Magnetic suspension nozzle calibration system and method

Abstract

The invention is applicable to the field of printing, and discloses a magnetic suspension nozzle calibration system and method.The magnetic suspension nozzle calibration system comprises a mover backboard, a mover assembly, a stator coil assembly, a stator backboard, a sensor module and a controller, the mover backboard is used for mounting a nozzle, the mover assembly comprises a permanent magnet array, and the controller is connected with the mover assembly; the permanent magnet array is installed on the back face of the rotor back plate, the stator coil assembly comprises a plurality of electromagnetic coils, the electromagnetic coils are installed on the front face of the stator back plate, magnetic suspension gaps are formed between the permanent magnet array and the electromagnetic coils, and six-degree-of-freedom magnetic suspension of the spray head is achieved through magnetic field interaction of the electromagnetic coils and the permanent magnet array. The sensor module is used for measuring the pose of the nozzle in real time, the controller is respectively connected with the sensor module and the plurality of electromagnetic coils, the controller receives the current pose of the nozzle fed back by the sensor module and controls the current of each electromagnetic coil based on the current pose of the nozzle so as to adjust the real-time pose of the nozzle, and online high-precision calibration of the nozzle can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of printing, and in particular to a magnetic levitation printhead calibration system and method. BACKGROUND

[0002] New display manufacturing, especially OLED, QLED, Micro-LED and other next-generation display technologies, are increasingly using high-precision inkjet printing (IJP) technology as its core preparation process. This technology precisely deposits functional materials (such as light-emitting materials, electrode materials, optical film paste, etc.) in the form of droplets on the substrate, realizing the graphical preparation of pixel points, circuits and functional layers, with the advantages of high material utilization, wide substrate range (can cope with flexible curved surfaces), simplified process, and easy realization of large size, etc.

[0003] However, inkjet printing technology faces extremely harsh challenges in large-scale, high-density, high-precision display panel manufacturing, of which the precise positioning and attitude control of the printing printhead are the core keys to determine product yield and performance.

[0004] High pixelization accuracy: when preparing an RGB pixel array, the size of each sub-pixel is only microns to tens of microns. The landing position error of ink droplets must be strictly controlled within ±5µm or even smaller, and any slight deviation can cause fatal defects such as color mixing, uneven brightness or circuit short circuit, resulting in high material and production cost loss.

[0005] Complex substrate surface topography: the substrate of the display panel is not an ideal plane. The printhead must be able to dynamically adjust its height (Z-axis) and attitude (pitch, yaw angle) to ensure that the nozzle and the substrate surface maintain the optimal distance (usually hundreds of microns) throughout the printing stroke, otherwise it will cause droplet flight curve deviation, landing position error or droplet shape out of control.

[0006] Challenge of large-size substrates: G8.5 and above generation production lines need to cope with large glass substrates of more than 2.2 meters x 2.5 meters. The printing system must have extremely high motion stability and global consistency, and any vibration, thermal deformation or guide rail error in the span range will directly translate into printing defects.

[0007] Capacity and throughput pressure: display manufacturing is a capital-intensive industry with extremely high requirements for equipment capacity (UPH, Units Per Hour). The calibration process of the printhead must be fast, automated, and preferably completed in real time online to minimize production downtime. Traditional calibration methods that rely on manual intervention have become a bottleneck in improving production efficiency.

[0008] Therefore, the prior art still needs to be improved and developed. SUMMARY

[0009] The first object of the present application is to provide a magnetic levitation printhead calibration system, which aims to solve the technical problems of poor calibration accuracy and inconvenience of online calibration of existing inkjet printing devices.

[0010] To achieve the above object, the present application provides the following scheme: A magnetic levitation printhead calibration system, comprising a mover back plate, a mover assembly, a stator coil assembly, a stator back plate, a sensor module and a controller, the mover back plate is used for installing a printhead, the mover assembly comprises a permanent magnet array, the permanent magnet array is installed on the back surface of the mover back plate, the stator coil assembly comprises a plurality of electromagnetic coils, a plurality of the electromagnetic coils are installed on the front surface of the stator back plate, a magnetic levitation gap is arranged between the permanent magnet array and the electromagnetic coils, the sensor module is arranged on the mover back plate and located beside the printhead, the sensor module is used for measuring the pose of the printhead in real time, the controller is connected with the sensor module and the plurality of electromagnetic coils respectively, the controller receives the current printhead pose fed back by the sensor module, and controls the current of each electromagnetic coil based on the current printhead pose, so as to adjust the real-time pose of the printhead.

