Touch display module assembly coordination system and method
By synchronously collecting data on multiple physical quantities using multiple sensors, generating a deviation propagation matrix, and performing virtual verification, the problem of isolated states of each process in the touch display module assembly system is solved, thereby improving assembly accuracy and reliability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN TAIQI PHOTOELECTRIC CO LTD
- Filing Date
- 2026-03-05
- Publication Date
- 2026-06-05
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In existing touch display module assembly systems, the states of each process cannot be correlated with each other, lacking overall optimization. This results in insufficient assembly accuracy and reliability, inadequate equipment self-adjustment capabilities, limited compensation methods, and an inability to achieve rapid linkage and overall optimization.
By synchronously collecting data on multiple physical quantities through multiple sensors, a unified time-series process status data is formed. This data is then written to a collaborative platform in real time for difference calculation and consistency checks, generating a deviation propagation matrix, constructing compensation instructions and performing virtual verification, and dynamically updating parameters to achieve collaborative optimization between processes.
It has enabled precise control and quality improvement in the assembly of touch display modules, improved assembly accuracy and reliability, and optimized production efficiency.
Smart Images
Figure CN122152159A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of collaborative assembly of touch display modules, specifically to a collaborative assembly system and method for touch display modules. Background Technology
[0002] The assembly process of touch display modules involves precise operations with multiple steps and parameters, including real-time monitoring of key parameters such as bonding displacement, pressure control, dispensing trajectory, curing temperature distribution, and surface stress. Although existing assembly systems can collect partial data from a single step, they still have the following main shortcomings: Isolated process information: The status of each process cannot be seen by each other, and there is a lack of overall correlation analysis, making it difficult to quantify and transmit deviations between processes in a timely manner; Insufficient self-adjustment capability of equipment: Existing systems rely heavily on manual experience or static parameter settings, and the equipment in each process cannot adjust itself based on real-time data from other processes, making it easy for deviations to accumulate during production. Limitations of compensation methods: Existing compensation methods mostly target a single parameter or a single process for local correction, lacking a comprehensive control strategy that couples multiple physical quantities, resulting in delayed abnormal response and inability to achieve rapid linkage and overall optimization; Therefore, it is essential to design a collaborative system and method for assembling touch display modules to improve assembly accuracy and reliability. Summary of the Invention
[0003] To address the shortcomings of existing technologies, this invention provides a collaborative system and method for assembling touch display modules, which has the advantages of improving assembly accuracy and reliability, and solves the problems mentioned in the background technology.
[0004] To achieve the aforementioned goals of improving assembly accuracy and reliability, this invention provides the following technical solution: a collaborative assembly method for touch display modules, comprising the following steps: Acquire bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution and surface stress, and perform time serialization and encoding processing according to a unified data structure to form process status data containing multiple physical quantities and visual anomaly markers. Process status data is written to the collaborative platform in real time. The difference between parameters of adjacent processes is calculated one by one according to the process flow sequence. The direction consistency check and noise removal are performed to generate a difference dataset that reflects the quantitative difference relationship between processes. Based on the differential dataset, displacement, pressure, trajectory and temperature deviations are extracted, the inter-process propagation coefficient is calculated, the deviation propagation matrix is constructed and compensation instructions for pressure, speed, trajectory and exposure area are generated; The compensation command is sent into the virtual compensation process to verify the superposition effect. After the verification is successful, it is written to the corresponding device. When an abnormal trend occurs, the hierarchical linkage is triggered according to the deviation propagation matrix to generate a response sequence for group management by workstation. The final test data, along with the process status data and response sequence, are transmitted back to the collaborative platform to analyze the contribution of each workstation's actions to the deviation, dynamically update the long-term parameters of each process, and perform upper and lower limit constraints, baseline reset, and drift calibration on the parameters.
[0005] Preferably, the process for generating process state data that includes multiple physical quantities and visual anomaly markers is as follows: Key parameters of each process are collected synchronously by multiple sensors, including displacement sensors, pressure sensors, dispensing trajectory tracking systems, thermal imaging or temperature sensors, and surface stress detection devices. Different physical quantities are mapped to a unified time axis and standardized dimensions to form a multidimensional state vector; Simultaneously, visual inspection markers are used to label abnormal features, generating process status data that can be parsed and analyzed across processes on a collaborative platform.
