A printing system with self-compensation function and a self-compensation method

Abstract

This invention discloses a printing system and method with self-compensation function. The system includes: a motion platform for carrying a printing plate and controlling its movement along a set trajectory; a movable vision acquisition unit positioned between the stencil and the printing plate for acquiring marker points on the printing plate and the stencil; an external detection unit for detecting deviations between the stencil and the printing plate under pressure; and a control processing unit that constructs a comprehensive error model containing multiple geometric error parameters based on the marker points acquired by the movable vision acquisition unit, and generates a compensation mapping table by combining the deviations detected by the external detection unit. During actual printing, the control processing unit, based on the current dimensions of the printing plate and the stencil, calls the corresponding comprehensive error model parameters and the compensation mapping table to correct the set trajectory of the motion platform in real time. This invention achieves precise compensation for printing deviations.

Description

Technical Field

[0001] This invention relates to precision printing equipment technology, specifically to a printing system and self-compensation method with self-compensation function. Background Technology

[0002] With the development of industries such as electronics manufacturing and precision printing, the requirements for the alignment accuracy of printing presses and other equipment are increasing. Currently, most mainstream printing press calibration methods rely on static compensation based on actual printing results. While these methods can obtain good compensation parameters at the time of calibration, in actual dynamic operation, the presence of various time-varying or motion-related error sources leads to a decrease in accuracy after calibration. Specifically: 1. Deviation between the machine tool rotation center and the theoretical design center: This deviation introduces unexpected translational components when the platform performs pure rotational motion, resulting in misalignment.

[0003] 2. Verticality deviation of the Z-axis of the lifting platform: When the Z-axis is raised to press the steel mesh and the printing plate together, this deviation will cause relative slippage of the contact surfaces of the two, resulting in planar offset.

[0004] 3. The optical axes of the built-in multi-cameras are not parallel: There is an installation deviation between the camera that acquires the stencil and the camera that acquires the printed circuit board. This deviation will become apparent before and after the Z-axis movement, causing "static alignment and pressing misalignment".

[0005] 4. Non-uniformity of platform motion and non-orthogonality between axes: When the motion platform is in different ranges of a single axis or when performing multi-axis linkage, its positioning accuracy is not ideally linear, introducing random or systematic positioning errors.

[0006] Existing technologies primarily address static calibration issues, lacking systematic modeling and compensation schemes for the aforementioned dynamic errors. Therefore, there is an urgent need for a method capable of online, dynamic calibration and compensation for these combined errors to achieve high-precision and stable alignment of the printing press throughout its entire workspace. Summary of the Invention

[0007] The purpose of this invention is to provide a printing system and a self-compensation method with self-compensation function to achieve offset compensation of the printing press.

[0008] To solve the above-mentioned technical problems, the present invention provides a printing system with self-compensation function, including a stencil and a printing plate disposed below the stencil, and further comprising: A motion platform is used to support the printing plate and control its movement along a set trajectory. A movable vision acquisition unit is set between the stencil and the printing plate to acquire the marking points on the printing plate and the stencil. There are several sets of marking points, with each set of marking points positioned opposite each other on the upper surface of the printing plate and the lower surface of the stencil. An external detection unit is used to detect the deviation between the steel mesh and the printing plate under the pressing state; The control and processing unit is communicatively connected to the motion platform, the movable vision acquisition unit, and the external detection unit. Based on the marker points acquired by the movable vision acquisition unit, the control and processing unit constructs a comprehensive error model containing multiple geometric error parameters, and generates a compensation mapping table by combining the deviations detected by the external detection unit. During actual printing, the control and processing unit calls the corresponding comprehensive error model parameters and compensation mapping table according to the current printing plate and stencil size to correct the set trajectory of the motion platform in real time.

[0009] According to the above scheme, the motion platform is the XYY platform, which controls the printed circuit board to perform horizontal translation or horizontal rotation.