[0011] Preferably, the sensor module comprises a displacement sensor, a gap sensor and an inertial measurement unit, the displacement sensor, the gap sensor and the inertial measurement unit are all arranged on the mover back plate, and the displacement sensor, the gap sensor and the inertial measurement unit are arranged on the circumferential side of the printhead respectively, the displacement sensor is used for measuring the displacement of the printhead in at least one translational direction, the gap sensor is used for measuring the distance between the printhead and the substrate to be printed, and the inertial measurement unit is used for detecting the micro-vibration and angular velocity of the printhead, and the controller is connected with the displacement sensor, the gap sensor and the inertial measurement unit respectively.

[0012] Preferably, the gap sensor comprises a capacitive gap sensor body, a first sensor mounting bracket and a second sensor mounting bracket, the first sensor mounting bracket is installed on the side edge of the stator, the capacitive gap sensor body is installed on the first sensor mounting bracket, and the second sensor mounting bracket is connected with the first sensor mounting bracket and fixes the capacitive gap sensor body on the first sensor mounting bracket.

[0013] Preferably, the mover assembly further comprises an upper mover magnet fixing plate, a middle mover magnet fixing plate and a lower mover magnet fixing plate, the permanent magnet array is mounted on the middle mover magnet fixing plate, the upper mover magnet fixing plate and the lower mover magnet fixing plate are fixed on both sides of the middle mover magnet fixing plate respectively, and the upper mover magnet fixing plate is mounted on the back surface of the mover back plate.

[0014] Preferably, the stator coil assembly further comprises a stator coil mounting member and a stator coil cover plate, the stator coil mounting member is mounted on the front surface of the stator back plate, the electromagnetic coils are mounted on the stator coil mounting member, and the stator coil cover plate covers the plurality of electromagnetic coils and is connected with the stator coil mounting member.

[0015] A second object of the present application is to provide a magnetic levitation nozzle calibration method, which is implemented based on the magnetic levitation nozzle calibration system as described above, and comprises the following steps: based on the structure and dynamic characteristics of the magnetic levitation nozzle calibration system, an initial state space model is constructed as a prediction model, and an offline pre-training is performed on a Gaussian process regression model to obtain initial hyperparameters of the Gaussian process regression model; displacement information, angle information and acceleration information related to the nozzle are obtained through feedback of a sensor module and are fused to obtain a current nozzle pose; based on an actual nozzle pose at a previous moment and an electromagnetic coil current vector at the previous moment, a prediction value of the current nozzle pose is predicted by using the prediction model, and a residual error between the current nozzle pose and the prediction value of the current nozzle pose is calculated; an electromagnetic coil current vector at a current moment is obtained, and the actual nozzle pose at the previous moment, the electromagnetic coil current vector at the previous moment and the residual error between the current nozzle pose and the prediction value of the current nozzle pose are combined to form a new input and output sample pair, and the initial hyperparameters of the Gaussian process regression model are updated by using the new input and output sample pair; based on the current nozzle pose and a to-be-solved control quantity, a residual error prediction mean value and a prediction variance are output by using the updated Gaussian process regression model; the residual error prediction mean value is fused with the prediction model to obtain a corrected prediction model; the corrected prediction model is used as an internal prediction model, the actual nozzle pose is used as an initial state, and a target pose is used as a tracking target, a cost function is defined, a constraint boundary is adjusted based on the prediction variance, and an optimal control input sequence that satisfies the constraint boundary in a future finite time domain is solved; an instant control quantity in the optimal control sequence is converted into a current signal, and the current signal is used to drive the plurality of electromagnetic coils of the stator coil assembly to adjust a real-time pose of the nozzle.

[0016] Preferably, the prediction model is represented as:

[0017] In the formula, represents an actual nozzle pose at a discrete moment , and Indicates at discrete time The electromagnetic coil current vector, The state matrix, This is the control matrix.

[0018] Preferably, the initial state-space model is constructed as a prediction model based on the structural and dynamic characteristics of the magnetic levitation nozzle calibration system, and the Gaussian process regression model is pre-trained offline to obtain the initial hyperparameters of the Gaussian process regression model. This includes: constructing an initial state-space model as a prediction model based on the structural and dynamic characteristics of the magnetic levitation nozzle calibration system; acquiring historical operating data or simulation data as initial training data, wherein the initial training dataset includes multiple initial input and output sample pairs, each of which includes an electromagnetic coil current vector, a target pose, and an actual nozzle pose; calculating the predicted nozzle pose value for each initial input and output sample pair using the prediction model based on the initial training data, and calculating the residual between the predicted nozzle pose value and the actual nozzle pose; pre-training the Gaussian process regression model offline with the electromagnetic coil current vector and target pose of the initial input and output sample pairs as independent variables and the corresponding residual as dependent variables, and optimizing and determining the initial hyperparameters of the Gaussian process regression model by minimizing the residual prediction error, wherein the initial hyperparameters include kernel function parameters and noise variance.

