Laminating method of vehicle-mounted curved surface display module

Abstract

The invention provides a fitting method for a vehicle-mounted curved surface display module, and the method comprises the steps: obtaining temperature data collected by a sensor, and generating an environment temperature sequence containing a timestamp and a temperature value; determining a fitting strength predicted value according to the environment temperature sequence; if the fitting strength predicted value exceeds a preset strength threshold value, determining a fitting parameter adjustment coefficient according to historical data; obtaining a fitting process parameter combination corresponding to the adjustment coefficient from a preset database, and generating an optimized fitting parameter set; inputting the optimized fitting parameter set into a control system, adjusting fitting equipment parameters, and monitoring to generate a fitting strength real-time sequence; and if the fluctuation amplitude of the fitting strength real-time sequence is smaller than a preset fluctuation threshold value, storing the fitting parameter set and the environment temperature sequence to a database.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for laminating a vehicle-mounted curved display module. Background Art

[0002] The bonding strength of machine-cut curved display modules is closely related to the ambient temperature. However, the temperature changes greatly in the actual use environment, which may cause fluctuations in the bonding strength and affect the display effect. It is necessary to study how to ensure the bonding strength of the display module under actual temperature conditions and develop a bonding technology with temperature self-adaptation capability to achieve stable bonding at different temperatures. Summary of the Invention

[0003] The present invention provides a method for laminating a vehicle-mounted curved display module, which mainly includes: Acquire temperature data collected by the sensor and generate an ambient temperature sequence including a timestamp and a temperature value; determine a predicted bonding strength value based on the ambient temperature sequence; if the predicted bonding strength value exceeds a preset strength threshold, determine a bonding parameter adjustment coefficient based on historical data; obtain a bonding process parameter combination corresponding to the adjustment coefficient from a preset database to generate an optimized bonding parameter set; input the optimized bonding parameter set into a control system, adjust the bonding equipment parameters, and monitor and generate a real-time bonding strength sequence; if the fluctuation amplitude of the real-time bonding strength sequence is less than a preset fluctuation threshold, save the bonding parameter set and the ambient temperature sequence to a database; if the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, use an optimization algorithm to iteratively optimize the bonding parameter set, generate a new bonding parameter set, and readjust the bonding equipment parameters.

[0004] Furthermore, the method of acquiring temperature data collected by a sensor and generating an ambient temperature sequence including a timestamp and a temperature value includes: collecting ambient temperature data in real time through a temperature sensor; determining a timestamp corresponding to each data point in the ambient temperature data; generating an ambient temperature sequence in a time series format based on the ambient temperature data and the timestamp, wherein the ambient temperature sequence includes temperature values corresponding to multiple consecutive time points; preprocessing the ambient temperature sequence to filter out abnormal temperature values; determining a time interval and a temperature change trend of the sequence based on the preprocessed ambient temperature sequence; and obtaining a temperature data sequence that can be used for subsequent analysis by verifying the integrity of the ambient temperature sequence.

[0005] Furthermore, determining the predicted value of the fitting strength based on the ambient temperature sequence includes: obtaining a preset temperature-strength mapping model; inputting the ambient temperature sequence into the temperature-strength mapping model; calculating the predicted value of the fitting strength corresponding to each time point through the temperature-strength mapping model; generating a fitting strength prediction sequence based on the fitting strength prediction value; smoothing the fitting strength prediction sequence to eliminate short-term fluctuation interference; and determining whether the predicted value exceeds a preset strength threshold based on the smoothed fitting strength prediction sequence.

[0006] Furthermore, determining the fitting parameter adjustment coefficient based on historical data includes: obtaining historical temperature data and corresponding historical fitting strength data from a historical database; establishing a regression analysis model based on the historical temperature data and the historical fitting strength data; determining the correlation between temperature and fitting strength through the regression analysis model; calculating the fitting strength deviation based on the ambient temperature sequence and the correlation; determining the fitting parameter adjustment coefficient based on the fitting strength deviation; and obtaining the adjustment coefficient for parameter optimization by verifying the stability of the fitting parameter adjustment coefficient.

