Pressure feedback-based toy surface pattern transfer depth adaptive control method and device

Abstract

This invention provides a method and apparatus for adaptive control of the transfer depth of patterns on toy surfaces based on pressure feedback. The method includes acquiring three-dimensional data of the toy surface and extracting surface features; constructing a dynamic transfer rate prediction model that couples ink rheology and time-varying pressure, based on ink rheological properties; generating the pressure distribution matrix of the contact surface between the transfer head and the toy in real time, based on the initial pose of the transfer head and the target dynamic pressure impulse; calculating the pressure spatial gradient field; and obtaining the microscopic shear displacement field based on a hyperelastic constitutive model. The method also evaluates the microscopic contact area ratio based on the statistical characteristics of the pressure distribution matrix and calculates the adaptive impedance stiffness and adaptive holding time based on the microscopic contact area ratio and the dynamic transfer rate. This invention utilizes the microscopic capillary rheological effect of ink under pressure to fill microscopic pits, achieving cascaded control of macroscopic adhesion under pressure control and microscopic filling under time control, ensuring effective ink transfer while avoiding damage to the workpiece.

Description

Technical Field

[0001] This invention relates to the field of milling cutter wear assessment technology, specifically to an adaptive control method and device for the transfer depth of toy surface patterns based on pressure feedback. Background Technology

[0002] Toy surface pattern transfer is a common surface decoration process that uses a transfer head to transfer ink patterns onto the toy surface, thereby enhancing the product's appearance and added value. As the market's demands for toy appearance quality continue to rise, the need for transferring complex curved surfaces, high-precision designs, and high-quality patterns is increasing.

[0003] Currently, existing methods for controlling the transfer of patterns on toy surfaces typically employ static control strategies based on preset pressure or depth. This involves obtaining the workpiece surface via 3D scanning, planning the downward trajectory of the transfer head, and applying a constant target pressure or holding time after contact. However, this existing transfer control method suffers from the following technical drawbacks in practical applications:

[0004] 1. Existing technologies neglect the time-cumulative effect of ink rheology and the influence of microscopic suspensions, which easily leads to "blurred" transfer patterns. Existing methods mostly use absolute pressure as a static control target, without considering the microscopic fluid dynamics characteristics of ink under temperature and pressure. In the actual transfer process, when the microscopic roughness of the toy surface is large, air retention and microscopic suspensions are easily generated between the transfer head and the workpiece. At the same time, if the loading rate is slow or the holding time is short under the same pressure, the ink will detach before it has fully rheologically changed. In this case, simply increasing the static pressure depth will not only fail to promote the ink to effectively fill the microscopic pits, but will also easily damage the toy workpiece, making it difficult to ensure the saturation and clarity of the transfer.

[0005] 2. Existing technologies lack the ability to perceive and control microscopic slippage on the contact surface, which can easily lead to pattern stretching and deformation during the transfer of complex curved surfaces. When transferring complex curved surfaces with abrupt curvature changes (such as the edges of pits), the transfer head (such as a silicone head) will experience uneven local deformation under pressure. Existing technologies only focus on whether the average value of macroscopic contact pressure meets the standard, and cannot monitor the changes in pressure spatial gradient caused by local friction and deformation differences. Furthermore, they cannot perceive and suppress the microscopic slippage trend on the surface of the transfer head. This microscopic slippage directly leads to uncontrollable stretching or compression of the pattern during the transfer process, which seriously affects the pixel-level alignment accuracy of the pattern.

[0006] 3. Existing technologies struggle to balance the contradiction between macroscopically sufficient fit and avoiding excessive pressure locally. They also have poor adaptability to irregular curved surfaces. When faced with the irregular curved surface features of toys, existing technologies mostly employ rigid pressure control or fixed posture trajectories, which make it difficult to achieve compliant yielding at points of curvature abrupt change. In areas with high curvature, the lack of compliance often leads to stress concentration and damage to the workpiece, while in flat areas, excessive yielding may result in incomplete fit. Therefore, multi-dimensional adaptive decoupling control of force and position cannot be achieved.

[0007] 4. Existing physical control models are mostly statically set, making it difficult to adapt to the drift of material properties and environmental factors, resulting in poor long-term production consistency. The transfer process is highly dependent on the rheological properties of the ink, but in actual production, the activation energy and viscosity of different batches of ink fluctuate, and changes in environmental temperature and humidity can also cause changes in ink properties. Existing static physical models are fixed after initial calibration and cannot self-correct parameters according to the actual transfer effect, resulting in a significant decrease in transfer depth and quality when material or environmental drift occurs. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method and apparatus for adaptive control of toy surface pattern transfer depth based on pressure feedback, in order to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: an adaptive control method for the transfer depth of patterns on toy surfaces based on pressure feedback, comprising the following steps:

[0010] S1: Collect three-dimensional data of the toy surface and extract surface features. Combine the ink rheological properties to construct a dynamic transfer rate prediction model that couples ink layer rheology and time-varying pressure, and generate the initial pose of the transfer head and the target dynamic pressure impulse.

