A Positioning Calibration Method for Gantry Mechanisms of Bonding Machines Based on Solder Point Alignment

Abstract

This invention relates to the field of motor control, and more particularly to a positioning and calibration method for a gantry mechanism of a bonding machine based on solder joint alignment. The method includes: monitoring the speed difference and real-time displacement deviation of two independently controlled servo motors on both sides of the gantry mechanism; calculating the predicted displacement deviation on both sides of the gantry mechanism based on the speed difference and real-time displacement deviation; calculating the deflection angle of the gantry mechanism's crossbeam relative to the guide rail length direction based on the predicted displacement deviation; determining the target speed of the two servo motors based on the deflection angle; and determining the change process of the current speed towards the target speed. This invention utilizes the length of the crossbeam to ensure that the generated corrective torque is physically matched with the required angular deviation; it can automatically plan a smoother speed change rate under high-risk conditions with relatively large deflection angles; and it employs a relatively fast speed change process under normal fine-tuning conditions with relatively small deflection angles, thereby ensuring the system's rapid response capability to minor disturbances.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, and in particular to a positioning and calibration method for a gantry mechanism of a bonding machine based on solder joint alignment. Background Technology

[0002] In the field of precision electronic packaging, the gantry mechanism of bonding machines needs to achieve micron-level positioning accuracy to meet the alignment requirements of solder joints. Traditional gantry synchronous control methods mainly rely on closed-loop speed or displacement feedback compensation from dual-sided servo motors, such as using PID algorithms to eliminate positional deviations on both sides.

[0003] However, in actual operation, the displacement deviation compensation does not consider the differential effect of beam length on torsional stiffness. Under the same displacement deviation, the actual deflection angle of the short beam is significantly greater than that of the long beam. Traditional control outputs a fixed compensation amount, resulting in undercompensation for the short beam and overcompensation for the long beam, which exacerbates the wear of the guide rail slider under off-center load and positioning drift. In addition, the existing control method also has a contradiction in dynamic response. Under large deflection conditions, rapid compensation is prone to triggering mechanical vibration due to sudden changes in inertial torque, while slow adjustment cannot suppress the accumulation of deviation. Moreover, the speed step during the static-dynamic friction transition period can cause jamming or overshoot, disrupting the continuity of the weld point alignment trajectory.

[0004] In summary, current fixed parameter control cannot adapt to different crossbeam specifications, and parameter readjustment is required when replacing the mechanism. Furthermore, there is a lack of dynamic response optimization mechanism to address deflection risks. Therefore, there is an urgent need for a positioning and calibration method for the gantry mechanism of a bonding machine to achieve precise control of weld point alignment. Summary of the Invention

[0005] This invention provides a positioning and calibration method for a gantry mechanism of a bonding machine based on solder joint alignment, which can effectively solve the problems in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment includes:

[0008] Monitor the speed difference and real-time displacement deviation of the two independently controlled servo motors on both sides of the gantry mechanism;

[0009] The predicted displacement deviations on both sides of the gantry mechanism are calculated based on the speed difference and real-time displacement deviation.

[0010] The deflection angle of the crossbeam of the gantry mechanism relative to the length direction of the guide rail is calculated based on the predicted displacement deviation.

[0011] The target speed of the two servo motors is determined based on the deflection angle, and the process of the current speed changing towards the target speed is determined.

[0012] Furthermore, the predicted displacement deviations on both sides of the gantry mechanism are calculated using the following formula:

[0013] ;

[0014] in, To predict displacement deviation, The real-time displacement deviation at the current time t. The time interval from the current time to the predicted time. for The speed difference between the two servo motors at any given time.

[0015] Further, determining the target speed of the two servo motors based on the deflection angle includes:

[0016] Dynamic structural simulation was performed on the gantry mechanism and guide rail to obtain the real-time friction torque difference under the deflection angle;

[0017] The attitude correction torque is calculated based on the real-time friction torque difference.

