Coated glass carrying control method and system

Abstract

The invention relates to the technical field of transportation devices, and discloses a coated glass carrying control method and system, and the method comprises the steps: respectively collecting surface reflection spectrum data, dimension specification parameters and carrying force arm information; extracting film layer combination characteristics, and carrying out bearing capacity calculation by combining dimension specification parameters and carrying force arm information to obtain a theoretical stress extreme value; acquiring a pressure rise rate and a steady-state leakage rate, and performing airtight characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage rate to obtain an interface adsorption efficiency value; carrying out constraint boundary fusion on the theoretical stress extreme value and the interface adsorption efficiency value to obtain a double constraint envelope; time sequence deduction is conducted on the mechanical arm based on double constraint envelope, and a space motion time sequence is obtained; a self-adaptive inertia carrying instruction is generated based on the space motion time sequence; the carrying stability of the coated glass can be improved.

Description

Technical Field

[0001] This invention relates to the field of transportation equipment technology, and in particular to a method and system for controlling the handling of coated glass. Background Technology

[0002] Most existing coated glass handling technologies employ general-purpose vacuum adsorption or mechanical clamping methods, with their control logic primarily relying on fixed handling parameters set based on the geometric dimensions and weight of the original glass sheet. However, with the increasing complexity of coating processes, the bonding strength and surface characteristics of different film systems vary significantly. Traditional technologies lack non-contact sensing capabilities of the microscopic physical properties of the coated surface, often neglecting the impact of film bonding characteristics contained in surface reflectance spectral data on the overall load-bearing limit. In practical operations, simply setting static mechanical parameters based on size specifications cannot accurately calculate the theoretical stress extremes under specific working conditions. This leads to frequent film peeling, scratches, or hidden cracks when handling highly sensitive or large-sized coated glass due to localized gripping stress exceeding the film's tolerance threshold, making it difficult to improve handling cycle time while ensuring finished product yield.

[0003] Existing solutions typically rely solely on static vacuum pressure thresholds to determine the adsorption state, lacking in-depth analysis of the dynamic gas tightness characteristics of the adsorption gas path. Traditional methods fail to effectively monitor the pressure rise rate and steady-state leakage, and cannot quantitatively assess the interfacial adsorption efficiency between the glass and the suction device. Due to the lack of a dual constraint envelope mechanism based on rigid stress limits and flexible adsorption efficiency in the control system, the robotic arm cannot dynamically adjust acceleration and torque according to the real-time spatial motion sequence during spatial movements. This open-loop or weakly coupled control mode is highly susceptible to workpiece slippage, vibration, or even drop accidents during high-speed start-stop or complex trajectory changes due to insufficient adaptive inertia compensation, severely restricting the safety and intelligence level of high-end coated glass automated production lines. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method for controlling the handling of coated glass to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for controlling the handling of coated glass, comprising:

[0006] S1: Collect surface reflectance spectrum data, dimensional specifications, and handling arm information of the coated glass respectively;

[0007] S2: Extract the film bonding characteristics from the surface reflectance spectrum data, and combine them with the size specification parameters and the handling force arm information to calculate the load-bearing capacity and obtain the theoretical stress extreme value of the coated glass;

[0008] S3: Collect the pressure rise rate and steady-state leakage of the adsorption gas path, and perform airtightness characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass.

[0009] S4: The theoretical stress extreme value and the interface adsorption efficiency value are fused by constraint boundary to obtain the double constraint envelope of the coated glass;

[0010] S5: Based on the dual constraint envelope, perform time-series deduction on the robotic arm to obtain the spatial motion time sequence of the robotic arm;

[0011] S6: Generate an adaptive inertia transport command for the coated glass based on the spatial motion timing.

[0012] The process of collecting surface reflectance spectral data, dimensional specifications, and handling arm information of the coated glass includes:

[0013] The coated glass was scanned across the entire wavelength range to obtain the surface reflectance spectral data of the coated glass;

[0014] Capture the edge contour point cloud of the coated glass to obtain the dimensional specifications of the coated glass;

[0015] The load feedback signal of the robotic arm is collected to determine the handling arm information of the robotic arm.

[0016] The process of extracting the film bonding characteristics from the surface reflectance spectral data, and combining this with the dimensional specifications and the handling arm information to calculate the load-bearing capacity, yields the theoretical stress extreme value of the coated glass, including:

[0017] The surface reflectance spectral data is subjected to resolution time-frequency extraction to obtain the film bonding characteristics of the surface reflectance spectral data;

[0018] The dimensional specifications and the handling lever information are topologically discretized to obtain the global load distribution matrix of the coated glass;

[0019] The theoretical stress extremum of the coated glass is obtained by performing heterogeneous convolution between the film layer bonding features and the global load distribution matrix.

[0020] The pressure rise rate and steady-state leakage of the adsorption gas path are collected, including:

[0021] Monitor the real-time gas pressure data of the adsorption gas path and perform differential calculations on the real-time gas pressure data to obtain the pressure rise rate of the adsorption gas path.

[0022] The flow rate data of the adsorption gas path when it is in equilibrium is obtained as the steady-state leakage of the adsorption gas path.

[0023] The process of performing an inversion of the airtightness characteristics of the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass includes:

[0024] By mapping the pressure rise rate and the steady-state leakage amount to the gap flow resistance, the surface roughness of the adsorption contact surface is obtained.

[0025] Contact mechanics analysis was performed on the coated glass based on the surface roughness to obtain the interfacial adsorption efficiency value of the coated glass.