[0011] S2: Real-time acquisition of the pressure distribution matrix between the transfer head and the toy contact surface, calculation of the pressure spatial gradient field, and decoupling based on the hyperelastic constitutive model to obtain the microscopic shear displacement field;

[0012] S3: Evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and dynamic transfer rate.

[0013] S4: Calculate the macroscopic pose adjustment based on adaptive impedance stiffness, and perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment;

[0014] S5: Drive the transfer head to perform cascade control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model by combining the actual transfer rate feedback from machine vision.

[0015] In a preferred embodiment, the three-dimensional data of the toy surface is specifically point cloud data of the toy surface, the surface features are specifically the three-dimensional curvature and normal vector information of the area to be transferred, and the target dynamic pressure impulse is specifically the integral target value of the pressure applied by the transfer head over time to achieve the desired transfer depth.

[0016] As a preferred embodiment, the specific process of S1 includes:

[0017] The 3D point cloud data is meshed, and the curvature of each mesh vertex is calculated.

[0018] Based on the current transfer temperature and ink activation energy, calculate the theoretical pressure impulse required to achieve the target transfer rate under different curvatures;

[0019] Based on the theoretical pressure impulse and the planned printing speed, the target dynamic pressure impulse and initial pose sequence at the moment of contact of the transfer head are generated.

[0020] The process of generating the initial pose sequence is as follows:

[0021] Divide the target dynamic pressure impulse by the planned imprinting time to obtain the target normal contact force;

[0022] Based on the surface normal vector of each grid vertex, the target normal contact force is decomposed into the pressing depth and attitude tilt angle of the transfer head along the machine coordinate system, realizing the accurate mapping from time-varying pressure impulse to six-dimensional initial pose in space.

[0023] The formula for predicting the dynamic transition rate of the current grid point at time t is:

[0024] ;

[0025] in, This represents the ink layer transfer rate at time t, where t is the pressurization time. Real-time pressure at all times for Real-time temperature at any given moment This represents the ink rheological coefficient as a function of temperature. This represents the reference pressure, where n represents the pressure nonlinearity exponent. This represents the ink activation energy, where R is the gas constant.

[0026] As a preferred embodiment, the specific process of S2 includes:

[0027] By spatially differentiating the pressure distribution matrix, the pressure gradients along the X and Y axes can be obtained.

[0028] The pressure gradient, normal pressure, and interfacial friction coefficient are input into the hyperelastic constitutive model of silicone to inversely deduce the microscopic shear displacement field of the contact surface.

[0029] The formula for calculating the pressure space gradient field at the current moment is:

[0030] ;

[0031] in, Represents the pressure space gradient field. and These are the directional derivatives of the pressure distribution matrix along the X and Y axes of the contact surface, respectively.

[0032] Based on the Coulomb friction model and the hyperelastic constitutive model, the decoupling calculation formula for the microscopic shear displacement field is as follows:

[0033] ;

[0034] in, This represents the microscopic shear displacement field at the previous time step. This represents the compliance matrix of the silicone tip. This represents the geometric characteristic coefficient that transforms the pressure gradient into an equivalent driving stress. This represents the coefficient of kinetic friction at the interface. Indicates normal pressure, This represents the unit vector of the tangential slip direction.

[0035] As a preferred embodiment, the specific process of S3 is as follows:

[0036] A micro-contact area ratio evaluation function is constructed using the proportion of effective contact units and the pressure variance in the pressure distribution matrix.

[0037] When the micro-contact area ratio is below the threshold, the pressing depth is not increased, and the holding time is extended by calculation to achieve micro-capillary filling of ink.

[0038] The formula for calculating the micro-contact area ratio at the current moment is:

[0039] ;

[0040] in, The micro-contact area ratio is represented by N, the number of effective sensing units within the contact surface is represented by H, and H represents the step function (when...). (Take 1 if it is true, otherwise take 0). This represents the actual pressure value detected by the i-th sensing unit. Indicates the microscopic contact threshold pressure. This represents the standard deviation of the pressure matrix. Indicates average pressure. This represents the roughness influencing factor;

[0041] The formula for calculating the adaptive pressure holding time based on the micro-contact area ratio is as follows:

[0042] ;

[0043] in, Indicates the adaptive holding time. Indicates the time compensation coefficient. This indicates the ink viscosity at the current temperature.

[0044] Meanwhile, the formula for calculating the adaptive impedance stiffness based on the dynamic transfer rate error is as follows:

[0045] ;

[0046] in This represents the adaptive impedance stiffness at the current moment. This indicates the preset reference impedance stiffness. To indicate a minimum value and prevent the denominator from being 0, This represents the combined sensitivity factor that combines the surface curvature with the current pressure gradient magnitude.