[0006] Preferably, the process of calculating the difference between parameters of adjacent processes according to the process flow sequence is as follows: Based on process state data, multidimensional state vectors are read in the collaborative platform and standardized with unified dimensions. The difference between each physical quantity data in the continuous process is calculated, and the timestamp, spatial location, and corresponding process, equipment and physical quantity identifier of each deviation are recorded. A weighted algorithm is used to perform preliminary quantification of the deviations between different processes, forming a preliminary deviation matrix between processes.
[0007] Preferably, the process of generating a difference dataset reflecting the quantitative differences between processes is as follows: Perform directional consistency analysis on the preliminary inter-process deviation matrix; Anomalies are identified by changes in vector direction and amplitude, and data that does not conform to the overall movement trend or process logic is eliminated. The processed deviations are used to generate a multidimensional difference dataset according to the process sequence and physical quantity type.
[0008] Preferably, the process of constructing the deviation propagation matrix and generating compensation instructions for pressure, speed, trajectory, and exposure area is as follows: Read the multidimensional deviation information of each process in the difference dataset, and perform statistical analysis and weighting on the cumulative effect of deviation in time and space; A multi-physical quantity coupling model is established by combining historical process data, the propagation coefficient of each deviation to subsequent processes is calculated, and a complete deviation propagation matrix is constructed. Multi-physical quantity comprehensive compensation commands can be generated based on the deviation propagation matrix and directly sent to the device.
[0009] Preferably, the process of sending the compensation command into the virtual compensation process to verify the superposition effect is as follows: In a digital twin platform or simulation environment, the deviation propagation matrix and the multi-physical quantity comprehensive compensation command are superimposed and simulated. The effectiveness of the compensation strategy was verified by combining deviation propagation and the coupling effect of multiple physical quantities. Real-time adjustments are made to address inter-process conflicts, overcompensation, or undercompensation, optimizing the order and magnitude of instructions.
[0010] Preferably, the process of generating a response sequence for workstation-based group management is as follows: The collaborative platform collects the trend of parameters of each process in real time, and combines the compensation instructions that have been virtually verified and optimized with the deviation propagation matrix. Reverse tracing of abnormal deviations to identify the main contributing workstations and error coupling relationships; By combining historical deviation data, inter-process dependencies, and physical quantity coupling models, the magnitude, sequence of action, and timing of execution of response instructions are dynamically optimized and adjusted. Generate a multi-workstation collaborative response sequence according to preset priorities, workstation grouping, and linkage strategies.
[0011] Preferably, the process of analyzing the contribution of each workstation's actions to the deviation is as follows: Match the multi-station collaborative response sequence with the final detection data; By using deviation decomposition, statistical regression and multi-physical quantity coupling analysis methods, the deviation contribution value of each workstation in different time periods and different physical quantities is quantified. By combining historical process data and response execution, the contribution values are weighted, corrected, and cumulatively analyzed to generate a comprehensive contribution value of each workstation to the overall error.
[0012] Preferably, the process of setting upper and lower limits for parameters, resetting the baseline, and calibrating for drift is as follows: Based on the contribution of each workstation to the overall deviation, combined with historical process data, equipment performance characteristics and inter-process dependencies, the long-term control parameter change trend and cumulative deviation of each process are dynamically analyzed. The parameters are corrected in real time according to the upper and lower limits, and key links in the accumulation of deviations are identified and corrected accordingly. Simultaneously, the baseline is reset, and drift calibration techniques are used to gradually adjust for minor offsets.
[0013] A touch display module assembly collaborative system, comprising: State generation module: acquires bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution and surface stress, and encodes them according to a unified structure to form process state data with multiple physical quantities and visual anomaly markers; The difference calculation module writes process status data into the collaborative platform, calculates the difference item by item according to the process flow sequence, checks the consistency of direction and removes noise, and generates a quantitative difference dataset between processes. Deviation Analysis Module: Extracts displacement, pressure, trajectory, and temperature deviations based on the difference dataset, calculates the propagation coefficient, constructs the deviation propagation matrix, and generates compensation instructions; Command execution module: Sends compensation commands into the virtual compensation process to verify the superposition effect. After verification, it sends them to the equipment. When the trend is abnormal, it triggers hierarchical linkage to generate workstation response sequence according to the deviation propagation matrix. Parameter update module: The final detection data, process status data and response sequence are transmitted back to the platform to analyze the contribution of station actions to deviation, dynamically update the parameters of each process and perform constraint calibration.