[0019] Preferably, the modified prediction model is expressed as:

[0020] In the formula, Indicates at discrete time The actual position of the nozzle. Indicates at discrete time The actual position of the nozzle. Indicates at discrete time The electromagnetic coil current vector, The state matrix, For the control matrix, This represents the mean of the predicted residuals.

[0021] Preferably, the cost function is defined as follows: Then the cost function Represented as:

[0022] In the formula, Indicates the target pose. Indicates at discrete time The actual position of the nozzle. Indicates at discrete time The electromagnetic coil current vector, denotes the prediction horizon length, which corrects the number of time steps forward predicted by the prediction model in each optimization, denotes the control horizon length, which corrects the number of control steps actually optimized by the prediction model, , is a summation index variable used to iterate over discrete time steps.

[0023] In the scheme, the six-degree-of-freedom non-contact magnetic suspension driving of the nozzle is realized through the magnetic field interaction of the permanent magnet array and the electromagnetic coil, the friction and wear problems of the traditional mechanical transmission are completely avoided, the sensor module is matched to capture the nozzle pose in real time and accurately, and then the controller dynamically regulates the current of each electromagnetic coil based on the measured pose, which not only can realize the micron-level high-precision positioning and attitude calibration to ensure the stability of the printing quality, but also supports the online real-time calibration in the printing process without interrupting the production process, greatly shortens the calibration time, improves the equipment capacity, simplifies the manual intervention process, reduces the maintenance cost, and meets the strict requirements of high-precision, high-efficiency and high-stability in high-end display manufacturing. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor based on the structures shown in the drawings.

[0025] Figure 1 is a structural diagram of the magnetic suspension nozzle calibration system provided by the embodiment of the present application Figure 1 ; Figure 2 is a structural diagram of the magnetic suspension nozzle calibration system provided by the embodiment of the present application Figure 2 ; Figure 3 is a side view of the magnetic suspension nozzle calibration system provided by the embodiment of the present application; Figure 4 is Figure 3 the cross section along A-A direction in Figure 5 is a flow chart of the magnetic suspension nozzle calibration method provided by the embodiment of the present application.

[0026] Explanation of reference numerals: 10. Mover backplate; 20. Mover assembly; 21. Permanent magnet array; 22. Upper fixing plate of mover magnet; 23. Middle fixing plate of mover magnet; 24. Lower fixing plate of mover magnet; 30. Stator coil assembly; 31. Electromagnetic coil; 32. Stator coil mounting component; 33. Stator coil cover plate; 40. Stator backplate; 50. Sensor module; 51. Displacement sensor; 52. Gap sensor; 521. Capacitive gap sensor body; 522. First sensor mounting bracket; 523. Second sensor mounting bracket; 53. Inertial measurement unit; 60. Magnetic levitation gap; 70. Nozzle; 80. Printing substrate. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0028] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0029] It should also be noted that when a component is described as "fixed to" or "set on" another component, it can be directly on the other component or there may be an intervening component present. When a component is described as "connected to" another component, it can be directly connected to the other component or there may be an intervening component present.

[0030] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0031] like Figures 1 to 4 As shown, this is a magnetic levitation nozzle calibration system according to an embodiment of the present invention.

[0032] Please see Figures 1 to 4As shown, the magnetic suspension nozzle calibration system of the embodiment of the application comprises a mover back plate 10, a mover assembly 20, a stator coil assembly 30, a stator back plate 40, a sensor module 50, and a controller. The mover back plate 10 is used to mount a nozzle 70. The mover assembly 20 comprises a permanent magnet array 21 mounted on the back of the mover back plate 10. The stator coil assembly 30 comprises a plurality of electromagnetic coils 31 mounted on the front of the stator back plate 40. A magnetic suspension gap 60 is provided between the permanent magnet array 21 and the electromagnetic coils 31. The six-degree-of-freedom magnetic suspension of the nozzle 70 is achieved through the magnetic field interaction between the electromagnetic coils 31 and the permanent magnet array 21. The sensor module 50 is provided on the mover back plate 10 and located beside the nozzle 70. The sensor module 50 is used to measure the pose of the nozzle 70 in real time. The controller is connected with the sensor module 50 and the plurality of electromagnetic coils 31 respectively. The controller receives the current nozzle pose fed back by the sensor module 50 and controls the current of each electromagnetic coil 31 based on the current nozzle pose to adjust the real-time pose of the nozzle 70.