[0047] As a preferred embodiment, the specific process of S4 is as follows:

[0048] Calculate the macroscopic pose adjustment based on pressure deviation and adaptive impedance stiffness;

[0049] The decoupled microscopic shear displacement field is mapped to the tangential translation and torsional compensation of the transfer head;

[0050] The macroscopic adjustment amount and the reverse pre-compensation amount are superimposed and output to the motion controller;

[0051] The formula for calculating the macroscopic pose adjustment is:

[0052] ;

[0053] in, This represents the pose adjustment amount at the current moment. The inverse matrix representing the target impedance stiffness. This indicates the global pressure deviation at the current moment. Indicates target impedance damping. Indicates the rate of change of pose adjustment;

[0054] The formula for calculating the optimal target pose at the current moment after inverse pre-compensation is as follows:

[0055] ;

[0056] in, This represents the optimal target pose at the current moment (including XYZ translation and rotation). Indicates the initial pose. This represents the amount of macroscopic pose adjustment. This represents the macroscopic pose fusion gain. This represents the shearing pre-compensation gain matrix, with the negative sign indicating inverse compensation.

[0057] As a preferred embodiment, the specific process of S5 includes:

[0058] Input the optimal target pose into the multi-axis servo system to perform pose transformation;

[0059] Once the pressure reaches the target threshold, the adaptive pressure holding time will be used. Controlling the piezoelectric ceramic to remain locked completes the time-cascaded micro-filling process;

[0060] After the transfer is demolded, the actual transfer rate score of the transfer pattern is obtained through a machine vision system;

[0061] The rheological coefficient is corrected based on the gradient descent method by utilizing the deviation between the actual transfer rate and the target transfer rate.

[0062] The formula for updating the ink rheological coefficient in the dynamic transfer rate prediction model is as follows:

[0063] ;

[0064] in, This indicates the updated ink rheological coefficient. This indicates the ink rheological coefficient before the update. Indicates the learning rate. Indicates the target transfer rate. Indicates the actual transfer rate. This represents the partial derivative of the transfer rate with respect to the rheological coefficient.

[0065] As a preferred embodiment, a pressure feedback-based adaptive control device for the transfer depth of toy surface patterns, used to implement the above-mentioned pressure feedback-based adaptive control method for the transfer depth of toy surface patterns, includes:

[0066] The prediction and initialization module is used to collect three-dimensional data of the toy surface and extract surface features. It combines the rheological properties of ink to construct a dynamic transfer rate prediction model coupled with ink layer rheology and time-varying pressure, and generates the initial pose of the transfer head and the target dynamic pressure impulse.

[0067] The microscopic decoupling module is used to collect the pressure distribution matrix between the transfer head and the toy contact surface in real time, calculate the pressure spatial gradient field, and obtain the microscopic shear displacement field based on the hyperelastic constitutive model.

[0068] An adaptive parameter calculation module is used to evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and to calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and the dynamic transfer rate.

[0069] The pose compensation decision module is used to calculate the macroscopic pose adjustment based on the adaptive impedance stiffness, and to perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment.

[0070] The cascaded control and model update module is used to drive the transfer head to perform cascaded control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model in combination with the actual transfer rate feedback from machine vision.

[0071] This invention provides a method and apparatus for adaptive control of the transfer depth of patterns on toy surfaces based on pressure feedback, which has the following beneficial effects:

[0072] 1. This invention effectively solves the problem of ink transfer blurring caused by static pressure control in the prior art by constructing a dynamic transfer rate prediction model coupled with ink layer rheology and time-varying pressure, and introducing a micro-contact area ratio evaluation mechanism. At the same time, by upgrading the control target to a dynamic target pressure impulse, and calculating the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, when the micro-contact area ratio is lower than the threshold, the pressure depth is not increased but the holding time is adaptively extended. This invention utilizes the micro-capillary rheological effect of ink under pressure to fill micro-dimples, realizing the cascade control of pressure control for macro-adhesion and time control for micro-filling, ensuring effective ink transfer while avoiding damage to the workpiece.

[0073] 2. This invention calculates the pressure spatial gradient field and decouples it to obtain the microscopic shear displacement field, and then performs reverse pre-compensation, which effectively overcomes the problem of pattern stretching deformation caused by microscopic slippage during the transfer of complex curved surfaces. The pressure spatial gradient field is obtained by spatially differentiating the pressure distribution matrix of the contact surface. The microscopic shear displacement field is derived by combining the hyperelastic constitutive model and the friction model. It is mapped as tangential translation and torsional compensation amount and superimposed with macroscopic pose adjustment amount. This realizes the application of reverse compensation force to counteract the slippage trend before microscopic slippage occurs, thereby significantly improving the alignment accuracy and deformation control capability of the pattern on complex curved surfaces.