[0014] Compared with the prior art, the present invention provides a touch display module assembly collaboration system and method, which has the following beneficial effects: This invention generates process status data by uniformly collecting and time-series encoding multiple physical quantities such as bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution, and surface stress. This data is then written into a collaborative platform in real time. The platform performs difference calculations, directional consistency checks, and noise removal on parameters of adjacent processes to form a quantitative difference dataset between processes. Based on this dataset, the deviation is extracted and a deviation propagation matrix is constructed. Compensation instructions for pressure, speed, trajectory, and exposure area are generated. Through virtual compensation verification and hierarchical linkage execution response, the detection data and response sequence are finally transmitted back, and the long-term parameters of each process are dynamically updated. This achieves precise control of assembly deviations, collaborative optimization between processes, and a significant improvement in assembly quality and efficiency. Attached Figure Description
[0015] Figure 1 This is a schematic diagram of the method of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation
[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0017] Example 1: Please refer to Figure 1 As shown in the figure, a touch display module assembly collaboration method according to an embodiment of the present invention includes the following steps: S1: Acquire bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution, and surface stress. Perform time serialization and encoding processing according to a unified data structure to form process status data containing multiple physical quantities and visual anomaly markers.
[0018] The process of generating process state data containing multiple physical quantities and visual anomaly markers in S1 is as follows: Key parameters of each process are collected synchronously by multiple sensors, including displacement sensors, pressure sensors, dispensing trajectory tracking systems, thermal imaging or temperature sensors, and surface stress detection devices. Displacement sensors, pressure sensors, dispensing trajectory tracking systems, thermal imaging or temperature sensors, and surface stress detection devices are installed at corresponding positions in each process. Each sensor is connected to the data acquisition and control system via a wired or wireless interface. The sampling frequency is uniformly set, and timestamps and process number markers are added to the acquired signals. After the acquisition program is started, the sensors continuously record parameters according to the set frequency, generating raw signal data streams. Each data entry includes sensor type, process identifier, equipment number, acquisition timestamp, and measured value.
[0019] Different physical quantities are mapped to a unified time axis and standardized dimensions to form a multidimensional state vector; The raw signals from each sensor are aligned according to the acquisition timestamp. Linear interpolation or spline interpolation is used to upsample the low sampling frequency signals so that the data of different physical quantities have corresponding values at a unified time point. The values of each physical quantity are converted according to the pre-set dimensional standardization rules, such as converting displacement to millimeters, pressure to kilopascals, temperature to degrees Celsius, and surface stress to megapascals. The standardized values of each physical quantity at each time point are combined in matrix form to form a multidimensional state vector, with each state vector corresponding to a time point.
[0020] Simultaneously, visual inspection markers are used to label abnormal features, generating process status data that can be parsed and analyzed across processes on a collaborative platform; The process surface is photographed using a fixed or moving camera to collect a sequence of visual images for each process. The images are matched with multidimensional state vectors according to the acquisition timestamp. Computer vision algorithms are used to locate surface defects, offsets, bubbles, or stains in the images. Detected anomalies are generated into binary labels or multi-level anomaly label matrices. Each visual label is accompanied by the acquisition time, process number, and image number to form visual anomaly information that can be associated with the corresponding state vector. The time-aligned multidimensional state vector is combined with the corresponding visual anomaly label to form the final process state data record.
[0021] S2: Write process status data to the collaborative platform in real time, calculate the difference of parameters of adjacent processes one by one according to the process flow order, and perform directional consistency checks and noise removal to generate a difference dataset that reflects the quantitative difference relationship between processes.
[0022] In S2, the process of calculating the difference between parameters of adjacent processes according to the process flow sequence is as follows: Based on process state data, multidimensional state vectors are read in the collaborative platform and standardized with unified dimensions. In the collaborative platform, the multidimensional state vectors of each time point are read from the stored process state data table or serialized file in the order of the operation. The physical quantity values in each state vector are uniformly transformed, such as displacement values being uniformly transformed to millimeters, pressure values to kilopascals, and temperature values to degrees Celsius. The numerical values after the transformation can be normalized according to the minimum value minus the maximum value or the standard deviation to generate a multidimensional state vector matrix with a uniform scale.