[0033] It can be understood that the permanent magnet array 21 is arranged in a preset Halbach array or other optimized polarity arrangement to enhance the single-sided magnetic field and reduce coupling. The electromagnetic coils 31 are distributed in a matrix in space. The axis direction of each electromagnetic coil 31 is designed so that the magnetic field generated thereby can effectively interact with the permanent magnet field. By independently and accurately controlling the current size and direction of each electromagnetic coil 31, any direction and size of force and torque can be synthesized on the mover back plate 10 (carrying the nozzle 70), thereby realizing the six-degree-of-freedom fully decoupled contactless driving and suspension along the X, Y, and Z axes and rotation around the three axes.

[0034] In the embodiment, the magnetic suspension gap 60 is usually designed to be between 0.1 mm and 1.0 mm. The optimized design of the gap ensures sufficient control force and stiffness while minimizing the magnetic circuit nonlinearity and coupling effect, thereby providing a guarantee for realizing high linearity and high decoupling precision control.

[0035] In the embodiment, the controller adopts a FPGA+DSP heterogeneous architecture. The FPGA is responsible for high-speed synchronous acquisition of multi-sensor data (acquisition delay ≤0.1 ms) and PWM current driving of the electromagnetic coils 31. The DSP is responsible for running an adaptive algorithm fused with model predictive control (MPC) and Gaussian process (GP), thereby ensuring real-time generation of control instructions.

[0036] In other embodiments, the controller can be any processor or combination of processors capable of realizing the required high-speed data acquisition and real-time model predictive control algorithm calculation, such as an industrial PC, a multi-core microprocessor, etc.

[0037] In the embodiment, the nozzle 70 is installed on the mover back plate 10, and then the current nozzle pose is measured by the sensor module 50, and the measured current nozzle pose is fed back to the controller. The controller compares the current nozzle pose with the target pose, judges whether the deviation value of the current nozzle pose and the target pose is within the deviation threshold, if yes, it indicates that the nozzle 70 is installed in place, if not, it indicates that the nozzle 70 is not installed in place. According to the deviation value of the current nozzle pose and the target pose, a control instruction is generated, and the current of each electromagnetic coil 31 is controlled according to the control instruction, so as to adjust the real-time pose of the nozzle 70, until the current nozzle pose and the target pose are within the deviation threshold.

[0038] In the embodiment, the six-degree-of-freedom non-contact magnetic suspension driving of the nozzle 70 is realized by the magnetic field interaction of the permanent magnet array 21 and the electromagnetic coil 31, which completely avoids the friction and wear problems of traditional mechanical transmission. The sensor module is used to capture the nozzle pose in real time and accurately, and the controller is used to dynamically regulate the current of each electromagnetic coil 31 based on the measured pose. It can not only realize micron-level high-precision positioning and attitude calibration to ensure the stability of printing quality, but also support online real-time calibration during printing without interrupting the production process. The calibration time is greatly shortened, the equipment productivity is improved, the manual intervention process is simplified, the maintenance cost is reduced, and the high-precision, high-efficiency and high-stability requirements of high-end display manufacturing are met.

[0039] In the embodiment, the sensor module 50 includes a displacement sensor 51, a gap sensor 52 and an inertial measurement unit 53. The displacement sensor 51, the gap sensor 52 and the inertial measurement unit 53 are arranged on the mover back plate 10, and the displacement sensor 51, the gap sensor 52 and the inertial measurement unit 53 are arranged on the side of the nozzle 70. The displacement sensor 51 is used to measure the displacement of the nozzle 70 in at least one translation direction; the gap sensor 52 is used to measure the distance between the nozzle 70 and the substrate 80 to be printed; the inertial measurement unit 53 is used to detect the micro-vibration and angular velocity of the nozzle 70, and the controller is connected with the displacement sensor 51, the gap sensor 52 and the inertial measurement unit 53. The configuration of the multi-sensor fusion constitutes a comprehensive and high-bandwidth sensing network. The displacement sensor 51 provides a global positioning reference to ensure the absolute accuracy in the XY plane. The gap sensor 52 directly monitors the key printing parameter (the distance between the nozzle 70 and the substrate), which overcomes the influence of the substrate topography. The inertial measurement unit 53 captures high-frequency micro-vibration to provide information for feedforward compensation. The three work together to provide comprehensive and accurate system state feedback.