[0074] 3. This invention calculates adaptive impedance stiffness based on dynamic transfer rate error and comprehensive sensitivity factor, realizing macroscopic compliant fit of the transfer head to irregular curved surfaces, avoiding local stress concentration. The impedance stiffness is dynamically adjusted according to the comprehensive sensitivity factor combining the curvature of the surface and the current pressure gradient amplitude, and the macroscopic pose adjustment is calculated based on pressure deviation and damping. At curvature abrupt changes, the system exhibits low stiffness characteristics to achieve compliant yielding, and high stiffness characteristics in flat areas to ensure transfer pressure. Thus, multi-dimensional adaptive decoupling control of transfer force and position is realized, improving the system's adaptability to complex curved surfaces and operational safety.

[0075] 4. This invention combines the actual transfer rate feedback from machine vision with gradient correction of the rheological coefficient in the dynamic transfer rate prediction model, overcoming the influence of material property drift and environmental fluctuations on transfer consistency. After transfer and demolding, the actual transfer rate score is obtained through machine vision, and the ink rheological coefficient is corrected along the partial derivative direction based on the gradient descent method. This closed-loop self-learning mechanism enables the prediction model to automatically update core parameters according to the actual transfer effect, effectively resisting the interference caused by material and environmental drift, and ensuring the consistency of transfer depth and quality in long-term production. Attached Figure Description

[0076] Figure 1 This is a flowchart of the adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to the present invention.

[0077] Figure 2 This is a schematic diagram of the adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to the present invention.

[0078] Figure 3 This is a block diagram of the pressure feedback-based adaptive control device for the transfer depth of toy surface patterns according to the present invention. Detailed Implementation

[0079] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0080] like Figure 1 and 2 As shown, this embodiment of the invention provides an adaptive control method for the transfer depth of patterns on a toy surface based on pressure feedback, comprising the following steps:

[0081] S1: Collect three-dimensional data of the toy surface and extract surface features. Combine the ink rheological properties to construct a dynamic transfer rate prediction model that couples ink layer rheology and time-varying pressure, and generate the initial pose of the transfer head and the target dynamic pressure impulse.

[0082] In this embodiment, the three-dimensional data of the toy surface is the point cloud data of the toy surface obtained by a high-precision 3D scanner; the surface features refer to the three-dimensional curvature and normal vector information of the area to be transferred; the dynamic transfer rate prediction model is a nonlinear physical model that considers the fluid dynamics characteristics of ink, combines the pressure loading rate, holding time and temperature factors, and predicts the actual degree of ink transfer; the target dynamic pressure impulse refers to the integral target value of the pressure applied by the transfer head over time in order to achieve the desired transfer depth.

[0083] Specifically, to accurately reflect the microscopic physical nature of ink transfer during the transfer process, three-dimensional data of the toy surface is collected and surface features are extracted. A dynamic transfer rate prediction model coupling ink layer rheology and time-varying pressure is constructed based on the ink rheological properties, generating the initial pose of the transfer head and the target dynamic pressure impulse, including:

[0084] The 3D point cloud data is meshed, and the curvature of each mesh vertex is calculated.

[0085] Based on the current transfer temperature and ink activation energy, calculate the theoretical pressure impulse required to achieve the target transfer rate under different curvatures;

[0086] Based on the theoretical pressure impulse and the planned printing speed, the target dynamic pressure impulse and initial pose sequence at the moment of contact of the transfer head are generated.

[0087] The process of generating the initial pose sequence is as follows:

[0088] Divide the target dynamic pressure impulse by the planned imprinting time to obtain the target normal contact force;

[0089] Based on the surface normal vectors of each grid vertex, the target normal contact force is decomposed into the pressing depth and attitude tilt angle of the transfer head along the machine coordinate system, thereby achieving a precise mapping from time-varying pressure impulse to the initial six-dimensional spatial pose.

[0090] Specifically, the formula for predicting the dynamic transition rate of the current grid point at time t is:

[0091] ;

[0092] in, This represents the ink layer transfer rate at time t, where t is the pressurization time. Real-time pressure at all times for Real-time temperature at any given moment This represents the ink rheological coefficient as a function of temperature. This represents the reference pressure, where n represents the pressure nonlinearity exponent. This represents the ink activation energy, where R is the gas constant.

[0093] It should be noted that traditional methods only focus on the static mapping of absolute pressure, while the essence of ink transfer is the fluid transfer of ink under the influence of temperature and pressure. Under the same pressure, a slow loading rate or a short holding time will cause the ink to detach before sufficient rheological transformation, resulting in a blurry transfer. This model upgrades the control objective from the static target pressure to the dynamic target transfer rate, enabling the system to adaptively decide whether to increase the instantaneous pressure or extend the holding time based on real-time pressure history and temperature.