[0023] The difference between each physical quantity data in the continuous process is calculated, and the timestamp, spatial location, and corresponding process, equipment and physical quantity identifier of each deviation are recorded. Take the state vector of process n and the state vector of process n+1, perform a simple subtraction operation on each physical quantity to obtain the numerical deviation of the physical quantity between the two processes, and record the corresponding timestamp, spatial location, process number, equipment number and physical quantity identifier of each deviation. Repeat this operation for the entire process sequence to generate a complete set of inter-process deviations.
[0024] A weighted algorithm is used to perform preliminary quantification of the deviations between different processes, forming a preliminary inter-process deviation matrix; The generated set of inter-process deviations is weighted according to preset weights for the deviations of different physical quantities. The weights can be set based on process type, importance of physical quantity, or historical data statistics. The weighted deviations are arranged in the order of processes to form a preliminary inter-process deviation matrix.
[0025] The process of generating a difference dataset in S2 that reflects the quantitative differences between processes is as follows: Perform directional consistency analysis on the preliminary inter-process deviation matrix; Extract the deviation vector between each process from the preliminary process deviation matrix, calculate the direction angle or direction ratio for each physical component of each vector, compare the vector directions of adjacent processes, and record the physical quantity and process number whose direction deviation exceeds the preset threshold.
[0026] Anomalies are identified by changes in vector direction and amplitude, and data that does not conform to the overall movement trend or process logic is eliminated. Calculate the rate of change of the deviation amplitude and the deviation amplitude of the preceding and following processes, determine whether the rate of change exceeds the threshold, mark the deviation points that exceed the threshold and whose direction is inconsistent with the overall trend as outliers, remove all outliers from the deviation set, and retain their timestamp, spatial location and physical quantity identifier for recording.
[0027] The processed deviations are used to generate a multidimensional difference dataset according to the process sequence and physical quantity type. The deviation data after removing anomalies is arranged in the order of the processes. For each process, the deviation of each physical quantity is combined into a multi-dimensional vector as an independent dimension. Each vector contains the process number, equipment number, timestamp, physical quantity identifier and the processed deviation value. All process vectors are arranged in order to generate a complete multi-dimensional difference dataset.
[0028] S3: Extract displacement, pressure, trajectory, and temperature deviations based on the differential dataset, calculate the inter-process propagation coefficient, construct the deviation propagation matrix, and generate compensation instructions for pressure, speed, trajectory, and exposure area.
[0029] The process of constructing the deviation propagation matrix and generating compensation instructions for pressure, speed, trajectory, and exposure area in S3 is as follows: Read the multidimensional deviation information of each process in the difference dataset, and perform statistical analysis and weighting on the cumulative effect of deviation in time and space; Deviation vectors for each process are read sequentially from the multidimensional difference dataset. Each vector contains process number, physical quantity type, timestamp, spatial location, and deviation value. Deviations of the same physical quantity in consecutive processes are accumulated over time. Deviation values are summed in process order, and the location distribution of accumulated deviations at each time point is recorded. The mean, variance, and moving average of accumulated deviations are calculated to generate a deviation distribution matrix for each process in time and space. Deviations of different physical quantities are weighted using preset weights, and the deviation magnitude and weights are combined to form a weighted deviation matrix.
[0030] A multi-physical quantity coupling model is established by combining historical process data, the propagation coefficient of each deviation to subsequent processes is calculated, and a complete deviation propagation matrix is constructed. For each weighted deviation, identify its influence path on the same physical quantity in subsequent processes. Represent the influence of deviations in a matrix according to the process sequence. Each element represents the cumulative contribution of the source process deviation to the target process. For each influence path, calculate the propagation coefficient in combination with historical process data. Fill the propagation coefficients of the source process, the target process, and the corresponding physical quantity into the deviation propagation matrix.