[0040] In the embodiment, the displacement sensor 51 is a laser interferometer or a high-resolution grating ruler. The laser interferometer or the ultra-high-resolution grating ruler can provide sub-nanometer resolution to meet the extreme precision requirements of display manufacturing for nozzle 70 positioning.

[0041] In the present embodiment, the displacement sensor 51 comprises two interferometer mirror sets, one of which is arranged transversely on the stator back plate 40, and the other of which is arranged longitudinally on the stator back plate 40. This orthogonal arrangement can directly and independently measure the displacement of the nozzle 70 in the X and Y directions, avoiding calculation errors introduced by coordinate transformation, and improving the directness and accuracy of planar positioning measurement.

[0042] In the present embodiment, the gap sensor 52 is a capacitive or eddy current non-contact gap sensor 52. The capacitive or eddy current gap sensor 52 has the characteristics of non-contact, high resolution and fast response, and can accurately measure the micron-level working gap change without interfering with the printing process.

[0043] In the present embodiment, the gap sensor 52 comprises a capacitive gap sensor body 521, a first sensor mounting bracket 522 and a second sensor mounting bracket 523, the first sensor mounting bracket 522 is mounted on the side of the stator, the capacitive gap sensor body 521 is mounted on the first sensor mounting bracket 522, and the second sensor mounting bracket 523 is connected with the first sensor mounting bracket 522 and fixes the capacitive gap sensor body 521 on the first sensor mounting bracket 522.

[0044] In the present embodiment, the mover assembly 20 further comprises an upper mover magnet fixing plate 22, a middle mover magnet fixing plate 23 and a lower mover magnet fixing plate 24, the permanent magnet array 21 is mounted on the middle mover magnet fixing plate 23, the upper and lower mover magnet fixing plates 22 and 24 are respectively fixed on the two sides of the middle mover magnet fixing plate 23, and the upper mover magnet fixing plate 22 is mounted on the back of the mover back plate 10. The multi-layer fixing plate structure provides stable mechanical protection and accurate positioning for the permanent magnet array 21, preventing it from being damaged or displaced during high-speed movement or impact, ensuring the stability and consistency of the magnetic force output, and facilitating the assembly and maintenance of the entire mover assembly 20.

[0045] In the present embodiment, the stator coil assembly 30 further comprises a stator coil mounting member 32 and a stator coil cover plate 33, the stator coil mounting member 32 is mounted on the front surface of the stator back plate 40, the electromagnetic coils 31 are mounted on the stator coil mounting member 32, and the stator coil cover plate 33 covers the plurality of electromagnetic coils 31 and is connected with the stator coil mounting member 32. The stator coil mounting member 32 provides accurate positioning and heat dissipation channels for the electromagnetic coils 31, while the stator coil cover plate 33 plays a role in protecting the coils, guiding the cooling airflow and preventing foreign matter such as metal debris from entering the magnetic gap, improving the reliability, heat dissipation performance and cleanliness of the entire actuator, and meeting the environmental requirements of high-end manufacturing equipment.

[0046] Please refer to Figure 5As shown, the embodiment of the application also provides a nozzle calibration method, comprising: S101, based on the structure and dynamic characteristics of the magnetic suspension nozzle calibration system, an initial state space model is constructed as a prediction model, and a Gaussian process regression model is pre-trained offline to obtain initial hyperparameters of the Gaussian process regression model; S102, obtaining displacement information, angle information and acceleration information related to the nozzle 70 fed back by the sensor module 50 and fusing them to obtain the current nozzle pose; S103, based on the actual nozzle pose at the last moment and the electromagnetic coil current vector at the last moment, the prediction model is used to predict the current nozzle pose prediction value, and the residual error between the current nozzle pose and the current nozzle pose prediction value is calculated; S104, obtaining the electromagnetic coil current vector at the current moment, and using the actual nozzle pose at the last moment, the electromagnetic coil current vector at the last moment, and the residual error between the current nozzle pose and the current nozzle pose prediction value to form a new input and output sample pair, and using the new input and output sample pair to update the initial hyperparameters of the Gaussian process regression model; S105, based on the current nozzle pose and the to-be-solved control quantity, the updated Gaussian process regression model is used to output the residual prediction mean and the prediction variance; S106, fusing the residual prediction mean with the prediction model to obtain a corrected prediction model; S107, using the corrected prediction model as an internal prediction model, using the actual nozzle pose as the initial state, and using the target pose as the tracking target, defining a cost function, and adjusting the constraint boundary based on the prediction variance, to solve the optimal control input sequence that satisfies the constraint boundary in the future finite time domain; S108, converting the instant control quantity in the optimal control sequence into a current signal, and driving the multiple electromagnetic coils 31 of the stator coil assembly 30 through the current signal to adjust the real-time pose of the nozzle 70.