[0094] Furthermore, a target control benchmark conforming to the micro-rheological physical laws of ink was obtained, providing a physical basis for the subsequent realization of dynamic adaptive control of transfer depth.

[0095] S2: Real-time acquisition of the pressure distribution matrix between the transfer head and the toy contact surface, calculation of the pressure spatial gradient field, and decoupling based on the hyperelastic constitutive model to obtain the microscopic shear displacement field;

[0096] In this embodiment, the pressure distribution matrix is ​​the two-dimensional pressure distribution data of the contact surface acquired in real time by an array of thin-film pressure sensors embedded in the end face of the transfer head. The pressure spatial gradient field refers to the rate of change of the pressure distribution matrix in the two-dimensional direction of the contact surface, reflecting the degree of local pressure concentration. The microscopic shear displacement field refers to the distribution of surface microscopic slippage and tensile displacement caused by local friction and deformation differences when the transfer head (such as a silicone head) is pressed on an irregular curved surface.

[0097] Specifically, to address the pattern stretching and deformation problem caused by microscopic slippage of the silicone head during complex curved surface transfer, the pressure spatial gradient field is calculated, and the microscopic shear displacement field is obtained by decoupling based on a hyperelastic constitutive model, including:

[0098] By spatially differentiating the pressure distribution matrix, the pressure gradients along the X and Y axes can be obtained.

[0099] By inputting the pressure gradient, normal pressure, and interfacial friction coefficient into the hyperelastic constitutive model of silicone, the microscopic shear displacement field of the contact surface is inversely calculated.

[0100] The formula for calculating the pressure space gradient field at the current moment is:

[0101] ;

[0102] in, Represents the pressure space gradient field. and These are the directional derivatives of the pressure distribution matrix along the X and Y axes of the contact surface, respectively.

[0103] Based on the Coulomb friction model and the hyperelastic constitutive model, the decoupling calculation formula for the microscopic shear displacement field is as follows:

[0104] ;

[0105] in, This represents the microscopic shear displacement field at the previous time step. This represents the compliance matrix of the silicone tip. This represents the geometric characteristic coefficient that transforms the pressure gradient into an equivalent driving stress. This represents the coefficient of kinetic friction at the interface. Indicates normal pressure, This represents the unit vector of the tangential slip direction.

[0106] Understandably, when the transfer head is in contact with the edge of the pit or the abrupt change in curvature, the macroscopic pressure may have reached the target, but the excessive local pressure gradient causes the silicone head to have a microscopic slippage tendency. At this time, the microscopic shear displacement field deduced will deviate significantly from the zero point, accurately indicating the direction and magnitude of the microscopic displacement of the pattern that is about to be stretched or compressed.

[0107] Furthermore, it breaks through the limitation of traditional technology that only focuses on the absolute value of macroscopic pressure, and infers microscopic deformation through the pressure gradient field, providing direct data basis for subsequent reverse pre-compensation of pattern anti-deformation.

[0108] S3: Evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and dynamic transfer rate.

[0109] Specifically, the microscopic contact area ratio refers to the ratio of the actual effective contact area between the transfer head and the toy surface (excluding air gaps between microscopic roughness peaks) to the theoretical projected area. Adaptive impedance stiffness determines the degree of yielding and compliance of the transfer head under force; adaptive holding time refers to the cascaded control parameters that extend the pressure application time to allow ink to fill the microscopic pores through rheology when the pressure is insufficient to overcome the suspension caused by microscopic roughness.

[0110] To overcome the problem of poor transfer printing caused by differences in micro-roughness, the micro-contact area ratio is evaluated based on the statistical characteristics of the pressure distribution matrix. Adaptive impedance stiffness and adaptive holding time are calculated based on the micro-contact area ratio and dynamic transfer rate, including:

[0111] A micro-contact area ratio evaluation function is constructed using the proportion of effective contact units and the pressure variance in the pressure distribution matrix.

[0112] When the micro-contact area ratio is below the threshold, the pressure depth is not increased, and the holding time is extended by calculation to achieve micro-capillary filling of the ink.

[0113] The formula for calculating the micro-contact area ratio at the current moment is:

[0114] ;

[0115] in, The micro-contact area ratio is represented by N, the number of effective sensing units within the contact surface is represented by H, and H represents the step function (when...). (Take 1 if it is true, otherwise take 0). This represents the actual pressure value detected by the i-th sensing unit. Indicates the microscopic contact threshold pressure. This represents the standard deviation of the pressure matrix. Indicates average pressure. This represents the roughness influencing factor;

[0116] In this embodiment, the formula for calculating the adaptive pressure holding time based on the micro-contact area ratio is as follows:

[0117] ;

[0118] in, Indicates the adaptive holding time. Indicates the time compensation coefficient. This indicates the ink viscosity at the current temperature.