[0031] Based on the deviation propagation matrix, multi-physical quantity comprehensive compensation commands can be directly issued to the device; Based on the deviation propagation matrix, the cumulative effect of deviation in each target process is mapped to a compensation amount. For the physical quantity of each target process, the deviations of the source process are linearly superimposed according to the propagation coefficient to obtain a comprehensive deviation value. The comprehensive deviation value is converted into an operation command that the equipment can recognize, including pressure adjustment value, speed adjustment amount, dispensing trajectory correction offset and exposure area correction parameters. The compensation commands of all target processes are summarized in the process sequence to form a complete multi-physical quantity comprehensive compensation command.
[0032] S4: Send the compensation command into the virtual compensation process to verify the superposition effect. After the verification is successful, write it to the corresponding device. When an abnormal trend occurs, trigger hierarchical linkage according to the deviation propagation matrix to generate a response sequence for group management by workstation.
[0033] In S4, the process of sending the compensation command into the virtual compensation process to verify the superposition effect is as follows: In a digital twin platform or simulation environment, the deviation propagation matrix and the multi-physical quantity comprehensive compensation command are superimposed and simulated. In a digital twin platform or simulation environment, the deviation propagation matrix is first loaded into the calculation module. The inter-process deviation values corresponding to the matrix elements are mapped to the virtual units of each process. Multi-physical quantity comprehensive compensation instructions are applied to the virtual units one by one. For each instruction, the target process, physical quantity type and compensation value are read. In the simulation environment, the process model is operated to offset or the parameters are adjusted, including pressure, speed, dispensing trajectory position and exposure area settings. The instructions of all processes are superimposed and calculated in the process order, and the physical quantity change values of each time step and spatial position are recorded.
[0034] The effectiveness of the compensation strategy was verified by combining deviation propagation and the coupling effect of multiple physical quantities. During the superposition simulation, the changes of each physical quantity are multiplied with the corresponding propagation coefficients in the deviation propagation matrix to obtain the comprehensive deviation of each process at each time step. The possible coupling relationships between different physical quantities are calculated by matrix multiplication or numerical summation, the state vector of each process unit is updated, and the values of physical quantities such as displacement, pressure, trajectory and temperature after superposition are recorded to form a virtual execution trajectory data table.
[0035] Real-time adjustments are made to address inter-process conflicts, overcompensation, or undercompensation, optimizing the instruction sequence and range. The superimposed simulation results are checked step by step to determine whether there are numerical conflicts or deviation superposition anomalies. Entries in adjacent processes whose physical quantities exceed the set threshold or have symbol conflicts are marked. For the marked processes, the corresponding compensation instructions are scaled, shifted, or reordered. The adjusted instruction parameters are recorded. The adjusted compensation instructions are reloaded into the virtual unit for iterative simulation. Deviation superposition and coupling effects are repeatedly calculated until all process instructions have completed a standardized superposition process. Finally, the compensation instruction sequence for each process and the corresponding virtual execution trajectory data table are output.
[0036] The process of generating a response sequence for workstation-based group management in S4 is as follows: The collaborative platform collects the trend of parameters of each process in real time, and combines the compensation instructions that have been virtually verified and optimized with the deviation propagation matrix. Key parameters for each process are collected in real time through the collaborative platform interface, including displacement sensors, pressure sensors, dispensing trajectory tracking system, temperature sensors, and exposure area monitoring. Sensor outputs are read in each sampling cycle, and compensation instructions verified by virtual compensation are loaded into the collaborative platform and associated with the deviation propagation matrix of the corresponding process. Each compensation instruction is labeled with the workstation, physical quantity type, compensation range, and expected execution time. The relevant rows and columns of the deviation propagation matrix are read, and the inter-process propagation coefficient is multiplied by the compensation range to generate the adjusted compensation value. The adjustment results are stored in a structured format, including workstation number, physical quantity, compensation value, and so on.
[0037] Reverse tracing of abnormal deviations to identify the main contributing workstations and error coupling relationships; Perform process-by-process calculations on the real-time collected trend data and deviation propagation matrix, locate the deviation abrupt change or continuous deviation segment in the trend curve, and record the time and workstation position. Use the matrix reverse calculation method to propagate the abnormal deviation to the preceding workstation, mark the workstation that contributes the most to the abnormality, and analyze the coupling relationship between processes. Add the key workstations that affect multiple physical quantities to the key workstation list.