[0018] The target speed of the two servo motors is generated based on the attitude correction torque.

[0019] Furthermore, dynamic structural simulation is performed on the gantry mechanism and guide rail, including:

[0020] A digital multibody model is constructed, which simplifies the gantry mechanism and guide rail into rigid bodies and kinematic pairs, and imports material properties and contact parameters.

[0021] The digital multibody model solves for the dynamic response under stress using the Newton-Euler equations.

[0022] Further, obtaining the real-time frictional torque difference at the deflection angle includes:

[0023] The deflection angle is input into the digital multibody model to drive the crossbeam to undergo virtual torsion around the center point;

[0024] The simulation engine outputs the positive pressure values ​​of the sliders on both sides in real time and establishes a quantitative mapping between the pressure distribution and the deflection angle.

[0025] Use the preset Stribeck curve to obtain the transient friction coefficient;

[0026] The frictional force on both sides is calculated based on the product of the transient friction coefficient and the normal force value.

[0027] The difference in frictional torque between the two sides is calculated using the length of the lever arm from the center point of the beam to both sides as the real-time frictional torque difference.

[0028] Further, calculating the attitude correction torque based on the real-time friction torque difference includes:

[0029] The vibration amplitude spectrum of the beam is predicted by the vibration transfer function;

[0030] The vibration amplitude exceeding the limit frequency band is determined based on the vibration amplitude spectrum, and an anti-vibration gain coefficient is generated for the vibration amplitude exceeding the limit frequency band;

[0031] The compensation torque corresponding to the vibration resistance gain coefficient is superimposed on the real-time friction torque difference to output the attitude correction torque.

[0032] Furthermore, the vibration resistance gain coefficient is the proportion of vibration energy that needs to be suppressed in the over-limit frequency band.

[0033] Further, generating the target speed of the two servo motors based on the attitude correction torque includes:

[0034] Based on the real-time displacement deviation direction of the crossbeam, the leading side and the lagging side in the target motion direction are defined, and the attitude correction torque is decomposed into the leading side torque component and the lagging side torque component.

[0035] Calculate the leading-side velocity compensation value based on the leading-side torque component, and calculate the lagging-side velocity compensation value based on the lagging-side torque component.

[0036] The initial target velocities on both sides are generated based on the preset reference velocity, the leading-side velocity compensation value, and the lagging-side velocity compensation value.

[0037] The initial target velocity is adjusted by frequency band amplitude constraint based on the vibration amplitude spectrum, and the target velocities on both sides are output.

[0038] Further, the initial target velocity is adjusted by frequency band amplitude constraint based on the vibration amplitude spectrum, and the target velocities on both sides are output, including:

[0039] Extract the over-limit frequency bands whose amplitude exceeds the threshold from the vibration amplitude spectrum;

[0040] Based on the center frequency and bandwidth of each of the aforementioned over-limit frequency bands, generate band-stop filter coefficients;

[0041] A band-stop filter is used to synchronously filter the initial target velocity on the leading side and the initial target velocity on the lagging side;

[0042] The filtered speed is used as the target speed output for the servo motors on both sides.

[0043] Furthermore, determining the change process of the current speed towards the target speed based on the deflection angle includes dynamically planning the rate of change of speed based on the deflection angle, wherein:

[0044] When the deflection angle is greater than or equal to a preset safety threshold, the first rate of change of velocity is used;

[0045] When the deflection angle is less than the preset safety threshold, the second rate of change of speed is used;

[0046] The first rate of change of velocity is lower than the second rate of change of velocity.