[0026] The formula for calculating the interfacial adsorption efficiency value is as follows:

[0027] ;

[0028] in, This is the interfacial adsorption efficiency value. It is the reference static friction coefficient. It refers to the complexity of the surface texture of the coated glass. This is the steady-state leakage amount. It is the equivalent leakage path length. It is the adsorption pressure difference. It is the dynamic attenuation coefficient. It is the rate of pressure rise. It is an exponential function.

[0029] The step of fusing the theoretical stress extremum and the interfacial adsorption efficiency value into a constraint boundary to obtain the dual constraint envelope of the coated glass includes:

[0030] By inversely mapping the theoretical stress extremum, the rigid damage boundary of the coated glass is obtained;

[0031] The flexible adsorption boundary of the coated glass was obtained by performing a balance analysis on the interfacial adsorption efficiency value.

[0032] The overlapping safe region of the rigid damage boundary and the flexible adsorption boundary is extracted as the dual constraint envelope of the coated glass.

[0033] The step of performing time-series deduction on the robotic arm based on the dual constraint envelope to obtain the spatial motion time sequence of the robotic arm includes:

[0034] Based on the dual constraint envelope, the transport path of the robotic arm is subjected to full path constraint mapping to obtain the local acceleration threshold sequence of the robotic arm;

[0035] The target velocity profile curve of the robotic arm is obtained by smoothly fitting the local acceleration threshold sequence.

[0036] The spatial motion timing of the robotic arm is obtained by performing periodic discrete interpolation on the target velocity profile curve.

[0037] The process of generating adaptive inertia transport commands for the coated glass based on the spatial motion time sequence includes:

[0038] The instantaneous pose data of the spatial motion sequence is analyzed, and the rotational inertia is calculated in combination with the size specification parameters to obtain the dynamic load inertia of the robotic arm.

[0039] The feedforward current compensation value of the robotic arm is generated by the dynamic load inertia.

[0040] Based on the feedforward current compensation value, the basic servo commands of the spatial motion timing are modified by torque superposition to obtain the adaptive inertia handling command of the coated glass.

[0041] To address the above problems, the present invention also provides a handling control system for coated glass, the system comprising:

[0042] The data acquisition module is used to collect surface reflectance spectrum data, dimensional specifications, and handling arm information of the coated glass.

[0043] The load-bearing capacity calculation module is used to extract the film bonding characteristics of the surface reflectance spectrum data, and combine the size specification parameters and the handling lever arm information to calculate the load-bearing capacity and obtain the theoretical stress extreme value of the coated glass.

[0044] The airtightness characteristic inversion module is used to collect the pressure rise rate and steady-state leakage of the adsorption gas path, and perform airtightness characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass.

[0045] The constraint boundary fusion module is used to fuse the theoretical stress extreme value and the interface adsorption efficiency value into a constraint boundary to obtain the dual constraint envelope of the coated glass.

[0046] The timing simulation module is used to perform timing simulation on the robotic arm based on the dual constraint envelope to obtain the spatial motion timing of the robotic arm.

[0047] The instruction generation module is used to generate adaptive inertia transport instructions for the coated glass based on the spatial motion timing.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. This invention acquires surface reflectance spectral data and extracts film bonding characteristics. It then calculates the theoretical stress extremum by combining dimensional specifications and lever arm information. Simultaneously, it inverts the interfacial adsorption efficiency value based on the pressure rise rate and steady-state leakage of the adsorption gas path, achieving precise quantification of the rigid damage boundary and flexible adsorption boundary of the coated glass. The advantage of this approach lies in constructing a dual constraint envelope of theoretical stress and adsorption efficiency, solving the surface damage problem caused by neglecting the microscopic tolerance of the film layer in traditional handling, as well as the risk of drop due to insufficient airtightness assessment. While ensuring the physical integrity of the coated glass, it improves the safety and adsorption reliability of the handling process by extracting the overlapping safe area of ​​the rigid and flexible boundaries.

[0050] 2. This invention uses a dual-constraint envelope to perform time-series deduction on the robotic arm to obtain the spatial motion timing, and further combines instantaneous pose data to analyze the dynamic load inertia, generating feedforward current compensation values ​​to correct the basic servo commands. Another advantage of this scheme is that it achieves adaptive inertia handling control, which can adjust the torque output in real time according to different glass specifications and motion states, overcoming the lag of traditional PID control under variable load conditions; through smooth fitting of local acceleration thresholds, it suppresses residual vibration and workpiece slippage of the robotic arm during high-speed start-up and shutdown phases, thereby improving the handling cycle time and operational efficiency of automated production lines while ensuring motion stability. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating a method for controlling the handling of coated glass according to an embodiment of the present invention.

[0052] Figure 2 This is a functional block diagram of a handling control system for coated glass provided in an embodiment of the present invention. Detailed Implementation

[0053] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0054] This application provides a method for controlling the handling of coated glass. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for controlling the handling of coated glass can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks, and big data and artificial intelligence platforms.

[0055] Reference Figure 1 The diagram shown is a flowchart illustrating a method for controlling the handling of coated glass according to an embodiment of the present invention. In this embodiment, the method for controlling the handling of coated glass includes:

[0056] S1: Collect surface reflectance spectrum data, dimensional specifications, and handling arm information of the coated glass.

[0057] In this embodiment of the invention, the acquisition of surface reflectance spectral data of the coated glass, dimensional specifications, and handling arm information of the robotic arm includes:

[0058] The coated glass was scanned across the entire wavelength range to obtain the surface reflectance spectral data of the coated glass;

[0059] Capture the edge contour point cloud of the coated glass to obtain the dimensional specifications of the coated glass;

[0060] The load feedback signal of the robotic arm is collected to determine the handling arm information of the robotic arm.