[0010] According to the above scheme, the movable vision acquisition unit includes a two-dimensional slide rail and a pair of monocular cameras slidably connected to the two-dimensional slide rail; the two-dimensional slide rail controls the pair of monocular cameras to perform horizontal translation, and the two monocular cameras respectively acquire the marked points on the printed circuit board and the steel mesh.

[0011] According to the above scheme, the methods for constructing a comprehensive error model include: The initial pose of the printing plate is calculated by acquiring at least two marker points of the printing plate in its initial state using a movable vision acquisition unit. The printing plate is moved to multiple points along a certain trajectory by a motion platform. At each point, at least two marker points of the printing plate are collected by a movable vision acquisition unit, and the actual pose of the printing plate at each point is calculated. Based on the initial pose of the printed circuit board and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

[0012] According to the above scheme, the methods for generating the compensation mapping table include: The printing plate is controlled by a motion platform to align the sets of markers acquired by the movable vision acquisition unit within the field of view. The printing plate is pressed into the stencil using a motion platform; The deviation between the steel mesh and the printing plate under the pressing state is detected by an external detection unit, and the printing plate is moved by a motion platform to eliminate the deviation. The printing plate is controlled to leave the stencil via a motion platform; Each set of marker points is collected again using the movable vision acquisition unit, and the residual pose deviation is calculated. The position of the stencil was changed multiple times, and the residual pose deviation under different relative poses was calculated to construct a compensation mapping table; the relative pose represents the pose deviation between the stencil and the printed circuit board.

[0013] According to the above scheme, the control processing unit constructs a calibration database by calculating the comprehensive error model and compensation mapping table under different printed circuit board and stencil sizes.

[0014] This invention also provides a self-compensation method for a printing system. The printing system includes a stencil, a printing plate, a motion platform, a movable vision acquisition unit, an external detection unit, and a control processing unit. The movable vision acquisition unit is disposed between the stencil and the printing plate. The lower surface of the stencil and the upper surface of the printing plate are provided with corresponding sets of marking points. The method includes the following steps: S1. Collect the marking points on the printed circuit board and steel mesh through the movable vision acquisition unit, and construct a comprehensive error model containing multiple geometric error parameters; S2. Generate a compensation mapping table based on the marker points collected by the movable vision acquisition unit and the deviation detected by the external detection unit. S3. During actual printing, based on the current printing plate and stencil size, the corresponding comprehensive error model parameters and compensation mapping table are called to correct the set trajectory of the motion platform in real time.

[0015] According to the above scheme, the methods for constructing a comprehensive error model include: The initial pose of the printing plate is calculated by acquiring at least two marker points of the printing plate in its initial state using a movable vision acquisition unit. The printing plate is moved to multiple points along a certain trajectory by a motion platform. At each point, at least two marker points of the printing plate are collected by a movable vision acquisition unit, and the actual pose of the printing plate at each point is calculated. Based on the initial pose of the printed circuit board and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

[0016] According to the above scheme, the methods for generating the compensation mapping table include: The printing plate is controlled by a motion platform to align the sets of markers acquired by the movable vision acquisition unit within the field of view. The printing plate is pressed into the stencil using a motion platform; The deviation between the steel mesh and the printing plate under the pressing state is detected by an external detection unit, and the printing plate is moved by a motion platform to eliminate the deviation. The printing plate is controlled to leave the stencil via a motion platform; Each set of marker points is collected again using the movable vision acquisition unit, and the residual pose deviation is calculated. The position of the stencil was changed multiple times, and the residual pose deviation under different relative poses was calculated to construct a compensation mapping table; the relative pose represents the pose deviation between the stencil and the printed circuit board.

[0017] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the self-compensation method of the printing system described above.