[0047] The technical solution of this invention can achieve the following technical effects:

[0048] This invention utilizes the length of the beam to calculate the target velocity based on the deflection angle, ensuring that the generated corrective torque is physically matched with the required angular deviation. During implementation, the real-time dynamic planning based on the deflection angle automatically plans a smoother rate of velocity change under high-risk conditions with relatively large deflection angles. This avoids drastic jumps in speed commands, reduces mechanical vibrations caused by sudden changes in inertial force and torsional torque, and provides a more sufficient and stable transition time for the static-to-dynamic transition of frictional force, preventing jamming or overshoot caused by sudden frictional changes, especially considering the aggravated nonlinearity of frictional force. Under normal fine-tuning conditions with relatively small deflection angles, the control system can employ a relatively fast velocity change process, ensuring rapid response to minor disturbances and maintaining high dynamic performance. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart of a bonding machine gantry mechanism positioning and calibration method based on solder joint alignment;

[0051] Figure 2 This is a flowchart for determining the target speed of two servo motors based on the deflection angle.

[0052] Figure 3 This is a flowchart for dynamic programming of the rate of change of velocity based on the deflection angle. Detailed Implementation

[0053] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0054] like Figure 1 As shown, the positioning and calibration method for the gantry mechanism of a bonding machine based on solder joint alignment includes:

[0055] A1: Monitor the speed difference and real-time displacement deviation of the two independently controlled servo motors on both sides of the gantry mechanism; in implementing this step, as a specific implementation method, the actual speed feedback signals of the two servo motors, such as encoder data, can be collected in real time, and the real-time speed difference can be directly calculated through a subtractor; as for the real-time displacement deviation, it can be obtained by real-time measurement using laser equipment; in this embodiment, the gantry mechanism includes a crossbeam, support structures at both ends of the crossbeam, and a slider located at the bottom of the support structure, with the slider slidably connected to the guide rail;

[0056] A2: Calculate the predicted displacement deviation on both sides of the gantry mechanism based on the speed difference and real-time displacement deviation;

[0057] A3: Calculate the deflection angle of the gantry beam relative to the length direction of the guide rail based on the predicted displacement deviation; it can intuitively reflect the non-perpendicular state of the current gantry beam length direction and the guide rail length direction. Specifically, the deflection angle is the ratio of the predicted displacement deviation to the beam length.

[0058] A4: Determine the target speed of the two servo motors based on the deflection angle, and determine the process of the current speed changing towards the target speed.

[0059] In step A3 of this embodiment, the calculation of the deflection angle includes the key geometric parameter of the beam length. In real-world scenarios, the same displacement deviation will result in drastically different torsional effects on beams of different lengths. In this invention, the deflection angle is directly related to the torsional deformation of the beam itself and the off-center load on the guide rail slider pair. Using the deflection angle as the control basis can directly suppress the fundamental physical factors that lead to equipment damage and precision degradation.

[0060] The compensation action required to correct the deflection angle in this invention essentially generates a corrective torsional torque. The effectiveness of this torque is highly dependent on the length of the crossbeam. The process of calculating the target velocity based on the deflection angle utilizes the length of the crossbeam, ensuring that the generated corrective torque and the required angle deviation are physically matched. During implementation, the real-time dynamic planning process based on the deflection angle in step A4 achieves the following technical advantages:

[0061] In high-risk conditions with relatively large deflection angles, the system will automatically plan a smoother rate of change of speed, thereby avoiding drastic jumps in speed commands and reducing mechanical vibrations caused by sudden changes in inertial force and torsional torque. At the same time, in response to the increased nonlinearity of friction in this situation, the smooth change process provides a more sufficient and stable transition time for the friction to change from static to dynamic, which can prevent jamming or overshoot caused by sudden changes in friction.

[0062] Under normal fine-tuning conditions with a relatively small deflection angle, the control system can employ a relatively fast speed change process, thereby ensuring the system's rapid response to minor disturbances and maintaining high dynamic performance. At the same time, due to the small torsional stress, rapid adjustment will not cause significant vibration.

[0063] As a preferred embodiment of the above, the predicted displacement deviation on both sides of the gantry mechanism is calculated using the following formula:

[0064] ;

[0065] in, To predict displacement deviation, The real-time displacement deviation at the current time t. The time interval from the current time to the predicted time. for The speed difference between the two servo motors at any given time.