[0047] In this embodiment, by first constructing an initial state space prediction model and pre-training a Gaussian process regression model to lay the foundation for accurate control, then fusing multi-sensor data to obtain the real pose of the nozzle 70, and dynamically adapting the non-linear characteristics of the system and external disturbances through residual calculation and online model updating, a high-precision corrected prediction model is obtained by fusing the residual prediction mean, and finally based on the corrected model, the optimal control sequence is solved in combination with the cost function and the dynamic constraint boundary to drive the electromagnetic coils 31, not only realizing high-precision calibration of the nozzle pose and ensuring the stability of the printing quality, but also completing the calibration in real time during the printing process without interrupting the production, greatly shortening the calibration time and improving the production efficiency, while having strong adaptability and robustness, which can effectively cope with complex working conditions such as substrate topography changes and environmental vibrations, reduce the cost of manual intervention and maintenance, and adapt to the harsh demands of high-end display manufacturing.

[0048] In the embodiment, in step S101, the prediction model is represented as:

[0049] In the formula, represents the actual pose of the nozzle at the discrete time , represents the current vector of the electromagnetic coil at the discrete time , is a state matrix, is a control matrix.

[0050] In the embodiment, the actual pose of the nozzle includes a translation amount along the direction and a rotation angle around the X, Y and Z axes , such as .

[0051] In the formula, T represents a transpose operation, which converts an original row vector into a column vector.

[0052] In the embodiment, .

[0053] In the formula, T represents a transpose operation, which converts an original row vector into a column vector, respectively represent the electromagnetic coil current for controlling the translation of the nozzle along the X, Y and Z axes, respectively represent the electromagnetic coil current for controlling the rotation of the nozzle around the X, Y and Z axes.

[0054] In the embodiment, based on the structure and dynamic characteristics of the magnetic suspension nozzle calibration system, an initial state space model is constructed as a prediction model, and a Gaussian process regression model is pre-trained offline to obtain initial hyperparameters of the Gaussian process regression model, including: based on the structure and dynamic characteristics of the magnetic suspension nozzle calibration system, an initial state space model is constructed as a prediction model; historical operation data or simulation data are obtained as initial training data, and the initial training data set includes multiple groups of initial input and output sample pairs, each group of initial input and output sample pairs including an electromagnetic coil current vector, a target pose and an actual pose of the nozzle; the prediction model is used to calculate the nozzle pose prediction value of each group of initial input and output sample pairs based on the initial training data, and the residual error between the nozzle pose prediction value and the actual pose of the nozzle is calculated; the electromagnetic coil current vector and the target pose of the initial input and output sample pair are taken as independent variables, and the corresponding residual error is taken as a dependent variable, and the Gaussian process regression model is pre-trained offline, and the initial hyperparameters of the Gaussian process regression model are determined by optimization and minimization of the residual prediction error, including kernel function parameters and noise variance.

[0055] In the embodiment, the initial input and output sample pair is represented as , represents the electromagnetic coil current vector at the time, represents the target pose at the time, represents the actual nozzle pose at the time.

[0056] The residual of the nozzle pose prediction value and the actual nozzle pose is represented as:

[0057] In the formula, represents the nozzle pose prediction value at the time.

[0058] In the embodiment, the initial training data set covers typical working conditions of the magnetic suspension nozzle calibration system, including historical operation data or simulation data under different substrate sizes (such as G8.5 generation substrate) and different printing speeds (50-200 mm / s). Each group of samples contains an electromagnetic coil current vector (6 coil currents, range 0-5 A), a target pose (accuracy ±0.1 μm), and an actual nozzle pose (measured data after sensor fusion). The sample amount of the data set is not less than 10,000 groups.

[0059] In the step S102 in the embodiment, the extended Kalman filter or the unscented Kalman filter algorithm is used to fuse the displacement information, the angle information, and the acceleration information related to the nozzle 70 to obtain the current nozzle pose.

[0060] In the step S104 in the embodiment, when the Gaussian process regression model is updated using the new input and output sample pair, the sparse Gaussian process approximation or the fixed-size sliding time window strategy is used. That is, only the most representative sample pair in the recent period of time is retained, or an induced point set is used to approximate the whole data set, so that the learning ability is guaranteed while the calculation load is greatly reduced, and the time limit requirement of the real-time control cycle is met.