[0018] Beneficial effects This invention features a movable visual acquisition unit that can accurately acquire relative marker points on the stencil and the printing plate, overcoming the limitations of fixed acquisition methods. It comprehensively captures the pose correlation information of the two at different positions, providing rich and accurate raw data for error modeling. The control and processing unit constructs a comprehensive error model based on these marker points, incorporating multiple geometric error parameters. Compared to existing compensation schemes that only address single errors, this model systematically covers various geometric errors that may occur during the printing process, achieving a comprehensive and accurate description of the errors. Simultaneously, it generates a compensation mapping table by combining the deviation detected by the external detection unit under the pressed state of the stencil and the printing plate. This ensures that the compensation basis includes both pose information during movement and the actual deviation after pressing, guaranteeing the targeted and comprehensive nature of the compensation. During actual printing, the control processing unit flexibly calls upon the corresponding comprehensive error model parameters and compensation mapping table based on the specific dimensions of the current printing plate and stencil to correct the set trajectory of the motion platform in real time. This effectively solves the problem of accuracy decay in dynamic operation caused by existing static calibration methods, avoids misalignment caused by dynamic errors such as machine rotation center deviation and inter-axis non-orthogonality, significantly improves the alignment accuracy during the printing process, ensures that the printing press maintains a stable high-precision printing effect throughout the entire working space, and better meets the stringent requirements for equipment alignment accuracy in industries such as electronics manufacturing and precision printing. Attached Figure Description

[0019] Figure 1 This is a flowchart of a dynamic calibration method for a printing system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a printing system with self-compensation function according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the motion platform structure according to an embodiment of the present invention.

[0020] In the diagram: 1. Printed plate; 2. Stencil; 3. Movable vision acquisition unit; 4. Marker point. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0022] See Figure 1 This embodiment provides a dynamic calibration method for a printing press based on a movable vision system. The core concept is to use a vision system that can move above the moving plane of the printing plate 1 to simulate the perspective of multiple built-in cameras, dynamically acquiring data throughout the platform's entire travel or key areas. Combined with an established platform motion comprehensive error model, multiple geometric error parameters, including rotation center deviation, are calculated in one go. Furthermore, by constructing a deviation mapping table under different working conditions (such as different Z-axis heights and different printing plate 1 sizes), real-time compensation for errors such as optical axis deviation and perpendicularity deviation is achieved. Details are as follows.

[0023] See Figure 2 This embodiment discloses a printing system with self-compensation function, including a stencil 2 (the stencil 2 is fixed by clamping) and a printing plate 1 disposed below the stencil 2, and further including: A motion platform is used to carry the printing plate 1 and control the printing plate 1 to move along a set trajectory (the set trajectory of the motion platform can be a path generated based on space filling or a pose generated based on the end effector). A movable visual acquisition unit 3 is set between the stencil 2 and the printing plate 1 to acquire the marker points 4 on the printing plate 1 and the stencil 2. There are several sets of marker points 4, and each set of marker points 4 is respectively set on the upper surface of the printing plate 1 and the lower surface of the stencil 2. An external detection unit is used to detect the deviation between the steel mesh 2 and the printing plate 1 under the pressing state (magnifying the texture and bonding details between the steel mesh 2 and the printing plate 1 for manual observation). The control and processing unit is communicatively connected to the motion platform, the movable vision acquisition unit 3, and the external detection unit. Based on the marker points 4 acquired by the movable vision acquisition unit 3, the control and processing unit constructs a comprehensive error model containing multiple geometric error parameters, and generates a compensation mapping table by combining the deviation detected by the external detection unit. During actual printing, the control and processing unit calls the corresponding comprehensive error model parameters and compensation mapping table according to the current size of the printing plate 1 and the stencil 2, and corrects the set trajectory of the motion platform in real time.

[0024] Furthermore, the motion platform is an XYY platform, which controls the horizontal translation or rotation of the printing plate 1. Specifically, to achieve the pressing between the printing plate 1 and the stencil 2, the motion platform also has a vertically upward degree of freedom (i.e., the Z-axis), which can control the vertical translation of the printing plate 1.

[0025] In other embodiments of the present invention, the motion platform may be other types of platforms, and therefore the observable quantities of the motion platform are not necessarily X, Y, θ. It should be understood that as long as the motion platform can be mathematically modeled according to the structural configuration, the original structural model parameters can be modified in reverse (to correct assembly errors) in conjunction with the actual motion performance.