[0119] Meanwhile, the formula for calculating the adaptive impedance stiffness based on the dynamic transfer rate error is as follows:

[0120] ;

[0121] in This represents the adaptive impedance stiffness at the current moment. This indicates the preset reference impedance stiffness. To indicate a minimum value and prevent the denominator from being 0, This represents the combined sensitivity factor that combines the surface curvature with the current pressure gradient magnitude.

[0122] Understandably, when a toy's surface roughness causes air to trap, simply increasing the pressure will only damage the toy. Instead, extending the holding time will prevent this. This method allows the ink to slowly rheomorphize and fill microscopic pits under pressure, fundamentally solving the problem of blurring caused by microscopic suspension. This establishes a new cascade control paradigm where pressure determines macroscopic adhesion and time determines microscopic filling.

[0123] S4: Calculate the macroscopic pose adjustment based on adaptive impedance stiffness, and perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment;

[0124] In this embodiment, the optimal target pose includes not only the macroscopic Z-axis pressing depth and the attitude tilt angle along the curved surface required to maintain the target pressure, but also the small tangential translation and deflection angle applied in advance to counteract micro-slippage; reverse pre-compensation refers to controlling the transfer head to move in the reverse direction in advance according to the predicted pattern stretching direction, and using contact friction to counteract the micro-slippage trend of the silicone head.

[0125] Specifically, in order to completely eliminate the deformation of the curved transfer pattern, the macroscopic pose adjustment is calculated based on adaptive impedance stiffness, and inverse pre-compensation is performed by integrating the microscopic shear displacement field, including:

[0126] Calculate the macroscopic pose adjustment based on pressure deviation and adaptive impedance stiffness;

[0127] The decoupled microscopic shear displacement field is mapped to the tangential translation and torsional compensation of the transfer head;

[0128] The macroscopic adjustment amount and the reverse pre-compensation amount are superimposed and output to the motion controller;

[0129] The formula for calculating the macroscopic pose adjustment is:

[0130] ;

[0131] in, This represents the pose adjustment amount at the current moment. The inverse matrix representing the target impedance stiffness. This indicates the global pressure deviation at the current moment. Indicates target impedance damping. This represents the rate of change of pose adjustment, used to provide damping to suppress oscillations;

[0132] The formula for calculating the optimal target pose at the current moment after inverse pre-compensation is as follows:

[0133] ;

[0134] in, This represents the optimal target pose at the current moment (including XYZ translation and rotation). Indicates the initial pose. This represents the amount of macroscopic pose adjustment. This represents the macroscopic pose fusion gain. This represents the shearing pre-compensation gain matrix, with the negative sign indicating inverse compensation.

[0135] Therefore, when the transfer head enters the high curvature abrupt change zone, the system not only maintains the smooth fit of the pressure through macroscopic adjustment, but also locks the microscopic slippage trend in advance through reverse pre-compensation, fundamentally ensuring the pixel-level alignment accuracy and deformation control of the pattern.

[0136] S5: Drive the transfer head to perform cascade control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model in combination with the actual transfer rate feedback from machine vision.

[0137] In this embodiment, in order to achieve a high-precision and high-consistency transfer effect, it is necessary to coordinate the macroscopic pose control and microscopic time cascade control, and introduce a model self-learning mechanism based on rheological physics closed loop.

[0138] Specifically, the cascaded control of the transfer head, driven by the optimal target pose and adaptive holding time, and updated in conjunction with machine vision feedback, includes:

[0139] Input the optimal target pose into the multi-axis servo system to perform pose transformation;

[0140] Once the pressure reaches the target threshold, the adaptive pressure holding time will be used. Controlling the piezoelectric ceramic to remain locked completes the time-cascaded micro-filling process;

[0141] After the transfer is demolded, the actual transfer rate score of the transfer pattern is obtained through a machine vision system;

[0142] The rheological coefficient is corrected based on the gradient descent method by utilizing the deviation between the actual transfer rate and the target transfer rate.

[0143] Specifically, the formula for updating the ink rheological coefficient in the dynamic transfer rate prediction model is as follows:

[0144] ;

[0145] in, This indicates the updated ink rheological coefficient. This indicates the ink rheological coefficient before the update. Indicates the learning rate. Indicates the target transfer rate. This represents the actual transfer rate calculated by the machine vision system based on the color saturation, edge sharpness, and area without defects in the transferred pattern. This represents the partial derivative of the transfer rate with respect to the rheological coefficient, characterizing the sensitivity of the rheological parameters to the final transfer effect.

[0146] Specifically, the rheological properties of different batches of ink fluctuate slightly, and changes in environmental temperature and humidity can also alter ink viscosity, making it difficult for static physical models to maintain accuracy over time. By extracting the actual transfer effect through visual closed-loop processing and correcting the core rheological parameters along the gradient direction, the prediction model gains adaptability.