[0061] An industrial-grade gantry frame is erected above the roller conveyor belt in front of the handling robot's gripping station, integrating a hyperspectral imager covering the visible to near-infrared bands. As the coated glass passes through the scanning area at a constant speed with the conveyor belt, the hyperspectral imager uses high-frequency linear array push-broom technology to capture the subtle spectral changes in light after interference, diffraction, and absorption on the film surface without contacting the glass surface. This is not a simple photograph but a deep analysis of the light wave energy distribution. After receiving the raw light signal, the back-end spectral processing unit uses photoelectric conversion and spectral calibration algorithms to remove ambient light noise and convert the physical light signal into a digital spectral curve matrix.

[0062] The beneficial effect is that it effectively transforms the invisible microstructure characteristics of the film into computer-readable surface reflectance spectral data. This data not only contains colorimetric information about the color, but also key physical properties such as the number and density of the implied film layers and the surface oxidation state, providing a unique optical fingerprint for subsequent evaluation of the wear resistance and pressure resistance of the glass surface.

[0063] Multiple sets of structured light cameras or line laser profilometers are deployed around the gripping point of the robotic arm. When the glass is in place and triggers the photoelectric sensing element, the sensor projects high-density coded stripes or laser lines onto the glass surface and uses the principle of triangulation to capture the precise positioning points of the glass edge and the elevation change data of the surface.

[0064] After filtering, denoising, and stitching massive point cloud data, a three-dimensional digital twin model of the glass entity is constructed. The computing unit can not only accurately calculate the length, width, and thickness of the glass, but also calculate the flatness error and warpage of the glass surface by fitting a plane.

[0065] The beneficial effect is that it outputs dimensional specifications including geometric dimensions and form and position tolerances, ensuring that the robotic arm can adjust the adhesion angle of the suction cup array according to the actual deformation of the glass when planning the gripping posture, thus achieving a leap from two-dimensional planar gripping to three-dimensional flexible adhesion.

[0066] A highly sensitive six-dimensional torque sensor is connected in series between the end effector and the flange of the robotic arm, and current loop monitoring is activated in the servo drivers of each joint. During the trial adsorption or pre-lifting phase, the sensor collects and separates the force components along the three spatial axes and the torque components around each axis in real time. Combining the current joint angles and link length parameters of the robotic arm, the static equilibrium equation is solved using the Jacobian matrix to accurately calculate the offset vector of the glass's center of gravity relative to the gripping center.

[0067] The beneficial effect is that by transforming abstract electrical signals into specific mechanical descriptions, the handling arm information of the robotic arm is determined. This information quantifies the actual torque burden caused by the load on the robotic arm and provides indispensable dynamic boundary conditions for subsequent calculation of dynamic inertia and feedforward torque compensation.

[0068] S2: Extract the film bonding characteristics from the surface reflectance spectral data, and combine them with the size specifications and the handling arm information to calculate the load-bearing capacity, thereby obtaining the theoretical stress extreme value of the coated glass.

[0069] In this embodiment of the invention, the step of extracting the film bonding characteristics from the surface reflectance spectral data and combining them with the dimensional specifications and the handling lever arm information to calculate the load-bearing capacity, thereby obtaining the theoretical stress extreme value of the coated glass, includes:

[0070] The surface reflectance spectral data is subjected to resolution time-frequency extraction to obtain the film bonding characteristics of the surface reflectance spectral data;

[0071] The dimensional specifications and the handling lever information are topologically discretized to obtain the global load distribution matrix of the coated glass;

[0072] The theoretical stress extremum of the coated glass is obtained by performing heterogeneous convolution between the film layer bonding features and the global load distribution matrix.

[0073] Surface reflectance spectral data are retrieved from memory, and high-resolution time-frequency analysis techniques such as wavelet transform or short-time Fourier transform are used to decompose and reconstruct the data at multiple scales. Electromagnetic interference and stray light noise present in the production environment are removed by threshold filtering. Subsequently, the spectral response peaks and bandwidth variations in specific bands are focused to identify characteristic frequency components that are highly correlated with the film lattice structure and interlayer van der Waals forces.

[0074] The beneficial effect is that it aims to extract film bonding characteristics that directly reflect film density and peel resistance from massive spectral information. These characteristics are no longer simple optical values ​​but digital vectors representing the physical strength of the film, providing a core materials science basis for predicting whether film peeling will occur during dynamic handling of glass.

[0075] The acquired dimensional parameters are mapped onto a virtual 3D coordinate system to generate a solid model of the glass. The point of action of the suction cup and the line of action of gravity are determined based on the robotic arm's handling force information. The glass model is then topologically discretized into thousands of tiny mesh elements. The force state of each node under the combined action of gravity, acceleration, inertial force, and suction force is calculated based on the principles of mechanics of materials.

[0076] The beneficial effect is that a global load distribution matrix covering the entire surface of the glass is constructed. This matrix records in detail the stress tensor distribution of each micro-element of the glass from the edge to the center under the current handling posture, thus accurately transforming the macroscopic handling scenario into a microscopic numerical mechanical model.

[0077] The extracted membrane bonding features were used as convolution kernels and subjected to high-dimensional convolution operations with the generated global load distribution matrix. The stress tolerance response of specific membrane microstructures under different stress regions was simulated. Emphasis was placed on high-risk regions of stress concentration in the load matrix, and membrane features were used to determine whether these regions exceeded the material's yield strength or peeling threshold.

[0078] The beneficial effect is that, through this simulation and data deduction, a quantitative theoretical stress extreme value is finally output. This value clearly defines the maximum acceleration limit and torque limit that the robotic arm can apply under the current handling conditions in order to ensure the integrity of the membrane layer, providing an absolute physical constraint boundary for the subsequent generation of safe motion commands.