[0026] Furthermore, the movable visual acquisition unit 3 includes a two-dimensional slide rail and a pair of monocular cameras slidably connected to the two-dimensional slide rail; the two-dimensional slide rail controls the pair of monocular cameras to perform horizontal translation, and the two monocular cameras respectively acquire the marker points 4 on the printed circuit board 1 and the steel mesh 2. In other embodiments of the present invention, the movable visual acquisition unit 3 includes a two-dimensional slide rail and a monocular camera, the monocular camera being slidably connected to the two-dimensional slide rail via a horizontal rotation mechanism, the horizontal rotation mechanism controlling the monocular camera to rotate around a horizontal axis to change the acquisition direction of the monocular camera, achieving upward or downward acquisition; or, the movable visual acquisition unit 3 includes a two-dimensional slide rail, a monocular camera, and two 45° beam splitters; the movable visual acquisition unit 3 can also be other peripherals such as a calibration ball capable of accurately acquiring the pose of feature points. In other embodiments of the present invention, the monocular camera of the movable visual acquisition unit 3 can be driven by other actuators or a robot system.

[0027] Furthermore, methods for constructing a comprehensive error model include: The movable vision acquisition unit 3 acquires at least two marker points 4 of the printing plate 1 in its initial state, and calculates the initial pose of the printing plate 1. The printing plate 1 is moved to multiple points along a certain trajectory by the motion platform. At each point, at least two marker points 4 of the printing plate 1 are collected by the movable vision acquisition unit 3, and the actual pose of the printing plate 1 at each point is calculated. Based on the initial pose of the printed circuit board 1 and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board 1 at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

[0028] Specifically, the pose of the printed circuit board 1 (or stencil 2) is obtained as follows: multiple marker points 4 are collected, and the plane containing the multiple marker points 4 is the plane containing the printed circuit board 1. If two marker points 4 are collected, they form a line segment. The two-dimensional coordinates of the midpoint of the plane or line segment are represented as (X,Y). The plane rotation angle θ of the printed circuit board 1 can be extracted by connecting multiple sets of line segments, and the pose of the printed circuit board 1 is then represented as (X,Y,θ). The geometric parameters are calculated using the least squares method. In other embodiments of the present invention, gradient descent, particle swarm optimization, etc., can also be used for calculation.

[0029] Geometric parameters include rotation center offset, non-orthogonality error between axes, and other parameters.

[0030] In other embodiments of the present invention, the comprehensive error model can be solved using a neural network. The neural network is trained using the collected data and can directly output the compensated motion result according to the target pose requirement.

[0031] Furthermore, the methods for generating the compensation mapping table include: The printing plate 1 is controlled by a motion platform to align the sets of marker points 4 collected by the movable vision acquisition unit 3 within the field of view. The printing plate 1 and the steel mesh 2 are pressed together by a motion platform; The deviation between the steel mesh 2 and the printed plate 1 under the pressing state is detected by an external detection unit, and the printed plate 1 is moved by a motion platform to eliminate the deviation. (Ideally, the texture of the steel mesh 2 and the printed plate 1 can be made to coincide by aligning them through the movable vision acquisition unit 3, which achieves the final compensation requirement; however, since the optical axes of a pair of monocular cameras are not perfectly aligned, the Z-axis of the motion platform is not completely perpendicular to the steel mesh 2, making it impossible to achieve texture coincidence through a monocular camera in actual situations.) The printing plate 1 is controlled to move away from the stencil 2 via a motion platform; The movable visual acquisition unit 3 is used to acquire each group of marker points 4 again, and the residual pose deviation is calculated. The position of the stencil 2 is changed multiple times (in this embodiment, this is achieved by loosening the stencil 2, manually moving the stencil 2, and then clamping it again), and the residual pose deviation under different relative poses is calculated to construct a compensation mapping table (the compensation mapping table is used to compensate for the optical axis deviation and the Z-axis perpendicularity deviation); the relative pose represents the pose deviation between the stencil 2 and the printed circuit board 1.