[0038] By combining historical deviation data, inter-process dependencies, and physical quantity coupling models, the magnitude, sequence of action, and timing of execution of response instructions are dynamically optimized and adjusted. Based on historical deviation data and process dependencies, the compensation magnitude is adjusted using linear or nonlinear calculation methods. The execution order of instructions is sorted, prioritizing instructions that have the greatest impact on subsequent processes. At the same time, the execution time interval is recorded. For workstations with multiple physical quantities coupled, the compensation values of each physical quantity are combined and calculated to generate a comprehensive instruction for that workstation.
[0039] Generate a multi-workstation collaborative response sequence according to preset priorities, workstation grouping, and linkage strategies; Workstations are grouped according to their priority, function, and spatial location. Compensation instructions for each group are arranged in an optimized order to form a time sequence within the group. For each instruction, the workstation number, physical quantity type, compensation value, and execution timestamp are recorded. The response sequences of all workstations are then summarized to form a multi-workstation collaborative response sequence.
[0040] S5: The final detection data, process status data, and response sequence are transmitted back to the collaborative platform to analyze the contribution of each station's actions to the deviation, dynamically update the long-term parameters of each process, and perform upper and lower limit constraints, baseline reset, and drift calibration on the parameters.
[0041] The process of analyzing the contribution of each workstation's actions to the deviation in S5 is as follows: Match the multi-station collaborative response sequence with the final detection data; Action sequence data is collected from each workstation, including timestamps, action types, execution parameters, and execution order. At the same time, final inspection data, including dimensional deviations, positional offsets, and surface defect indices, are also collected, along with timestamps, process numbers, and equipment numbers. The multi-workstation response sequences and final inspection data are synchronized and aligned according to the timestamps. An index relationship is established between each workstation action and its corresponding inspection result to generate a time series mapping table.
[0042] By using deviation decomposition, statistical regression and multi-physical quantity coupling analysis methods, the deviation contribution value of each workstation in different time periods and different physical quantities is quantified. After alignment, the deviation is decomposed according to the category of physical quantity. The overall detection deviation is broken down into components of each physical quantity. Linear or nonlinear regression methods are used to analyze the correlation between the actions of each workstation and the component deviations, and a deviation contribution matrix is established. At the same time, the coupling relationship between different physical quantities is analyzed, and the numerical contribution of each workstation action to the overall deviation is calculated in each time period, forming a three-dimensional deviation matrix of workstation-time-physical quantity.
[0043] By combining historical process data and response execution, the contribution values are weighted, corrected, and cumulatively analyzed to generate a comprehensive contribution value of each workstation to the overall error. Read historical process data, including workstation action history, equipment status, operating parameters, and historical deviation distribution. Based on the consistency and reliability of workstation execution, perform weighted correction on each unit in the deviation contribution matrix, adjust the contribution value of abnormal or missing data, accumulate and sum the contribution value of each workstation in the time series, and summarize the weighted accumulated workstation deviation contribution by workstation. Calculate the average deviation contribution, variance, and peak contribution during critical periods for each workstation in the entire production process, organize the data into a table or matrix, and output the comprehensive contribution value of each workstation to the final product deviation.
[0044] The process of setting upper and lower limits for parameters, resetting the baseline, and calibrating drift in S5 is as follows: Based on the contribution of each workstation to the overall deviation, combined with historical process data, equipment performance characteristics and inter-process dependencies, the long-term control parameter change trend and cumulative deviation of each process are dynamically analyzed. From the comprehensive deviation contribution value matrix collected from each workstation, the time series data of key control parameters of each process are read. At the same time, historical process data, equipment operation logs and process sequence dependencies are called up. The change curve of each process parameter is plotted according to the time series. The short-term fluctuation and long-term cumulative deviation of the parameters are calculated using the sliding window or exponential smoothing method. The links with deviation amplitude exceeding the preset threshold or with obvious cumulative growth are marked.
[0045] The parameters are corrected in real time according to the upper and lower limits, and key links in the accumulation of deviations are identified and corrected accordingly. For each process's key control parameters, preset upper and lower thresholds are applied. Data exceeding these thresholds is truncated or interpolated for correction, and the values and timestamps before and after correction are recorded. Combined with cumulative deviation analysis results, key processes or key parameter points that significantly contribute to overall deviation are identified. These points are then subject to focused correction, including adjusting target values, modifying control curves, or replacing abnormal data segments, ensuring all parameters remain within the set controllable range.