[0079] S3: Collect the pressure rise rate and steady-state leakage of the adsorption gas path, and perform airtightness characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass.

[0080] In this embodiment of the invention, the collection of the pressure rise rate and steady-state leakage of the adsorption gas path includes:

[0081] Monitor the real-time gas pressure data of the adsorption gas path and perform differential calculations on the real-time gas pressure data to obtain the pressure rise rate of the adsorption gas path.

[0082] The flow rate data of the adsorption gas path when it is in equilibrium is obtained as the steady-state leakage of the adsorption gas path.

[0083] In automated handling of high-end coated glass, the stability of the gas path seal directly determines the safety of the operation. Traditional handling equipment often relies solely on the on / off signal of a vacuum pressure switch to determine whether the adsorption standard has been met; that is, as long as the negative pressure value is below a certain threshold, the robotic arm is allowed to lift. When dealing with special coated glass with microscopic rough textures or porous structures, the static pressure value alone cannot reflect the dynamic holding capacity of the adsorption interface. Once the vacuum generator malfunctions and stops or the robotic arm performs high-dynamic variable-speed movements, even a small leak can quickly cause the vacuum level to drop below the safety line, leading to a fall accident. An industrial-grade high-frequency piezoelectric pressure sensor is integrated into the vacuum manifold or independent gas path branch adjacent to the suction cup array. This sensor has a millisecond-level ultra-fast response capability, enabling it to continuously collect real-time gas pressure data inside the adsorption gas path at a high sampling frequency during the pressure holding phase after the adsorption action is established or during the vacuum failure test phase. After receiving this dense time-series signal, a Kalman filter algorithm is first used to filter out high-frequency noise generated by airflow disturbances. Then, numerical differentiation calculations are performed on the smoothed pressure curve to solve for the slope of the pressure change over time.

[0084] The beneficial effect is that it transforms static pressure readings into dynamic trends, accurately quantifying the rate of vacuum loss per unit time. The final pressure rise rate of the adsorption gas path is not only a key indicator for airtightness testing, but also a physical basis for predicting the remaining safe operating time of the robotic arm in the event of a sudden gas outage, ensuring that the control logic can dynamically adjust the conveying acceleration according to the rate of leakage.

[0085] A thermal mass flow meter is connected in series on the main air supply line between the vacuum generator and the suction cup array. When the suction cups contact the glass and the vacuum is activated, the air pressure in the gas path gradually decreases and eventually stabilizes at a constant value, at which point the gas path enters a state of hydrodynamic equilibrium. The flow monitoring unit reads the instantaneous gas velocity in the pipe, which is numerically exact equal to the rate at which outside air seeps into the gas path through the microscopic gaps at the contact surface between the suction cups and the glass. The processing core captures the flow reading at this moment and defines it as the steady-state leakage of the adsorption gas path.

[0086] The beneficial effect is that it directly reflects the flow resistance characteristics and roughness level of the coated glass surface, which are the core physical variables for subsequent calculation of the interface friction coefficient and anti-slip capability, thus transforming the abstract influence of surface texture into measurable fluid parameters.

[0087] In this embodiment of the invention, the step of performing an inversion of the airtightness characteristics of the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass includes:

[0088] By mapping the pressure rise rate and the steady-state leakage amount to the gap flow resistance, the surface roughness of the adsorption contact surface is obtained.

[0089] Contact mechanics analysis was performed on the coated glass based on the surface roughness to obtain the interfacial adsorption efficiency value of the coated glass.

[0090] During the handling of coated glass, the microscopic morphology of the glass surface is a hidden killer affecting the adsorption quality. Different batches of coating processes can cause minute undulations in the film lattice stacking. These nano- to micro-scale textures create countless invisible micro-leakage channels between the suction cup adhesive and the rigid glass surface. Traditional surface roughness detection requires offline use of a white light interferometer, which is impossible on a continuous production line, preventing the control end from sensing the decrease in gripping force caused by changes in surface texture. This paper retrieves real-time pressure rise rate and steady-state leakage rate as boundary conditions and inputs them into a gap flow resistance model simplified from the Navier-Stokes equations. This model treats the contact gap between the suction cup lip and the glass surface as a porous medium or a parallel plate slit, utilizing the laminar flow characteristics of gas in the slit to solve for the equivalent gap height that hinders gas flow. An iterative algorithm approximates the true flow resistance coefficient, thus mathematically reconstructing the height of the microscopic texture that hinders sealing without physical probes contacting the glass surface.

[0091] The beneficial effect is that the physical inversion process ultimately outputs the surface roughness of the adsorption contact surface with a resolution of micrometers. This parameter is no longer static data in the laboratory, but a dynamic physical index that reflects the ease of gas escape under the current working conditions, providing the most basic geometric basis for subsequent evaluation of friction.

[0092] The calculated surface roughness is used as the input parameter for the rigid rough surface, combined with the elastic modulus and Poisson's ratio of the suction cup rubber material. The deformation process of the rubber material filling the micro-pits on the glass surface under the current vacuum negative pressure is simulated. The ratio of the actual contact area to the nominal contact area between the rubber and the film layer, as well as the meshing depth between the micro-protrusions, are calculated. Based on these micro-mechanical behaviors, the current equivalent friction coefficient and ultimate shear strength are further derived, and their real-time values ​​are compared with the benchmark values ​​under the theoretical perfect adsorption state.