[0032] In other embodiments of the present invention, the compensation mapping table can be replaced with polynomial coefficients, and the compensation amount calculation formula is generated by polynomial fitting.

[0033] Furthermore, the control processing unit constructs a calibration database by calculating a comprehensive error model and compensation mapping table under different sizes of printed circuit board 1 and stencil 2.

[0034] Specifically, during the calibration phase, different sizes of printed circuit board 1 and stencil 2 are used to repeatedly obtain the corresponding comprehensive error model and compensation mapping table, thus forming the calibration database. During actual compensation, the calibration file closest to the current size is selected, and benchmark compensation is performed using interpolation based on the comprehensive error model (compensating for the specific dimensional parameters of the motion platform). Then, the adjustment amount of printed circuit board 1 during pressing is compensated according to the compensation mapping table (i.e., the pose parameters X, Y, θ of printed circuit board 1). The compensation is performed sequentially, and the final result is calculated.

[0035] This invention also provides a self-compensation method for a printing system. The printing system includes a stencil 2, a printing plate 1, a motion platform, a movable visual acquisition unit 3, an external detection unit, and a control processing unit. The movable visual acquisition unit 3 is disposed between the stencil 2 and the printing plate 1. The lower surface of the stencil 2 and the upper surface of the printing plate 1 are provided with a plurality of corresponding sets of marking points 4. The method includes the following steps: S1. The movable vision acquisition unit 3 acquires the marking points 4 on the printing plate 1 and the steel mesh 2 to construct a comprehensive error model containing multiple geometric error parameters. S2. Generate a compensation mapping table based on the marker point 4 collected by the movable visual acquisition unit 3 and the deviation detected by the external detection unit. S3. During actual printing, based on the current dimensions of printing plate 1 and stencil 2, the corresponding comprehensive error model parameters and compensation mapping table are called to correct the set trajectory of the motion platform in real time.

[0036] Understandably, before performing the S1 and S2 calibration steps, it is necessary to initialize the system to zero, install the reference stencil 2 and printing plate 1 (since the printing system may install stencils 2 and printing plates 1 of different sizes, the size of the stencil 2 and printing plate 1 used in a single calibration may not be the same as that used in actual use; the reference stencil 2 and printing plate 1 refer to the stencil 2 and printing plate 1 of appropriate size that are manually selected within the operating range of the printing system), and perform the initial static calibration (such as standard nine-point calibration) of the monocular camera and the printing system within the movable vision acquisition unit 3.

[0037] Furthermore, methods for constructing a comprehensive error model include: The movable vision acquisition unit 3 acquires at least two marker points 4 of the printing plate 1 in its initial state, and calculates the initial pose of the printing plate 1. The printing plate 1 is moved to multiple points along a certain trajectory by the motion platform. At each point, at least two marker points 4 of the printing plate 1 are collected by the movable vision acquisition unit 3, and the actual pose of the printing plate 1 at each point is calculated. Based on the initial pose of the printed circuit board 1 and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board 1 at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

[0038] Furthermore, the methods for generating the compensation mapping table include: The printing plate 1 is controlled by a motion platform to align the sets of marker points 4 collected by the movable vision acquisition unit 3 within the field of view. The printing plate 1 and the steel mesh 2 are pressed together by a motion platform; The deviation between the steel mesh 2 and the printing plate 1 under the pressing state is detected by an external detection unit, and the printing plate 1 is moved by a motion platform to eliminate the deviation. The printing plate 1 is controlled to move away from the stencil 2 via a motion platform; The movable visual acquisition unit 3 is used to acquire each group of marker points 4 again, and the residual pose deviation is calculated. The position of the stencil 2 is changed multiple times, and the residual pose deviation under different relative poses is calculated to construct a compensation mapping table; the relative pose represents the pose deviation between the stencil 2 and the printed circuit board 1.