[0147] Specifically, when material properties drift, the automatic correction prediction model makes the impulse distribution and holding time of the next transfer more accurate; combined with macroscopic impedance compliance control, microscopic shear pre-compensation and spatiotemporal cascade holding mechanism, the full-dimensional adaptive control of the transfer depth and microscopic quality of the toy surface pattern is realized.

[0148] like Figure 3 As shown, this embodiment also provides a pressure feedback-based adaptive control device for the transfer depth of toy surface patterns, used to implement the above-mentioned pressure feedback-based adaptive control method for the transfer depth of toy surface patterns, including:

[0149] The prediction and initialization module is used to collect three-dimensional data of the toy surface and extract surface features. It combines the rheological properties of ink to construct a dynamic transfer rate prediction model coupled with ink layer rheology and time-varying pressure, and generates the initial pose of the transfer head and the target dynamic pressure impulse.

[0150] The microscopic decoupling module is used to collect the pressure distribution matrix between the transfer head and the toy contact surface in real time, calculate the pressure spatial gradient field, and obtain the microscopic shear displacement field based on the hyperelastic constitutive model.

[0151] An adaptive parameter calculation module is used to evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and to calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and the dynamic transfer rate.

[0152] The pose compensation decision module is used to calculate the macroscopic pose adjustment based on the adaptive impedance stiffness, and to perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment.

[0153] The cascaded control and model update module is used to drive the transfer head to perform cascaded control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model in combination with the actual transfer rate feedback from machine vision.

[0154] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for adaptive control of the transfer depth of patterns on a toy surface based on pressure feedback, characterized in that, Includes the following steps: S1: Collect three-dimensional data of the toy surface and extract surface features. Combine the ink rheological properties to construct a dynamic transfer rate prediction model that couples ink layer rheology and time-varying pressure, and generate the initial pose of the transfer head and the target dynamic pressure impulse. S2: Real-time acquisition of the pressure distribution matrix between the transfer head and the toy contact surface, calculation of the pressure spatial gradient field, and decoupling based on the hyperelastic constitutive model to obtain the microscopic shear displacement field; S3: Evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and dynamic transfer rate. S4: Calculate the macroscopic pose adjustment based on adaptive impedance stiffness, and perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment; S5: Drive the transfer head to perform cascade control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model by combining the actual transfer rate feedback from machine vision.

2. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 1, characterized in that, The three-dimensional data of the toy surface specifically refers to the point cloud data of the toy surface, the surface features specifically refer to the three-dimensional curvature and normal vector information of the area to be transferred, and the target dynamic pressure impulse specifically refers to the integral target value of the pressure applied by the transfer head over time to achieve the desired transfer depth.

3. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 2, characterized in that, The specific process of S1 includes: The 3D point cloud data is meshed, and the curvature of each mesh vertex is calculated. Based on the current transfer temperature and ink activation energy, calculate the theoretical pressure impulse required to achieve the target transfer rate under different curvatures; Based on the theoretical pressure impulse and the planned printing speed, the target dynamic pressure impulse and initial pose sequence at the moment of contact of the transfer head are generated. The process of generating the initial pose sequence is as follows: Divide the target dynamic pressure impulse by the planned imprinting time to obtain the target normal contact force; Based on the surface normal vector of each grid vertex, the target normal contact force is decomposed into the pressing depth and attitude tilt angle of the transfer head along the machine coordinate system, realizing the accurate mapping from time-varying pressure impulse to six-dimensional initial pose in space. The formula for predicting the dynamic transition rate of the current grid point at time t is: ； in, This represents the ink layer transfer rate at time t, where t is the pressurization time. Real-time pressure at all times for Real-time temperature at any given moment This represents the ink rheological coefficient as a function of temperature. This represents the reference pressure, where n represents the pressure nonlinearity exponent. This represents the ink activation energy, where R is the gas constant.

4. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 1, characterized in that, The specific process of S2 includes: By spatially differentiating the pressure distribution matrix, the pressure gradients along the X and Y axes can be obtained. The pressure gradient, normal pressure, and interfacial friction coefficient are input into the hyperelastic constitutive model of silicone to inversely deduce the microscopic shear displacement field of the contact surface. The formula for calculating the pressure space gradient field at the current moment is: ； in, Represents the pressure space gradient field. and These are the directional derivatives of the pressure distribution matrix along the X and Y axes of the contact surface, respectively. Based on the Coulomb friction model and the hyperelastic constitutive model, the decoupling calculation formula for the microscopic shear displacement field is as follows: ； in, This represents the microscopic shear displacement field at the previous time step. This represents the compliance matrix of the silicone tip. This represents the geometric characteristic coefficient that transforms the pressure gradient into an equivalent driving stress. This represents the coefficient of kinetic friction at the interface. Indicates normal pressure, This represents the unit vector of the tangential slip direction.

5. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 1, characterized in that, The specific process of S3 is as follows: A micro-contact area ratio evaluation function is constructed using the proportion of effective contact units and the pressure variance in the pressure distribution matrix. When the micro-contact area ratio is below the threshold, the pressing depth is not increased, and the holding time is extended by calculation to achieve micro-capillary filling of ink. The formula for calculating the micro-contact area ratio at the current moment is: ； in, The micro-contact area ratio is represented by N, the number of effective sensing units within the contact surface is represented by H, and H represents the step function (when...). (Take 1 if it is true, otherwise take 0). This represents the actual pressure value detected by the i-th sensing unit. Indicates the microscopic contact threshold pressure. This represents the standard deviation of the pressure matrix. Indicates average pressure. This represents the roughness influencing factor; The formula for calculating the adaptive pressure holding time based on the micro-contact area ratio is as follows: ； in, Indicates the adaptive holding time. Indicates the time compensation coefficient. This indicates the ink viscosity at the current temperature. Meanwhile, the formula for calculating the adaptive impedance stiffness based on the dynamic transfer rate error is as follows: ； in This represents the adaptive impedance stiffness at the current moment. This indicates the preset reference impedance stiffness. To indicate the minimum value and prevent the denominator from being 0, This represents the combined sensitivity factor that combines the surface curvature with the current pressure gradient magnitude.

6. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 1, characterized in that, The specific process of S4 is as follows: Calculate the macroscopic pose adjustment based on pressure deviation and adaptive impedance stiffness; The decoupled microscopic shear displacement field is mapped to the tangential translation and torsional compensation of the transfer head; The macroscopic adjustment amount and the reverse pre-compensation amount are superimposed and output to the motion controller; The formula for calculating the macroscopic pose adjustment is: ； in, This represents the pose adjustment amount at the current moment. The inverse matrix representing the target impedance stiffness. This indicates the global pressure deviation at the current moment. Indicates target impedance damping. Indicates the rate of change of pose adjustment; The formula for calculating the optimal target pose at the current moment after inverse pre-compensation is as follows: ； in, This represents the optimal target pose at the current moment (including XYZ translation and rotation). Indicates the initial pose. This represents the amount of macroscopic pose adjustment. This represents the macroscopic pose fusion gain. This represents the shearing pre-compensation gain matrix, with the negative sign indicating inverse compensation.

7. The adaptive control method for the transfer depth of toy surface patterns based on pressure feedback according to claim 1, characterized in that, The specific process of S5 includes: Input the optimal target pose into the multi-axis servo system to perform pose transformation; Once the pressure reaches the target threshold, the adaptive pressure holding time will be used. Controlling the piezoelectric ceramic to remain locked completes the time-cascaded micro-filling process; After the transfer is demolded, the actual transfer rate score of the transfer pattern is obtained through a machine vision system; The rheological coefficient is corrected based on the gradient descent method by utilizing the deviation between the actual transfer rate and the target transfer rate. The formula for updating the ink rheological coefficient in the dynamic transfer rate prediction model is as follows: ； in, This indicates the updated ink rheological coefficient. This indicates the ink rheological coefficient before the update. Indicates the learning rate. Indicates the target transfer rate. Indicates the actual transfer rate. This represents the partial derivative of the transfer rate with respect to the rheological coefficient.

8. A pressure feedback-based adaptive control device for the transfer depth of toy surface patterns, used to implement the pressure feedback-based adaptive control method for the transfer depth of toy surface patterns according to any one of claims 1-7, characterized in that, include: The prediction and initialization module is used to collect three-dimensional data of the toy surface and extract surface features. It combines the rheological properties of ink to construct a dynamic transfer rate prediction model coupled with ink layer rheology and time-varying pressure, and generates the initial pose of the transfer head and the target dynamic pressure impulse. The microscopic decoupling module is used to collect the pressure distribution matrix between the transfer head and the toy contact surface in real time, calculate the pressure spatial gradient field, and obtain the microscopic shear displacement field based on the hyperelastic constitutive model. An adaptive parameter calculation module is used to evaluate the micro-contact area ratio based on the statistical characteristics of the pressure distribution matrix, and to calculate the adaptive impedance stiffness and adaptive pressure holding time based on the micro-contact area ratio and the dynamic transfer rate. The pose compensation decision module is used to calculate the macroscopic pose adjustment based on the adaptive impedance stiffness, and to perform inverse pre-compensation by fusing the microscopic shear displacement field to obtain the optimal target pose at the current moment. The cascaded control and model update module is used to drive the transfer head to perform cascaded control based on the optimal target pose and adaptive holding time, and update the dynamic transfer rate prediction model in combination with the actual transfer rate feedback from machine vision.

Images