[0093] The beneficial effect is that the mechanical analysis process ultimately generates a normalized interfacial adsorption efficiency value, which intuitively quantifies how tightly the suction cup grips. If this value is lower than the safety threshold, the controller will automatically limit the motion acceleration of the robotic arm, thereby completely avoiding the risk of slippage and detachment caused by differences in film texture at the physical level.

[0094] In this embodiment of the invention, the formula for calculating the interfacial adsorption efficiency value is as follows:

[0095] ;

[0096] in, This is the interfacial adsorption efficiency value. It is the reference static friction coefficient. It refers to the complexity of the surface texture of the coated glass. This is the steady-state leakage amount. It is the equivalent leakage path length. It is the adsorption pressure difference. It is the dynamic attenuation coefficient. It is the rate of pressure rise. It is an exponential function.

[0097] This refers to the interfacial adsorption efficiency value, which is no longer a unit of pressure but a normalized dimensionless confidence index. Physically, it represents the ratio of the effective friction force that the suction cup can provide to the theoretical maximum friction force at the current moment, taking into account all leaks and dynamic disturbances. If this value approaches the reference static friction coefficient, it indicates that the grip is extremely stable; if this value drops significantly, it means that although the vacuum gauge shows normal negative pressure, the interface is actually on the verge of slippage.

[0098] It is the reference static friction coefficient, which represents the physical frictional property of the suction cup rubber material and a specific coated glass surface under ideal, complete contact conditions. It depends on the intermolecular forces between the two materials and is a physical constant determined by a materials science laboratory, setting the theoretical ceiling for gripping ability.

[0099] This refers to the complexity of the surface texture of the coated glass. Coated glass is not absolutely smooth; its surface contains nanoscale lattice stacks or micrometer-scale roller marks. This parameter quantifies the degree to which these micro-textures interfere with the sealing ring's fit. Physically speaking, The higher the value, the rougher the surface, and the easier it is for air molecules to enter the vacuum chamber through the tiny gaps caused by the texture, thus amplifying the negative impact of leakage on adsorption efficiency.

[0100] It is the steady-state leakage rate, which is the volume of air per second that seeps into the suction cup, as measured by the flow meter during the vacuum stabilization phase.

[0101] It is the equivalent leakage path length, a physical measure of the width of the annular band where the suction cup lip contacts the glass, representing the physical distance that air molecules need to travel from the outside to the internal vacuum cavity.

[0102] This is the adsorption pressure difference, which represents the difference between the negative pressure inside the vacuum chamber and the external atmospheric pressure. It is the core driving energy for maintaining the adsorption force. In the formula, it is in the denominator, which physically means that, for the same leakage amount, the larger the adsorption pressure difference, the smaller the destructive effect of air leakage on the overall grasping stability.

[0103] It is the dynamic attenuation coefficient, used for adjustment. Weighting of the impact on the outcome.

[0104] The pressure rise rate is the slope of the vacuum decrease captured by the sensor, representing the rate at which the leak worsens.

[0105] It is an exponential function, reflecting the nonlinear explosive characteristics of dynamic risk. In a physical environment, when the high-speed movement of the robotic arm causes a slight deformation of the suction cup, the rate of pressure rise... It will fluctuate. An exponential relationship means: A tiny increase in this value can lead to an explosive growth in the value. This simulates the avalanche effect, where once the seal begins to loosen rapidly, the adsorption efficiency drops sharply, forcing the robotic arm to slow down in advance in the control logic to avoid a catastrophic detachment.

[0106] The permeation behavior of microfluidics was simulated using the cube root function. It describes the permeation behavior under pressure difference. Driven by this, the ease with which air flows through a steady-state leakage length of is determined. Physically, a leakage penalty factor is constructed, which increases with the severity of the leakage or the smaller the pressure difference, leading to an increase in the denominator and ultimately lowering the overall adsorption efficiency.

[0107] S4: The theoretical stress extremum and the interface adsorption efficiency value are fused by constraint boundary to obtain the double constraint envelope of the coated glass.

[0108] In this embodiment of the invention, the step of fusing the theoretical stress extremum and the interfacial adsorption efficiency value into a constraint boundary to obtain the dual constraint envelope of the coated glass includes:

[0109] By inversely mapping the theoretical stress extremum, the rigid damage boundary of the coated glass is obtained;

[0110] The flexible adsorption boundary of the coated glass was obtained by performing a balance analysis on the interfacial adsorption efficiency value.

[0111] The overlapping safe region of the rigid damage boundary and the flexible adsorption boundary is extracted as the dual constraint envelope of the coated glass.

[0112] The theoretical stress extremum is retrieved, representing the maximum internal force that the glass can withstand without physical damage. Inverse derivation is performed using the finite element stiffness matrix and the Jacobian transpose of the robotic arm. It simulates the robotic arm's grasping motion in different spatial postures, calculating the maximum linear and angular acceleration of the end effector required to generate the ultimate stress at the glass's weakest point, and further decomposing it into the ultimate torque values ​​of the six joint axes. Essentially, it translates the fragility of materials science into a no-go zone for robotics. Through global traversal calculations in Cartesian and joint spaces, the processor ultimately defines an insurmountable rigid damage boundary in the controller's underlying logic.

[0113] The beneficial effect is that this boundary defines the upper limit of dynamics that the robotic arm cannot exceed at any time, thus ensuring at the physical level that the glass will not suffer structural damage due to excessive movement during the entire handling process.