[0046] At the same time, the baseline is reset, and drift calibration technology is used to gradually adjust for minor offsets; Determine the current baseline parameter value for each process, compare the actual collected parameters with the baseline value, calculate the offset, and use the step-by-step correction method or proportional compensation method to adjust the small offset step by step, gradually pulling the parameter value back to the baseline range. At the same time, update the baseline record in the control system, and record the offset magnitude, adjustment time and corresponding workstation for each adjustment operation to ensure that the baseline of the next cycle corresponds and matches the current equipment status and process execution.
[0047] Example 2: As Figure 2 As shown, a touch display module assembly collaboration system includes: State generation module: acquires bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution and surface stress, and encodes them according to a unified structure to form process state data with multiple physical quantities and visual anomaly markers; The difference calculation module writes process status data into the collaborative platform, calculates the difference item by item according to the process flow sequence, checks the consistency of direction and removes noise, and generates a quantitative difference dataset between processes. Deviation Analysis Module: Extracts displacement, pressure, trajectory, and temperature deviations based on the difference dataset, calculates the propagation coefficient, constructs the deviation propagation matrix, and generates compensation instructions; Command execution module: Sends compensation commands into the virtual compensation process to verify the superposition effect. After verification, it sends them to the equipment. When the trend is abnormal, it triggers hierarchical linkage to generate workstation response sequence according to the deviation propagation matrix. Parameter update module: The final detection data, process status data and response sequence are transmitted back to the platform to analyze the contribution of station actions to deviation, dynamically update the parameters of each process and perform constraint calibration.
[0048] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0049] 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 collaborative assembly method for a touch display module, characterized in that, Includes the following steps: Acquire bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution and surface stress, and perform time serialization and encoding processing according to a unified data structure to form process status data containing multiple physical quantities and visual anomaly markers. Process status data is written to the collaborative platform in real time. The difference between parameters of adjacent processes is calculated one by one according to the process flow sequence. The direction consistency check and noise removal are performed to generate a difference dataset that reflects the quantitative difference relationship between processes. Based on the differential dataset, displacement, pressure, trajectory and temperature deviations are extracted, the inter-process propagation coefficient is calculated, the deviation propagation matrix is constructed and compensation instructions for pressure, speed, trajectory and exposure area are generated; The compensation command is sent into the virtual compensation process to verify the superposition effect. After the verification is successful, it is written to the corresponding device. When an abnormal trend occurs, the hierarchical linkage is triggered according to the deviation propagation matrix to generate a response sequence for group management by workstation. The final test data, along with the process status data and response sequence, are transmitted back to the collaborative platform to analyze the contribution of each workstation's actions to the deviation, dynamically update the long-term parameters of each process, and perform upper and lower limit constraints, baseline reset, and drift calibration on the parameters.
2. The method for collaborative assembly of a touch display module according to claim 1, characterized in that, The process of generating process state data that includes multiple physical quantities and visual anomaly markers is as follows: Key parameters of each process are collected synchronously by multiple sensors, including displacement sensors, pressure sensors, dispensing trajectory tracking systems, thermal imaging or temperature sensors, and surface stress detection devices. Different physical quantities are mapped to a unified time axis and standardized dimensions to form a multidimensional state vector; Simultaneously, visual inspection markers are used to label abnormal features, generating process status data that can be parsed and analyzed across processes on a collaborative platform.
3. The method for collaborative assembly of a touch display module according to claim 2, characterized in that, The process of calculating the difference between parameters of adjacent processes according to the process flow sequence is as follows: Based on process state data, multidimensional state vectors are read in the collaborative platform and standardized with unified dimensions. The difference between each physical quantity data in the continuous process is calculated, and the timestamp, spatial location, and corresponding process, equipment and physical quantity identifier of each deviation are recorded. A weighted algorithm is used to perform preliminary quantification of the deviations between different processes, forming a preliminary deviation matrix between processes.