[0066] In this step, the instantaneous speed difference monitored in real time is used as input, and the displacement deviation that may accumulate after future time intervals is calculated in advance through integral calculation. The prediction model used has a small amount of computation and a fast operation speed, which can be completed in a very short time period. During the implementation process, the model directly responds to the change in speed difference, so that any factor that causes the instantaneous speed of the two servo motors to be inconsistent will be quickly captured by the model and transformed into the predicted displacement deviation trend, such as load disturbance, friction difference, small deviation of motor response characteristics, control signal fluctuation, etc.

[0067] Maintaining the crossbeam's posture during weld alignment is fundamental to ensuring the consistency and repeatability of the welding head's trajectory throughout its entire stroke. This invention obtains predicted displacement deviations through real-time speed difference integration, enabling proactive sensing of asynchronous movements on both sides of the gantry mechanism. Furthermore, by converting the predicted displacement deviations into physical angles and dynamically and in a closed-loop manner adjusting the target speeds and their changes of the servo motors on both sides, it achieves active, smooth, and precise compensation for the crossbeam's posture.

[0068] As a preferred embodiment of the above, such as Figure 2 As shown, the target speed of the two servo motors is determined based on the deflection angle, including:

[0069] B1: Perform dynamic structural simulation of the gantry mechanism and guide rail to obtain the real-time frictional torque difference under the deflection angle;

[0070] B2: Calculate attitude correction torque based on real-time friction torque difference;

[0071] B3: Generate the target speed of the two servo motors based on the attitude correction torque.

[0072] As a preferred embodiment of the above, dynamic structural simulation of the gantry mechanism and guide rail is performed, including:

[0073] B11: Construct a digital multibody model. The digital multibody model simplifies the gantry mechanism and guide rail into rigid bodies and kinematic pairs, and imports material properties and contact parameters.

[0074] B12: The digital multibody model solves the dynamic response under force using the Newton-Euler equations, enabling online access.

[0075] The above process can be completed during the initialization phase of the control system. The material properties include, but are not limited to, elastic modulus and density. The contact parameters can be specifically the initial coefficients of the slider-guide friction pair. The Newton-Euler equations are the standard mathematical tools for establishing the relationship between rigid body motion and force. The numerical solution methods used, such as recursive algorithms, are existing technologies in engineering software and will not be elaborated here.

[0076] In this preferred scheme, the dynamic stress of the beam-guide rail is simulated online using a digital multibody model, and the deflection angle is converted into structural stress data in real time. This provides a real mechanical state input for torque compensation and avoids the torque mismatch problem caused by neglecting the influence of geometric deformation on contact stress in traditional control.

[0077] As a preferred embodiment of the above, obtaining the real-time friction torque difference at the deflection angle includes:

[0078] B13: Input the deflection angle into the digital multibody model to drive the crossbeam to undergo virtual torsion around the center point;

[0079] B14: The simulation engine outputs the positive pressure values ​​of the sliders on both sides in real time and establishes a quantitative mapping between pressure distribution and deflection angle.

[0080] B15: Obtain the transient friction coefficient by calling the preset Stribeck curve based on the current slider speed;

[0081] B16: Calculate the frictional force on both sides based on the product of the transient friction coefficient and the normal force value;

[0082] B17: Calculate the difference in friction torque between the two sides using the length of the lever arm from the center point of the beam to both sides as the real-time friction torque difference.

