on the discrete edge points detected by Canny, calculate the sub-pixel level edge position, and obtain the sub-pixel level edge image; Project the sub-pixel level edge image onto the polar coordinate system, calculate the gray level integral projection at each angle in the polar coordinate system, detect the cross-shaped edge lines through Hough transform and delete them, perform gray level fitting on the remaining candidate lines after deleting the cross-edge lines, and use the calculated intersection point of the gray level fitting as the center of the reference point.

4. The chip mounter collaborative control system for flexible production according to claim 3, wherein The chip positioning unit specifically includes: Quickly screening candidate chip areas through a multi-resolution positioning strategy in the global image. When in the case of low resolution, downsample the image and use morphological operations to extract the approximate chip contour. The morphological operations include dilation and erosion; When in the case of high resolution, refine the candidate chip area of the approximate chip contour and perform local feature extraction. The local feature extraction includes: extracting the chip pin feature points using Harris corner detection within the candidate area.

5. The chip mounter collaborative control system for flexible production according to claim 4, wherein The specific process of extracting the chip pin feature points includes: Calculating the gradient of each reference pixel point in the refined candidate chip area, constructing the gradient covariance matrix of each reference pixel point based on the gradient information. The mathematical expression of the covariance matrix is: ； wherein, and are the gradients of the image in the x and y directions respectively, and x and y are the reference pixel points in the image; Calculating the Harris corner response value. The mathematical expression of the Harris corner response value is: ； where k is an empirical constant, and its value range is [0.04, 0.06]; Screen the reference pixel points whose corner response values are higher than the preset response threshold, and retain the points with the maximum gradient value. Perform eigenvalue decomposition on the covariance matrix based on basic matrix calculations to obtain the eigenvalues and , construct an ideal chip model, and output the characteristic points of the chip pins.

6. The chip mounter collaborative control system for flexible production according to claim 5, characterized in that Constructing the ideal chip model includes: Based on the chip pin feature points and the pre-extracted actual point set, perform ICP iterative closest point matching, calculate the rotation matrix, translation vector, and scaling factor, and generate a pose matrix; The mathematical expression of the pose matrix is: ； where N is the pose matrix, R is the rotation matrix, T is the translation vector, and S is the scaling factor; Convert the chip local coordinate system to the machine coordinate system to obtain the actual position and orientation of the chip, and output the reference point coordinates in the machine coordinate system, the rotation angle of the chip relative to the machine coordinate system, and the chip pose matrix. The reference point coordinates are the chip pin feature points.

7. The chip mounter collaborative control system for flexible production according to claim 6, characterized in that, The decision-making module includes a planning unit and a control unit: The planning unit includes: Generate a multi-dimensional collaborative operation path based on the chip pin feature points, encode the chip placement operation machine equipment without repetition and obtain an operation code, and construct a mechanical equipment link coordinate system based on the operation code; Generating a multi-dimensional collaborative operation path based on the chip pin feature points specifically includes: Randomly extract chip pin feature points as path nodes, generate a collision-free path through the RRT algorithm, and perform multi-objective optimization on the operation path of chip placement. The optimization objectives include path length, time, and safety; Obtaining the operation code includes: Generate time series path points according to the beat period of the mounter, superimpose the Z-axis height coordinate information of the chip pin feature point machine coordinate system on the path of the discrete time series path points to generate a three-dimensional trajectory path, and allocate the corresponding coding prefix for each process stage according to the process stage. The process stages include loading, placement, and detection, and output a three-dimensional dynamic operation code; Constructing a mechanical equipment link coordinate system includes: Construct an adaptive mechanical equipment coordinate system through forward kinematics and inverse kinematics, control the position and posture of the end effector of the mechanical equipment, and calibrate and note the position and posture of the end effector in the mechanical equipment coordinate system. Calculate the position and posture of the end effector of the mechanical equipment through the chain multiplication of the link transformation matrix, optimize the joint angles of the end effector of the mechanical equipment through the particle swarm algorithm, and perform precise control in combination with the Jacobian matrix constraint.

8. The chip mounter collaborative control system for flexible production according to claim 7, characterized in that The control unit is used to perform master-slave collaborative formation control on the multi-dimensional collaborative operation path: Master-slave collaborative formation control includes master device global path planning and slave device tracking feedback; Master device global path planning includes: input the chip placement task sequence and the global map, use the improved APF-RRT algorithm for path planning, construct the gravitational field of the target point and the repulsive field of the obstacle, and generate a collision-free and smooth optimized path; Using the improved APF-RRT algorithm for path planning includes: S01: Initialize the optimization environment, set the global path parameters. The global path parameters include the joint point coordinates, angles of the initial end effector of the mechanical equipment, and obstacle information. The obstacle information includes the coordinate set of the mechanical equipment link coordinate system that the end effector of the mechanical equipment is prohibited from reaching, and set the optimization target as the expected joint point coordinates; S02: Perform initialization global path planning according to the multi-dimensional collaborative operation path; S03: Calculate the resultant force received by the end effector of the current mechanical equipment. The resultant force is the resultant force of all forces obtained by performing a force analysis according to Newton's laws of motion. Set a resultant force threshold F_min. When the resultant force is greater than the resultant force threshold F_min, it indicates that the current path planning is effective. Continue to detect the resultant force in real time and execute step S05; When the resultant force is less than the resultant force threshold F_min, it indicates that the current path planning is invalid, and execute step S04; S04: Conduct a finite number of searches through an improved RRT algorithm, and select the node on the RRT growth tree that is closest to the current key angle as the temporary target joint angle, and continue to execute step S02; S05: Determine whether the joint angle of the end effector of the current mechanical equipment has reached the expected coordinates of the target joint point: When the joint angle of the end effector of the current mechanical equipment has reached the expected coordinates of the target joint point, the path planning optimization task is completed and the loop is exited; When the joint angle of the end effector of the current mechanical equipment has not reached the expected coordinates of the target joint point, return to step S02; Device tracking feedback includes: Using a feedback linearization control architecture, performing error compensation based on the Jacobian matrix, establishing a six-dimensional state equation through an LQR controller and solving the optimal control law to achieve precise trajectory tracking, using a virtual spring-damper model to maintain the desired distance between the master and slave devices, and synchronizing the device states in the communication network through a distributed consensus algorithm.

9. The chip mounter collaborative control system for flexible production according to claim 8, wherein The execution module includes: Based on the PID control theory, design a three-loop controller for position, speed, and acceleration. Construct an error model according to the multi-dimensional collaborative operation path output by the decision-making module, optimize the acceleration and deceleration process of the motion trajectory with an S-curve, continuously monitor the chip position deviation through a vision system, adjust the placement head attitude using fuzzy adaptive control, and compensate for the drift of the vision system caused by environmental changes through an online calibration algorithm; Constructing an error model includes: Define the errors of position, speed, and acceleration. Based on the mechanical equipment link coordinate system output by the decision-making module, obtain the real-time actual pose of the chip accordingly, and compare the coordinates with the preset target pose to obtain an error vector.

10. The chip mounter collaborative control system for flexible production according to claim 9, wherein S-curve acceleration and deceleration optimization includes: The acceleration profile is generated by a fifth-order polynomial interpolation method, constraining the continuity of the initial / terminal velocity, acceleration, and chip displacement; Update the PCB reference point and the chip pin feature point according to the preset period, calculate the offset of the mechanical equipment link coordinate system, fit the calibration error model using the RANSAC algorithm, calculate the influence of the device running time and temperature on the vision system in real time, feedback the calibration error as an error function to the multi-dimensional collaborative operation path, and correct the pose of the end effector through master-slave collaborative formation control.

Images