[0114] After establishing the breakage prevention boundary, another critical risk in handling operations is workpiece slippage or detachment. This typically occurs when the robotic arm stops abruptly or makes a high-speed turn, where the resulting centrifugal and inertial forces exceed the frictional holding force of the suction cup. Due to the significant differences in the coefficient of friction of different coated glass surfaces, generic motion parameters often lead to workpieces with low friction coefficients flying off during handling. To mitigate safety hazards, this embodiment introduces a real-time dynamic equilibrium analysis mechanism. The interface adsorption efficiency value is read, which integrates vacuum degree, leakage rate, and surface roughness information, accurately reflecting the current gripping strength. A dynamic equilibrium equation system incorporating gravity, Coriolis force, centrifugal force, and Euler force is constructed to simulate the force conditions of the glass during various high-dynamic maneuvers in three-dimensional space. The focus is on solving for the maximum combined external force that can be applied at the end of the robotic arm under the critical state of no tangential slippage or normal separation.

[0115] The beneficial effect is that it defines a flexible adsorption boundary based on aerodynamics for the robotic arm. This boundary dynamically limits the maximum acceleration of the robotic arm in all directions, ensuring that no matter how complex the motion trajectory is, the inertial peeling force acting on the glass is always less than the physical adsorption force provided by the suction cup, thereby completely eliminating the occurrence of part dropping accidents.

[0116] In actual production, the two boundaries mentioned above often conflict: to prevent adsorption slippage, deceleration may need to be limited, but this could lead to longer handling cycles; while to pursue efficiency or meet certain path points, higher acceleration may be needed, but this could trigger stress limits. If the control logic cannot coordinate these two aspects, the robotic arm will be unable to operate normally due to frequent alarms and shutdowns. The path planning controller treats the rigid damage boundary and the flexible adsorption boundary as two independent hypersurfaces in the six-dimensional state space. It performs a Boolean AND operation, that is, within each interpolation cycle in the time domain, it compares the limit values ​​of the two boundaries point by point and always selects the more conservative one as the current effective constraint.

[0117] For example: In a certain pose, the rigid boundary allows 5 The acceleration, but the flexible boundary only allows 3 Then the maximum forced lock is 3. At another steady-state location, the extremely strong adsorption force allows for 10... However, the stress is limited to 6. Then it is locked as 6 The fusion operation ultimately eliminated all potentially dangerous areas, extracting an absolutely safe double-constraint envelope.

[0118] The beneficial effect is that the dual constraint envelope provides a unique green channel for subsequent trajectory planning, ensuring that every generated motion command simultaneously meets the dual safety standards of not breaking and not falling.

[0119] S5: Based on the dual constraint envelope, perform time-series deduction on the robotic arm to obtain the spatial motion time sequence of the robotic arm.

[0120] In this embodiment of the invention, the step of performing time-series deduction on the robotic arm based on the dual constraint envelope to obtain the spatial motion time sequence of the robotic arm includes:

[0121] Based on the dual constraint envelope, the transport path of the robotic arm is subjected to full path constraint mapping to obtain the local acceleration threshold sequence of the robotic arm;

[0122] The target velocity profile curve of the robotic arm is obtained by smoothly fitting the local acceleration threshold sequence.

[0123] The spatial motion timing of the robotic arm is obtained by performing periodic discrete interpolation on the target velocity profile curve.

[0124] Since the handling path typically involves multiple stages such as vertical lifting, horizontal transfer, and lowering, the posture and equivalent moment of inertia of the robotic arm at different path points are constantly changing. Furthermore, the gravitational torque and wind resistance experienced by the glass fluctuate dramatically depending on its location. Simply using a single global acceleration limit often leads to inefficiency in low-risk sections and breaches of safety limits in high-risk sections. A high-precision path mapping mechanism is employed, extracting the planned spatial geometric path from memory and discretizing it into thousands of tightly packed path control points. The generated dual constraint envelope, including rigid damage boundaries and flexible adsorption boundaries, is invoked to perform dynamic verification on each discrete point. The calculation module analyzes, one by one, the maximum instantaneous acceleration and maximum jerk allowed at the current robotic arm configuration to prevent membrane stress damage and adsorption slippage. In effect, a safety ruler is tailored for each step on the path, arranging these discrete constraint values ​​in path order and outputting a sequence of local acceleration thresholds that vary with position.

[0125] The beneficial effect is that the local acceleration threshold sequence indicates which part of the path can be sprinted at full speed and which part must be traversed with caution, thereby achieving a full-coverage mapping of physical constraints on the spatial trajectory.

[0126] The local acceleration threshold sequence was retrieved, and a seven-segment velocity planning algorithm was used to mathematically smooth the original sequence. While keeping the overall curve below a safe threshold, peaks and valleys were actively smoothed to eliminate abrupt acceleration changes, and the derivative of acceleration with respect to time was strictly limited. The mathematical optimization process aimed to find the physically smoothest velocity variation law, ensuring that the robotic arm transitions smoothly like a fluid when performing actions. After tens of thousands of iterations, a continuous, smooth, and everywhere differentiable target velocity profile curve was finally constructed.

[0127] The beneficial effect is that the target velocity profile curve not only defines the ideal velocity at each moment, but also fundamentally suppresses residual vibrations that may cause coating damage, laying a kinematic foundation for high-speed and high-precision material handling.

[0128] A high-real-time discrete interpolation strategy is employed. A high-precision clock generator within the motion control chip triggers an interrupt signal. At the arrival of each servo cycle, the interpolation algorithm extracts the corresponding target position, velocity, and feedforward acceleration values ​​from the target velocity profile curve based on the current timestamp. An inverse kinematics solver maps these Cartesian space target values ​​to the six joint spaces, calculating the minute angular increment each joint motor should rotate in the next millisecond. Macroscopic trajectory planning is decomposed into microscopic motor execution cycles, ensuring microsecond-level synchronization of multi-axis linkage. The control unit continuously sends data packets containing position, velocity, and torque information to each axis driver, forming a strictly time-ordered spatial motion sequence.