[0083] In this preferred embodiment, virtual torsion causes a normal pressure distribution on the sliders at both ends of the crossbeam. The transient friction coefficient is used to describe the characteristic of decreasing friction coefficient at low speeds and stabilizing at high speeds. The Stribeck curve is preset in the model, describing the nonlinear law of the sliding friction coefficient changing with speed: at low speeds, the friction coefficient decreases sharply with increasing speed, specifically in the transition stage from static friction to dynamic friction; at medium speeds, it is in an unstable mixed lubrication region; at high speeds, the friction coefficient tends to stabilize, at which point fluid lubrication dominates. The Stribeck curve can be established through experimental calibration and parametric modeling methods. In this embodiment, the preset curve is used to accurately capture the transient friction characteristics during the start-stop and speed change stages of the servo motor, solving the stuttering problem caused by sudden speed changes.

[0084] Based on pressure distribution mapping and the Stribeck friction model, the velocity-dependent friction torque difference can be accurately quantified, achieving synergistic suppression of torsional deformation compensation and friction disturbance, and eliminating trajectory deviation caused by load asymmetry.

[0085] As a preferred embodiment of the above, calculating the attitude correction torque based on the real-time friction torque difference includes:

[0086] B21: Predicting the vibration amplitude spectrum of the crossbeam using the vibration transfer function;

[0087] B22: Determine the frequency band exceeding the vibration amplitude limit based on the vibration amplitude spectrum, and generate the vibration resistance gain coefficient for the frequency band exceeding the vibration amplitude limit;

[0088] B23: The compensation torque corresponding to the vibration resistance gain coefficient is superimposed on the real-time friction torque difference to output the attitude correction torque.

[0089] In step B21 of this embodiment, the vibration transfer function is pre-calibrated. This process can be completed offline before the equipment leaves the factory. As a specific method of calibration, it may include:

[0090] The mass distribution and damping coefficient of the crossbeam are measured by sensors, such as by measuring the amplitude attenuation when the crossbeam is struck; a sinusoidal torque is input to the servo motor in the range of 0-200Hz, and the vibration response data at the end of the crossbeam is recorded; a mathematical model of the transfer function is established based on the input torque-output vibration data, and the predicted crossbeam vibration amplitude spectrum reflects the vibration response under the current torque disturbance; the input of the vibration transfer function is the real-time friction torque difference, and the output is the vibration amplitude spectrum at the end of the crossbeam, thereby quantifying the amplitude risk value of each frequency band.

[0091] In this embodiment, the vibration resistance gain coefficient is used to generate the reverse compensation torque. In the specific implementation process, it is necessary to first compare the vibration spectrum with the safety threshold, such as the maximum amplitude allowed by micron-level positioning, so as to mark the over-limit frequency band. In some embodiments of the present invention, the over-limit frequency band is the 40-60Hz resonance band.

[0092] As a preferred embodiment of the above, the vibration resistance gain coefficient is the proportion of vibration energy that needs to be suppressed in the out-of-limit frequency band. For example:

[0093] The 60Hz frequency band is an over-limit frequency band, with a corresponding safety threshold of 6μm. The actual amplitude is 10.8μm. Therefore, the required proportion of vibration energy to be suppressed is (10.8-6) / 6=80%, and the vibration resistance gain coefficient is 0.8.

[0094] During implementation, a reference compensation torque can be set for each frequency band during the system calibration stage, and the product of the reference compensation torque and the vibration resistance gain coefficient can be used as the corresponding compensation torque; when multiple frequency bands exceed the limit, multiple compensation torques can be superimposed accordingly.

[0095] As a preferred embodiment of the above, generating the target speed of the two servo motors based on the attitude correction torque includes:

[0096] B31: Based on the real-time displacement deviation direction of the crossbeam, define the leading side and the lagging side in the target motion direction, and decompose the attitude correction torque into the leading side torque component and the lagging side torque component.

[0097] B32: Calculate the leading-side velocity compensation value based on the leading-side torque component, and calculate the lagging-side velocity compensation value based on the lagging-side torque component;

[0098] B33: Generate the initial target velocities on both sides based on the preset reference velocity, the leading side velocity compensation value, and the lagging side velocity compensation value; the initial target velocity allocation can be obtained directly by summation in this step. Of course, the leading side velocity compensation value and the lagging side velocity compensation value need to be marked with positive or negative signs according to the velocity direction.