4. The touch display module assembly collaboration method according to claim 3, characterized in that, The process of generating a difference dataset that reflects the quantitative differences between processes is as follows: Perform directional consistency analysis on the preliminary inter-process deviation matrix; Anomalies are identified by changes in vector direction and amplitude, and data that does not conform to the overall movement trend or process logic is eliminated. The processed deviations are used to generate a multidimensional difference dataset according to the process sequence and physical quantity type.
5. The touch display module assembly collaboration method according to claim 4, characterized in that, The process of constructing the deviation propagation matrix and generating compensation instructions for pressure, speed, trajectory, and exposure area is as follows: Read the multidimensional deviation information of each process in the difference dataset, and perform statistical analysis and weighting on the cumulative effect of deviation in time and space; A multi-physical quantity coupling model is established by combining historical process data, the propagation coefficient of each deviation to subsequent processes is calculated, and a complete deviation propagation matrix is constructed. Multi-physical quantity comprehensive compensation commands can be generated based on the deviation propagation matrix and directly sent to the device.
6. The method for collaborative assembly of a touch display module according to claim 5, characterized in that, The process of sending the compensation command into the virtual compensation process to verify the superposition effect is as follows: In a digital twin platform or simulation environment, the deviation propagation matrix and the multi-physical quantity comprehensive compensation command are superimposed and simulated. The effectiveness of the compensation strategy was verified by combining deviation propagation and the coupling effect of multiple physical quantities. Real-time adjustments are made to address inter-process conflicts, overcompensation, or undercompensation, optimizing the order and magnitude of instructions.
7. The method for collaborative assembly of a touch display module according to claim 6, characterized in that, The process of generating a response sequence for workstation-based group management is as follows: The collaborative platform collects the trend of parameters of each process in real time, and combines the compensation instructions that have been virtually verified and optimized with the deviation propagation matrix. Reverse tracing of abnormal deviations to identify the main contributing workstations and error coupling relationships; By combining historical deviation data, inter-process dependencies, and physical quantity coupling models, the magnitude, sequence of action, and timing of execution of response instructions are dynamically optimized and adjusted. Generate a multi-workstation collaborative response sequence according to preset priorities, workstation grouping, and linkage strategies.
8. The method for collaborative assembly of a touch display module according to claim 7, characterized in that, The process of analyzing the contribution of each workstation's actions to the deviation is as follows: Match the multi-station collaborative response sequence with the final detection data; By using deviation decomposition, statistical regression and multi-physical quantity coupling analysis methods, the deviation contribution value of each workstation in different time periods and different physical quantities is quantified. By combining historical process data and response execution, the contribution values are weighted, corrected, and cumulatively analyzed to generate a comprehensive contribution value of each workstation to the overall error.
9. The method for collaborative assembly of a touch display module according to claim 8, characterized in that, The process of setting upper and lower limits for parameters, resetting the baseline, and performing drift calibration is as follows: Based on the contribution of each workstation to the overall deviation, combined with historical process data, equipment performance characteristics and inter-process dependencies, the long-term control parameter change trend and cumulative deviation of each process are dynamically analyzed. The parameters are corrected in real time according to the upper and lower limits, and key links in the accumulation of deviations are identified and corrected accordingly. Simultaneously, the baseline is reset, and drift calibration techniques are used to gradually adjust for minor offsets.
10. A touch display module assembly collaboration system, applied to the method described in any one of claims 1-9, characterized in that, include: State generation module: acquires bonding displacement, pressure curve, dispensing trajectory offset, curing temperature distribution and surface stress, and encodes them according to a unified structure to form process state data with multiple physical quantities and visual anomaly markers; The difference calculation module writes process status data into the collaborative platform, calculates the difference item by item according to the process flow sequence, checks the consistency of direction and removes noise, and generates a quantitative difference dataset between processes. Deviation Analysis Module: Extracts displacement, pressure, trajectory, and temperature deviations based on the difference dataset, calculates the propagation coefficient, constructs the deviation propagation matrix, and generates compensation instructions; Command execution module: Sends compensation commands into the virtual compensation process to verify the superposition effect. After verification, it sends them to the equipment. When the trend is abnormal, it triggers hierarchical linkage to generate workstation response sequence according to the deviation propagation matrix. Parameter update module: The final detection data, process status data and response sequence are transmitted back to the platform to analyze the contribution of station actions to deviation, dynamically update the parameters of each process and perform constraint calibration.