[0129] The beneficial effect is that the spatial motion timing precisely directs every joint of the robotic arm, achieving a perfect reproduction from theoretical planning to physical action.

[0130] S6: Generate an adaptive inertia transport command for the coated glass based on the spatial motion timing.

[0131] In this embodiment of the invention, generating the adaptive inertia transport command for the coated glass based on the spatial motion time sequence includes:

[0132] The instantaneous pose data of the spatial motion sequence is analyzed, and the rotational inertia is calculated in combination with the size specification parameters to obtain the dynamic load inertia of the robotic arm.

[0133] The feedforward current compensation value of the robotic arm is generated by the dynamic load inertia.

[0134] Based on the feedforward current compensation value, the basic servo commands of the spatial motion timing are modified by torque superposition to obtain the adaptive inertia handling command of the coated glass.

[0135] In the automated handling of high-end coated glass, the greatest dynamic challenge faced by robotic arms lies in the dramatic time-varying nature of load inertia. With the extension and retraction of the robotic arm joints and the variation in the size of the glass adsorbed at the end effector, the rotational inertia matrix of the entire motion system undergoes significant nonlinear changes. If the controller calculates based solely on fixed load parameters, when the robotic arm switches rapidly from a retracted to an extended posture, the torque output by the motor is often insufficient to overcome the actual inertia, leading to increased trajectory tracking errors and even end-effector jitter. This poses a significant safety hazard to the fragile coating layer. This paper extracts the instantaneous joint angle and angular velocity information for each interpolation cycle from the spatial motion time-series data stream, while simultaneously invoking the glass size specifications. Using the Newton-Euler iterative equations or the Lagrange dynamics equations, a multi-rigid-body dynamic model is constructed, incorporating the link mass, motor rotor inertia, and end-effector glass load. This model can calculate in real time the equivalent rotational inertia exhibited by the robotic arm in maintaining its predetermined motion state at the current millisecond.

[0136] The beneficial effect is that the mathematical and physical calculation process ultimately outputs a precise dynamic load inertia. This parameter quantifies the weight of the robotic arm at every point in space in real time, providing the most basic physical basis for the subsequent precise distribution of motor torque.

[0137] Traditional PID feedback control relies on the accumulation of position errors to generate corrective torque. This mechanism inherently suffers from phase lag when handling the high-speed start-up and shutdown of large-inertia coated glass, manifesting as sluggish response during startup and overshoot oscillation during shutdown. This new approach receives the calculated dynamic load inertia and combines it with the planned target angular acceleration value in the spatial motion sequence. Utilizing the physical law that torque equals moment of inertia multiplied by angular acceleration, it directly calculates the theoretical torque required to overcome the current inertial load. Based on the motor's torque constant and flux characteristics, this theoretical torque is converted into a corresponding current amplitude. In effect, the required force is calculated before the motor generates position errors, generating millisecond-level varying feedforward current compensation values.

[0138] The beneficial effect is that the feedforward current compensation value represents the additional current energy that must be injected to enable the robotic arm to move along a predetermined trajectory with zero delay, which fundamentally improves the dynamic response stiffness of the servo.

[0139] Within a microsecond-level time window, these two current commands are vector-superimposed to generate a pulse-width modulation signal that ultimately drives the power module to switch on and off. This superposition operation allows the motor to obtain a powerful current sufficient to overcome enormous inertia the instant it receives the motion command, while automatically reducing the current output during the constant-speed phase to save energy. Through this dynamic torque correction, the robotic arm behaves as if it has muscle memory, automatically adjusting the amount of force applied based on the size of the glass being grasped and its current posture.

[0140] The beneficial effect is that the signal processing process ultimately outputs an adaptive inertia handling command, which drives the robotic arm to complete the entire process from static acceleration to high-speed operation and then to precise positioning with high stability, completely eliminating the risk of swaying and slippage of the coated glass caused by changes in inertia.

[0141] like Figure 2 The diagram shown is a functional block diagram of a coated glass handling control system provided in an embodiment of the present invention.

[0142] The coated glass handling control system described in this invention can be installed in an electronic device. Depending on the functions implemented, the coated glass handling control system may include a data acquisition module, a load-bearing capacity calculation module, an airtightness characteristic inversion module, a constraint boundary fusion module, a timing deduction module, and an instruction generation module. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, and are stored in the memory of the electronic device.

[0143] In this embodiment, the functions of each module / unit are as follows:

[0144] The data acquisition module is used to collect surface reflectance spectrum data, size specifications, and handling arm information of the coated glass.

[0145] The load-bearing capacity calculation module is used to extract the film bonding characteristics of the surface reflectance spectrum data, and combine the size specification parameters and the handling force arm information to calculate the load-bearing capacity and obtain the theoretical stress extreme value of the coated glass.

[0146] The airtightness characteristic inversion module is used to collect the pressure rise rate and steady-state leakage of the adsorption gas path, and perform airtightness characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage to obtain the interfacial adsorption efficiency value of the coated glass.

[0147] The constraint boundary fusion module is used to fuse the theoretical stress extreme value and the interface adsorption efficiency value into a constraint boundary to obtain the dual constraint envelope of the coated glass.

[0148] The timing simulation module is used to perform timing simulation on the robotic arm based on the dual constraint envelope to obtain the spatial motion timing of the robotic arm.

[0149] The instruction generation module is used to generate adaptive inertia transport instructions for the coated glass based on the spatial motion timing.