[0099] B34: Adjust the initial target velocity based on the frequency band amplitude constraint according to the vibration amplitude spectrum, and output the target velocities on both sides.

[0100] In some embodiments of the present invention, the attitude correction torque is decomposed into a leading torque component and a lagging torque component, specifically:

[0101] Leading side moment component = Attitude correction moment * Lagging side lever arm length / Total lever arm length of crossbeam;

[0102] Lagging side torque component = Attitude correction torque - Leading side torque component;

[0103] Wherein, the length of the lagging side lever arm is the vertical projection distance from the center point of the beam to the lagging side guide rail, the length of the leading side lever arm is the vertical projection distance from the center point of the beam to the leading side guide rail, and the total lever arm length of the beam is the sum of the length of the leading side lever arm and the length of the lagging side lever arm.

[0104] In other embodiments of the present invention, calculating the leading-side velocity compensation value based on the leading-side torque component and calculating the lagging-side velocity compensation value based on the lagging-side torque component includes: calculating the leading-side velocity compensation value through a dynamic response model of the leading-side servo motor, wherein the dynamic response model is a first-order inertial system model; and calculating the lagging-side velocity compensation value through a dynamic response model of the lagging-side servo motor, wherein the dynamic response model is a first-order inertial system model.

[0105] During implementation, the leading side inputs the leading torque component to the dynamic response model, outputting a negative velocity compensation value and instructing deceleration to suppress displacement lead; the lagging side inputs the lagging torque component to the dynamic response model, outputting a positive velocity compensation value and instructing acceleration to compensate for displacement lag; the model parameters are calibrated through servo motor step response and braking tests to ensure that the dynamic response matches the actual mechanical characteristics. In this embodiment, the model inputs the torque component, and based on the law of inertia and energy transfer relationships, simulates the dynamic influence of this torque on the beam's motion state in real time, outputting a velocity compensation value; during parameter calibration, the model's inertial parameters are calibrated through a step response test, and the model's damping parameters are calibrated through a braking test, ensuring that the model's output velocity value accurately corresponds to the transient response of the real machinery.

[0106] In this preferred embodiment, the leading and lagging sides are dynamically defined by real-time displacement deviation, and a differentiated compensation strategy is executed accordingly. This enables precise synchronous control of the bonding machine gantry mechanism. The control process is entirely based on autonomous decision-making according to the motion state, rather than static structural parameters. The leading side, due to its ahead displacement, requires deceleration compensation to suppress overshoot, while the lagging side, due to its lagging displacement, requires acceleration compensation to catch up with the trajectory. The velocity compensation values ​​on both sides are generated through independent dynamic response models of servo motors. The leading side model emphasizes high-damping deceleration response, while the lagging side model emphasizes high-sensitivity acceleration response, thereby accurately adapting to the electromechanical characteristics under different motion states.

[0107] As a preferred embodiment of the above, adjusting the initial target velocity based on the vibration amplitude spectrum using frequency band amplitude constraints, and outputting the target velocities on both sides, includes:

[0108] B341: Extracting the over-limit frequency bands whose amplitude exceeds the threshold from the vibration amplitude spectrum;

[0109] B342: Generate band-stop filter coefficients based on the center frequency and bandwidth of each over-limit frequency band;

[0110] B343: A band-stop filter is used to synchronously filter the initial target velocity on the leading side and the initial target velocity on the lagging side;

[0111] B344: The filtered speed is used as the target speed output for the servo motors on both sides.

[0112] In this preferred embodiment, the initial target velocity is not a static scalar, but a dynamically updated time-series signal. The essence of band-stop filtering is to process the velocity command signal stream in real time. Through this preferred embodiment, the same filter bank is used but applied independently to both sides of the velocity command, ensuring synchronous decay of resonant energy while preserving the differentiated characteristics of dynamic compensation on both sides, thus solving the phase mismatch problem caused by step-by-step filtering.