[0061] In the step S106 in the embodiment, the modified prediction model is represented as:

[0062] In the formula, represents the actual nozzle pose at the discrete time , represents the actual nozzle pose at the discrete time , represents the electromagnetic coil current vector at the discrete time ,​ The state matrix, For the control matrix, This represents the predicted mean of the residuals.

[0063] In this embodiment, in step S107, the cost function Represented as:

[0064] In the formula, Indicates the target pose. Indicates at discrete time The actual position of the nozzle. Indicates at discrete time The electromagnetic coil current vector, This indicates the length of the prediction time domain, and corrects the number of time steps the prediction model advances in each optimization. This indicates the length of the control time domain and corrects the actual number of control steps optimized by the prediction model. , This is a summation index variable used to iterate through discrete time steps.

[0065] In this embodiment, adjusting the constraint boundary based on the prediction variance means mapping the prediction variance of the Gaussian process output to a constraint safety margin adjustment factor for control or state variables. For example, when the prediction variance is large (high model uncertainty), the current amplitude constraint is tightened or the pose tracking error constraint is relaxed to avoid instability caused by aggressive control when the model is unreliable; conversely, when the prediction variance is small, the current constraint is relaxed or the pose constraint is tightened to pursue better dynamic performance.

[0066] In this embodiment, the constraints include: electromagnetic coil 31 current amplitude constraint (single coil current ≤ 5A), nozzle pose range constraint (X / Y axis translation range ±100μm, Z axis translation range 50-500μm, rotation angle range around the three axes ±0.5°); the safety margin of the constraint boundary is dynamically adjusted based on the prediction variance output by the Gaussian process regression model, when the prediction variance > 1e -8 When the safety margin is increased by 20%, and the prediction variance is less than 1e... -10 At that time, the safety margin is reduced by 10%.

[0067] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A magnetic levitation showerhead calibration system, comprising: The application relates to a magnetic levitation nozzle calibration system, which comprises a mover back plate, a mover assembly, a stator coil assembly, a stator back plate, a sensor module and a controller, the mover back plate is used for mounting a nozzle, the mover assembly comprises a permanent magnet array, the permanent magnet array is mounted on the back of the mover back plate, the stator coil assembly comprises a plurality of electromagnetic coils, the plurality of electromagnetic coils are mounted on the front of the stator back plate, a magnetic levitation gap is arranged between the permanent magnet array and the electromagnetic coils, the sensor module is arranged on the mover back plate and located on the side of the nozzle, the sensor module is used for measuring the pose of the nozzle in real time, the controller is connected with the sensor module and the plurality of electromagnetic coils respectively, the controller receives the current nozzle pose fed back by the sensor module, and the current of each electromagnetic coil is controlled based on the current nozzle pose so as to adjust the real-time pose of the nozzle.

2. The magnetic levitation showerhead calibration system of claim 1, wherein, The sensor module comprises a displacement sensor, a gap sensor and an inertial measurement unit, the displacement sensor, the gap sensor and the inertial measurement unit are arranged on the mover back plate, and the displacement sensor, the gap sensor and the inertial measurement unit are arranged on the circumferential side of the nozzle respectively, the displacement sensor is used for measuring the displacement of the nozzle in at least one translation direction, the gap sensor is used for measuring the distance between the nozzle and a substrate to be printed, and the inertial measurement unit is used for detecting the micro-vibration and angular velocity of the nozzle, and the controller is connected with the displacement sensor, the gap sensor and the inertial measurement unit respectively.

3. The magnetic levitation showerhead calibration system of claim 2, wherein, The gap sensor comprises a capacitive gap sensor body, a first sensor mounting bracket and a second sensor mounting bracket, the first sensor mounting bracket is mounted on the side of the stator, the capacitive gap sensor body is mounted on the first sensor mounting bracket, and the second sensor mounting bracket is connected with the first sensor mounting bracket and fixes the capacitive gap sensor body on the first sensor mounting bracket.

4. The magnetic levitation showerhead calibration system of claim 1, wherein, The mover assembly further comprises a mover magnet upper fixing plate, a mover magnet middle fixing plate and a mover magnet lower fixing plate, the permanent magnet array is mounted on the mover magnet middle fixing plate, the mover magnet upper fixing plate and the mover magnet lower fixing plate are fixed on the two sides of the mover magnet middle fixing plate respectively, and the mover magnet upper fixing plate is mounted on the back of the mover back plate.