[0039] See Figure 3 For ease of understanding, the following example is provided in this embodiment.

[0040] A comprehensive error model was established using the original model from the XYY platform (wherein...) For the calculation results, (For input) is represented as follows:

[0041]

[0042]

[0043]

[0044]

[0045]

[0046]

[0047]

[0048]

[0049] In the above formula, This refers to the X-axis adjustment amount (horizontal direction) of the motion platform. , This indicates the adjustment amount of the Y1 and Y2 axes of the motion platform. The time indicates that the motion platform is translating. (Time indicates that the motion platform generates a horizontal rotation), X, Y, and θ represent the motion platform's rotation from the origin through the current... , , The adjusted movement is positive with counterclockwise for θ, positive with rightward for X, and positive with upward for Y.

[0050] In this example, the geometric parameters include (That is, the horizontal deviation between the Y1 and Y2 axes is 300mm, the horizontal deviation between the Y2 and X1 axes is 39.5mm, and the vertical deviation between the X1 axis and the line connecting the Y1 and Y2 axes is 136.5mm). These values ​​are used as default values ​​for iteration. The X, Y1, and Y2 axes are set to move from their negative limits to their positive limits, and multiple sets of adjustment values ​​and the actual pose of the printed circuit board at position 1 are collected. .

[0051] Using multiple sets of collected data, combined with the original model and the initial values ​​of geometric parameters, the corrected values ​​of the geometric parameters are obtained through the least squares algorithm.

[0052] Then, obtain different angles. Under the premise that the alignment is accurately determined by an external detection unit, the pose error between the printed circuit board 1 and the stencil 2 acquired by the monocular camera. .

[0053] In final use, the correction values ​​of the geometric parameters are used to align the monocular camera, combined with... The target position is corrected according to the requirements of the external detection unit, and the printing plate 1 and the stencil 2 are finally accurately aligned.

[0054] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the self-compensation method of the printing system described above.

[0055] The key technical point of this invention is: 1. Data acquisition mechanism: A single movable vision system is used to collect multiple data points during the dynamic movement of the platform, which is the basis for achieving systematic dynamic calibration.

[0056] 2. Unified modeling and decoupling of composite errors: Establish a platform integrated kinematic model that can simultaneously accommodate multiple parameters such as rotation center offset and inter-axis error, and solve it through a single calibration process.

[0057] 3. Layered compensation strategy: The first layer corrects the platform's basic motion commands through model parameters; the second layer compensates for residual deviations (mainly optical axis and perpendicularity errors) related to Z-axis height and product dimensions through lookup tables.

[0058] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0059] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A printing system with self-compensation function, comprising a stencil and a printing plate disposed below the stencil, characterized in that, include: A motion platform is used to support the printing plate and control its movement along a set trajectory. A movable vision acquisition unit is set between the stencil and the printing plate to acquire the marking points on the printing plate and the stencil. There are several sets of marking points, with each set of marking points positioned opposite each other on the upper surface of the printing plate and the lower surface of the stencil. An external detection unit is used to detect the deviation between the steel mesh and the printing plate under the pressing state; The control and processing unit is communicatively connected to the motion platform, the movable vision acquisition unit, and the external detection unit. Based on the marker points acquired by the movable vision acquisition unit, the control and processing unit constructs a comprehensive error model containing multiple geometric error parameters, and generates a compensation mapping table by combining the deviations detected by the external detection unit. During actual printing, the control and processing unit calls the corresponding comprehensive error model parameters and compensation mapping table according to the current printing plate and stencil size to correct the set trajectory of the motion platform in real time.

2. The printing system with self-compensation function according to claim 1, characterized in that, The motion platform is an XYY platform, which controls the printed circuit board to perform horizontal translation or horizontal rotation.

3. The printing system with self-compensation function according to claim 1, characterized in that, The movable vision acquisition unit includes a two-dimensional slide rail and a pair of monocular cameras slidably connected to the two-dimensional slide rail; the two-dimensional slide rail controls the pair of monocular cameras to perform horizontal translation, and the two monocular cameras respectively acquire the marked points on the printed circuit board and the steel mesh.