[0113] A band-stop filter is an electronic device that eliminates signals in a specific frequency band. By scanning the speed command signal in real time, it can remove speed fluctuation components in the resonant frequency band, and the speed command in the non-resonant frequency band can pass through without attenuation, thus ensuring the accuracy of the trajectory of the moving subject. That is, in this embodiment, the targeted band-stop filter based on the over-limit frequency band can filter out the speed fluctuation components generated by the speed signal.

[0114] In this embodiment, the vibration amplitude spectrum is converted into an anti-vibration gain coefficient and superimposed on the real-time friction torque difference in steps B22 and B23, thus constructing a pre-suppression mechanism for the vibration source. Its technical advantage lies in that, for the detected over-limit resonant frequency band, a reverse damping gain is injected in advance in the process of generating the attitude correction torque to actively cancel the resonant energy excited by the crossbeam due to friction fluctuations, thereby weakening the vibration intensity from the torque source. The above process can lay a low vibration basis for the secondary filtering in step B34, forming a dual vibration suppression closed loop from the torque source to the velocity end.

[0115] As a preliminary option for the above embodiments, such as Figure 3 As shown, the process of determining the change of current velocity towards target velocity based on the deflection angle includes dynamic programming of the velocity change rate based on the deflection angle, where:

[0116] When the deflection angle is greater than or equal to the preset safety threshold, the first rate of change of velocity is used;

[0117] When the deflection angle is less than the preset safety threshold, the second rate of change of speed is used;

[0118] The rate of change of the first velocity is lower than the rate of change of the second velocity.

[0119] During implementation, when the deflection angle is greater than or equal to the preset safety threshold, the system automatically adopts a low rate of change, i.e., the first rate of change, to ensure that the crossbeam transitions smoothly under large torsional stress, effectively suppressing mechanical vibrations induced by inertial impact and frictional abrupt changes. When the deflection angle is less than the preset safety threshold, the system switches to a high rate of change, i.e., the second rate of change, thereby leveraging the dynamic response capability of the servo motor to quickly eliminate minute displacement deviations. In this embodiment, the dual-mode speed regulation mechanism based on the angle threshold ensures system stability under high torsional risk while also taking into account response efficiency in fine-tuning mode.

[0120] In some embodiments of the present invention, the preset safety threshold can be specifically set based on experience, or, as an optimized acquisition scheme, the process of determining the preset safety threshold includes:

[0121] Alternating loads are applied to the digital multibody model to simulate the torsional condition of the beam; the maximum equivalent stress in the stress concentration area of ​​the beam is extracted and matched with the fatigue strength curve of the beam material; the deflection angle corresponding to the first time the maximum equivalent stress exceeds 60%-70% of the material yield strength is calibrated as the preset safety threshold.

[0122] Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to this invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for positioning and calibration of a gantry mechanism in a bonding machine based on solder joint alignment, characterized in that, include: Monitor the speed difference and real-time displacement deviation of the two independently controlled servo motors on both sides of the gantry mechanism; The predicted displacement deviations on both sides of the gantry mechanism are calculated based on the speed difference and real-time displacement deviation. The deflection angle of the crossbeam of the gantry mechanism relative to the length direction of the guide rail is calculated based on the predicted displacement deviation. Determining the target speed of the two servo motors based on the deflection angle includes: Dynamic structural simulation is performed on the gantry mechanism and guide rail to obtain the real-time friction torque difference at the deflection angle; the attitude correction torque is calculated based on the real-time friction torque difference; and the target speed of the two servo motors is generated according to the attitude correction torque. And determine the process of the change from the current speed to the target speed.