5. The magnetic suspension showerhead calibration system of claim 1, wherein, The stator coil assembly further comprises a stator coil mounting piece and a stator coil cover plate, the stator coil mounting piece is mounted on the front of the stator back plate, the electromagnetic coils are mounted on the stator coil mounting piece, and the stator coil cover plate covers the plurality of electromagnetic coils and is connected with the stator coil mounting piece.

6. A method of calibrating a magnetic levitation showerhead, the method comprising: The nozzle calibration method is realized based on the magnetic levitation nozzle calibration system according to any one of claims 1-5, and the nozzle calibration method comprises the following steps: Based on the structure and dynamic characteristics of the magnetic levitation nozzle calibration system, an initial state space model is constructed as a prediction model, and a Gaussian process regression model is pre-trained offline to obtain initial hyperparameters of the Gaussian process regression model; Obtain displacement information, angle information and acceleration information related to the spray head fed back by the sensor module and fuse them to obtain a current spray head pose; Based on the actual spray head pose at the previous moment and the electromagnetic coil current vector at the previous moment, a prediction model is used to predict a current spray head pose prediction value, and a residual error between the current spray head pose and the current spray head pose prediction value is calculated; An electromagnetic coil current vector at the current moment is obtained, and the actual spray head pose at the previous moment, the electromagnetic coil current vector at the previous moment and the residual error between the current spray head pose and the current spray head pose prediction value are combined to form a new input and output sample pair, and the new input and output sample pair is used to update the initial hyperparameters of the Gaussian process regression model; Based on the current spray head pose and the to-be-solved control quantity, the updated Gaussian process regression model is used to output a residual prediction mean value and a prediction variance; The residual prediction mean value is fused with the prediction model to obtain a corrected prediction model; The corrected prediction model is used as an internal prediction model, the actual spray head pose is used as an initial state, and a target pose is used as a tracking target, a cost function is defined, and a constraint boundary is adjusted based on the prediction variance to solve an optimal control input sequence in a future finite time domain that satisfies the constraint boundary; The instantaneous control quantity in the optimal control sequence is converted into a current signal, and the current signal is used to drive multiple electromagnetic coils of a stator coil assembly to adjust the real-time pose of the spray head.

7. The method of claim 6, wherein, The prediction model is represented as: wherein denotes the actual pose of the spray head at discrete time steps, denotes the electromagnetic coil current vector at discrete time steps, is a state matrix, is a control matrix.

8. The method of claim 6, wherein the magnetic levitation showerhead is calibrated by, Based on the structure and dynamic characteristics of the magnetic suspension spray head calibration system, an initial state space model is constructed as a prediction model, and the Gaussian process regression model is pre-trained offline to obtain initial hyperparameters of the Gaussian process regression model, including: Based on the structure and dynamic characteristics of the magnetic suspension spray head calibration system, an initial state space model is constructed as a prediction model; Historical operation data or simulation data are obtained as initial training data, and the initial training data set includes multiple groups of initial input and output sample pairs, each of which includes an electromagnetic coil current vector, a target pose and an actual spray head pose; Based on the initial training data, the prediction model is used to calculate a spray head pose prediction value of each initial input and output sample pair, and a residual error between the spray head pose prediction value and the actual spray head pose is calculated; The Gaussian process regression model is pre-trained offline with the electromagnetic coil current vector and the target pose of the initial input and output sample pair as independent variables and the corresponding residual error as a dependent variable, and the initial hyperparameters of the Gaussian process regression model are determined by optimization through minimization of the residual prediction error, including kernel function parameters and noise variance.

9. The method of claim 6, wherein, The corrected prediction model is represented as: wherein denotes the actual pose of the spray head at discrete time denotes the actual pose of the spray head at discrete time denotes the actual pose of the spray head at discrete time denotes the actual pose of the spray head at discrete time denotes the current vector of the electromagnetic coil at discrete time denotes the current vector of the electromagnetic coil at discrete time is a state matrix, is a control matrix, denotes the residual prediction mean.

10. The method of claim 6, wherein, The cost function is defined as The cost function is defined as is expressed as: wherein represents the target pose, represents the actual nozzle pose at discrete time steps, represents the electromagnetic coil current vector at discrete time steps, represents the prediction horizon length, correcting the number of time steps the prediction model is forward predicted in each optimization, represents the control horizon length, correcting the number of control steps the prediction model is actually optimized for, , is a summation index variable used to iterate over the discrete time steps.

Images