4. The printing system with self-compensation function according to claim 1, characterized in that, Methods for constructing a comprehensive error model include: The initial pose of the printing plate is calculated by acquiring at least two marker points of the printing plate in its initial state using a movable vision acquisition unit. The printing plate is moved to multiple points along a certain trajectory by a motion platform. At each point, at least two marker points of the printing plate are collected by a movable vision acquisition unit, and the actual pose of the printing plate at each point is calculated. Based on the initial pose of the printed circuit board and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

5. The printing system with self-compensation function according to claim 1, characterized in that, Methods for generating compensation mapping tables include: The printing plate is controlled by a motion platform to align the sets of markers acquired by the movable vision acquisition unit within the field of view. The printing plate is pressed into the stencil using a motion platform; The deviation between the steel mesh and the printing plate under the pressing state is detected by an external detection unit, and the printing plate is moved by a motion platform to eliminate the deviation. The printing plate is controlled to leave the stencil via a motion platform; Each set of marker points is collected again using the movable vision acquisition unit, and the residual pose deviation is calculated. The position of the stencil was changed multiple times, and the residual pose deviation under different relative poses was calculated to construct a compensation mapping table; the relative pose represents the pose deviation between the stencil and the printed circuit board.

6. The printing system with self-compensation function according to claim 1, characterized in that, The control processing unit constructs a calibration database by calculating a comprehensive error model and compensation mapping table for different printed circuit board and stencil sizes.

7. A self-compensation method for a printing system, characterized in that, The printing system includes a stencil, a printing plate, a motion platform, a movable vision acquisition unit, an external detection unit, and a control and processing unit. The movable vision acquisition unit is positioned between the stencil and the printing plate. The lower surface of the stencil and the upper surface of the printing plate have corresponding sets of marking points. The method includes the following steps: S1. Collect the marking points on the printed circuit board and steel mesh through the movable vision acquisition unit, and construct a comprehensive error model containing multiple geometric error parameters; S2. Generate a compensation mapping table based on the marker points collected by the movable vision acquisition unit and the deviation detected by the external detection unit. S3. During actual printing, based on the current printing plate and stencil size, the corresponding comprehensive error model parameters and compensation mapping table are called to correct the set trajectory of the motion platform in real time.

8. The self-compensation method for a printing system according to claim 7, characterized in that, Methods for constructing a comprehensive error model include: The initial pose of the printing plate is calculated by acquiring at least two marker points of the printing plate in its initial state using a movable vision acquisition unit. The printing plate is moved to multiple points along a certain trajectory by a motion platform. At each point, at least two marker points of the printing plate are collected by a movable vision acquisition unit, and the actual pose of the printing plate at each point is calculated. Based on the initial pose of the printed circuit board and the control commands of the motion platform, calculate the theoretical pose of the printed circuit board at each motion point. Based on the kinematic model of the motion platform theory, geometric parameters are introduced to construct a comprehensive error model; Based on the actual and theoretical poses of each motion point, the geometric parameters of the integrated error model are calculated.

9. The self-compensation method for a printing system according to claim 7, characterized in that, Methods for generating compensation mapping tables include: The printing plate is controlled by a motion platform to align the sets of markers acquired by the movable vision acquisition unit within the field of view. The printing plate is pressed into the stencil using a motion platform; The deviation between the steel mesh and the printing plate under the pressing state is detected by an external detection unit, and the printing plate is moved by a motion platform to eliminate the deviation. The printing plate is controlled to leave the stencil via a motion platform; Each set of marker points is collected again using the movable vision acquisition unit, and the residual pose deviation is calculated. The position of the stencil was changed multiple times, and the residual pose deviation under different relative poses was calculated to construct a compensation mapping table; the relative pose represents the pose deviation between the stencil and the printed circuit board.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the self-compensation method of the printing system according to any one of claims 7 to 9.

Images