2. The method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment according to claim 1, characterized in that, The predicted displacement deviation on both sides of the gantry mechanism is calculated using the following formula: ； in, To predict displacement deviation, The real-time displacement deviation at the current time t. The time interval from the current time to the predicted time. for The speed difference between the two servo motors at any given time.

3. The method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment according to claim 1, characterized in that, Dynamic structural simulation of the gantry mechanism and guide rail is performed, including: A digital multibody model is constructed, which simplifies the gantry mechanism and guide rail into rigid bodies and kinematic pairs, and imports material properties and contact parameters. The digital multibody model solves for the dynamic response under stress using the Newton-Euler equations.

4. The method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment according to claim 3, characterized in that, Obtaining the real-time friction torque difference at the deflection angle includes: The deflection angle is input into the digital multibody model to drive the crossbeam to undergo virtual torsion around the center point; The simulation engine outputs the positive pressure values ​​of the sliders on both sides in real time and establishes a quantitative mapping between the pressure distribution and the deflection angle. Use the preset Stribeck curve to obtain the transient friction coefficient; The frictional force on both sides is calculated based on the product of the transient friction coefficient and the normal force value. The difference in frictional torque between the two sides is calculated using the length of the lever arm from the center point of the beam to both sides as the real-time frictional torque difference.

5. The method for positioning and calibration of a gantry mechanism of a bonding machine based on solder joint alignment according to any one of claims 1, 3, and 4, characterized in that, The attitude correction torque is calculated based on the real-time friction torque difference, including: The vibration amplitude spectrum of the beam is predicted by the vibration transfer function; The vibration amplitude exceeding the limit frequency band is determined based on the vibration amplitude spectrum, and an anti-vibration gain coefficient is generated for the vibration amplitude exceeding the limit frequency band; The compensation torque corresponding to the vibration resistance gain coefficient is superimposed on the real-time friction torque difference to output the attitude correction torque.

6. The method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment according to claim 5, characterized in that, The vibration resistance gain coefficient is the proportion of vibration energy that needs to be suppressed in the over-limit frequency band.

7. The bonding machine gantry mechanism positioning and calibration method based on solder joint alignment according to claim 5, characterized in that, Generating the target speed of the two servo motors based on the attitude correction torque includes: Based on the real-time displacement deviation direction of the crossbeam, the leading side and the lagging side in the target motion direction are defined, and the attitude correction torque is decomposed into the leading side torque component and the lagging side torque component. Calculate the leading-side velocity compensation value based on the leading-side torque component, and calculate the lagging-side velocity compensation value based on the lagging-side torque component. The initial target velocities on both sides are generated based on the preset reference velocity, the leading-side velocity compensation value, and the lagging-side velocity compensation value. The initial target velocity is adjusted by frequency band amplitude constraint based on the vibration amplitude spectrum, and the target velocities on both sides are output.

8. The method for positioning and calibration of a bonding machine gantry mechanism based on solder joint alignment according to claim 7, characterized in that, The initial target velocity is adjusted by frequency band amplitude constraint based on the vibration amplitude spectrum, and the target velocities on both sides are output, including: Extract the over-limit frequency bands whose amplitude exceeds the threshold from the vibration amplitude spectrum; Based on the center frequency and bandwidth of each of the aforementioned over-limit frequency bands, generate band-stop filter coefficients; A band-stop filter is used to synchronously filter the initial target velocity on the leading side and the initial target velocity on the lagging side; The filtered speed is used as the target speed output for the servo motors on both sides.

9. The method for positioning and calibration of a gantry mechanism of a bonding machine based on solder joint alignment according to claim 1, characterized in that, The process of determining the change of current speed towards target speed based on the deflection angle includes dynamically planning the rate of change of speed based on the deflection angle, wherein: When the deflection angle is greater than or equal to a preset safety threshold, the first rate of change of velocity is used; When the deflection angle is less than the preset safety threshold, the second rate of change of speed is used; The first rate of change of velocity is lower than the second rate of change of velocity.

Images