[0150] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0151] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0152] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0153] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0154] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A handling control method of coated glass, characterized by, The method comprises: S1: respectively collecting surface reflection spectrum data, size specification parameters and carrying force arm information of the coated glass; S2: extracting film layer combination characteristics of the surface reflection spectrum data, and combining the size specification parameters and the carrying force arm information to calculate the bearing force, to obtain a theoretical stress extreme value of the coated glass; S3: collecting pressure rise rate and steady-state leakage of the adsorption gas path, and based on the pressure rise rate and the steady-state leakage, performing air tightness characteristic inversion on the coated glass to obtain an interface adsorption efficiency value of the coated glass; S4: performing constraint boundary fusion on the theoretical stress extreme value and the interface adsorption efficiency value to obtain a double constraint envelope of the coated glass; S5: based on the double constraint envelope, performing time sequence deduction on the mechanical arm to obtain a spatial motion time sequence of the mechanical arm; S6: based on the spatial motion time sequence, generating an adaptive inertia carrying instruction of the coated glass.

2. The handling control method of a coated glass according to claim 1, wherein, The method comprises: performing full-band scanning on the coated glass to obtain the surface reflection spectrum data of the coated glass; capturing edge contour point cloud of the coated glass to obtain the size specification parameters of the coated glass; collecting load feedback signal of the mechanical arm to determine the carrying force arm information of the mechanical arm.

3. The handling control method of a coated glass according to claim 1, wherein The method comprises: performing resolution time-frequency extraction on the surface reflection spectrum data to obtain the film layer combination characteristics of the surface reflection spectrum data; performing topological discretization on the size specification parameters and the carrying force arm information to obtain a global load distribution matrix of the coated glass; performing heterogeneous convolution on the film layer combination characteristics and the global load distribution matrix to obtain the theoretical stress extreme value of the coated glass.

4. The handling control method of a coated glass according to claim 1, wherein The method comprises: monitoring real-time air pressure data of the adsorption gas path, and performing differential calculation on the real-time air pressure data to obtain the pressure rise rate of the adsorption gas path; obtaining flow data of the adsorption gas path in the equilibrium state as the steady-state leakage of the adsorption gas path.

5. The method of claim 1, wherein the glass sheet is a coated glass sheet. The method comprises: performing gap flow resistance mapping on the pressure rise rate and the steady-state leakage to obtain surface roughness of the adsorption contact surface; based on the surface roughness, performing contact mechanics analysis on the coated glass to obtain the interface adsorption efficiency value of the coated glass.

6. The handling control method of a coated glass according to claim 5, wherein The calculation formula of the interface adsorption efficiency value is: ； wherein, is the interfacial adsorption effectiveness value, is the reference static friction coefficient, is the surface texture complexity of the coated glass, is the steady state leak rate, is the equivalent leak path length, is the adsorption pressure differential, is the dynamic decay coefficient, is the pressure rise rate, is an exponential function.

7. The method of claim 1, wherein the glass sheet is a coated glass sheet. The method comprises: performing reverse mapping on the theoretical stress extreme value to obtain a rigid damage boundary of the coated glass; Performing balance analysis on the interface adsorption performance value to obtain a flexible adsorption boundary of the coated glass; Extracting an overlapping safety area of the rigid damage boundary and the flexible adsorption boundary as a double constraint envelope of the coated glass.

8. The method of claim 1, wherein the glass sheet is a coated glass sheet. The double constraint envelope is used to perform timing deduction on the mechanical arm to obtain a spatial motion timing of the mechanical arm, including: The double constraint envelope is used to perform full-path constraint mapping on a carrying path of the mechanical arm to obtain a local acceleration threshold sequence of the mechanical arm; The local acceleration threshold sequence is subjected to smooth fitting to obtain a target speed profile curve of the mechanical arm; The target speed profile curve is subjected to equal-period discrete interpolation to obtain the spatial motion timing of the mechanical arm.

9. The method of claim 1, wherein the glass sheet is a coated glass sheet. The spatial motion timing is used to generate an adaptive inertia carrying instruction of the coated glass, including: Instantaneous pose data of the spatial motion timing is analyzed, and the size specification parameters are combined to perform rotational inertia calculation to obtain a dynamic load inertia of the mechanical arm; The dynamic load inertia is used to generate a feedforward current compensation value of the mechanical arm; The feedforward current compensation value is used to perform torque superposition correction on a basic servo instruction of the spatial motion timing to obtain the adaptive inertia carrying instruction of the coated glass.

10. A handling control system for coated glass, characterized by, A carrying control method of a coated glass is used to implement any one of claims 1-9, and the system includes: A data acquisition module is used to acquire surface reflectance spectrum data, size specification parameters, and carrying arm information of a mechanical arm of the coated glass respectively; A bearing force calculation module is used to extract film layer combination features of the surface reflectance spectrum data, and combine the size specification parameters and the carrying arm information to perform bearing force calculation to obtain a theoretical stress extreme value of the coated glass; An airtightness characteristic inversion module is used to acquire a pressure rise rate and a steady-state leakage of an adsorption gas path, and perform airtightness characteristic inversion on the coated glass based on the pressure rise rate and the steady-state leakage to obtain an interface adsorption performance value of the coated glass; A constraint boundary fusion module is used to perform constraint boundary fusion on the theoretical stress extreme value and the interface adsorption performance value to obtain a double constraint envelope of the coated glass; A timing deduction module is used to perform timing deduction on the mechanical arm based on the double constraint envelope to obtain a spatial motion timing of the mechanical arm; An instruction generation module is used to generate an adaptive inertia carrying instruction of the coated glass based on the spatial motion timing.

Images