[0007] Furthermore, obtaining the bonding process parameter combination corresponding to the adjustment coefficient from a preset database to generate an optimized bonding parameter set includes: obtaining a preset bonding parameter database; matching the corresponding bonding process parameter combination in the bonding parameter database according to the bonding parameter adjustment coefficient; determining the pressure parameter and colloid viscosity parameter corresponding to the adjustment coefficient by table lookup; generating an optimized bonding parameter set according to the pressure parameter and the colloid viscosity parameter; formatting the optimized bonding parameter set to adapt to the control system input requirements; and obtaining a parameter set that can be used for equipment adjustment by verifying the integrity of the optimized bonding parameter set.

[0008] Furthermore, inputting the optimized bonding parameter set into the control system to adjust the bonding equipment parameters includes: transmitting the optimized bonding parameter set to the bonding equipment control system; parsing the pressure parameters and colloid viscosity parameters in the optimized bonding parameter set through the control system; adjusting the pressure control module of the bonding equipment according to the pressure parameters; adjusting the colloid supply module of the bonding equipment according to the colloid viscosity parameters; verifying the accuracy of the bonding equipment parameter adjustment through a real-time feedback mechanism; and executing the bonding process operation according to the adjusted bonding equipment parameters.

[0009] Furthermore, the monitoring generates a real-time sequence of fitting strength, including: collecting strength data during the fitting process in real time through a high-precision sensor; generating a real-time sequence of fitting strength containing a timestamp based on the strength data; denoising the real-time sequence of fitting strength to filter out environmental interference; calculating the fluctuation amplitude of the sequence based on the real-time sequence of fitting strength after denoising; determining the changing trend of the fluctuation amplitude through statistical analysis; and generating a real-time sequence of fitting strength for subsequent judgment based on the changing trend.

[0010] Furthermore, if the fluctuation amplitude of the real-time sequence of the fitting strength is less than a preset fluctuation threshold, the fitting parameter set and the ambient temperature sequence are saved to the database, including: obtaining the preset fluctuation threshold; judging whether the fluctuation amplitude meets the requirement by comparing the fluctuation amplitude of the real-time sequence of the fitting strength with the preset fluctuation threshold; if the fluctuation amplitude is less than the preset fluctuation threshold, generating a saving instruction; according to the saving instruction, storing the optimized fitting parameter set and the ambient temperature sequence to the database; confirming that the saving operation is successful by verifying the integrity of the stored data; and updating the records of the historical database according to the saved data.

[0011] Furthermore, the use of an optimization algorithm to iteratively optimize the fitting parameter set includes: obtaining a preset machine learning algorithm; inputting the real-time sequence of fitting strength and the optimized fitting parameter set into the machine learning algorithm; analyzing the correlation between the fitting strength fluctuation and the parameter set through the machine learning algorithm; adjusting the pressure parameters and colloid viscosity parameters in the optimized fitting parameter set according to the correlation; generating a new fitting parameter set through iterative calculation; and obtaining a parameter set that can be used to readjust device parameters by verifying the validity of the new fitting parameter set.

[0012] Furthermore, if the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, a new bonding parameter set is generated and the bonding equipment parameters are readjusted, including: judging whether optimization is needed by comparing the fluctuation amplitude of the real-time bonding strength sequence with the preset fluctuation threshold; if the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, triggering the execution of the optimization algorithm; generating a new bonding parameter set through the optimization algorithm; transmitting the new bonding parameter set to the bonding equipment control system; readjusting the pressure parameters and colloid viscosity parameters of the bonding equipment according to the new bonding parameter set; and verifying the effectiveness of the readjusted bonding equipment parameters through real-time monitoring.

[0013] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses a method for adaptively controlling bonding strength based on ambient temperature. The method collects ambient temperature data in real time and establishes a temperature-strength mapping model. The bonding strength is predicted and compared with a preset threshold. A parameter adjustment coefficient is determined based on historical data regression analysis. The corresponding process parameter combination is obtained from a database and optimized. The optimized parameters are input into a control system to adjust equipment parameters. High-precision sensors are used to monitor bonding strength, and the need for further optimization is determined based on the amplitude of fluctuations. The present invention achieves adaptive control of bonding strength through closed-loop feedback of temperature monitoring, strength prediction, parameter optimization, and real-time monitoring. This method effectively improves the stability and consistency of bonding quality and is suitable for precision bonding processes that are significantly affected by ambient temperature. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The present invention is a flowchart of a method for laminating a vehicle-mounted curved display module. DETAILED DESCRIPTION

[0015] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0016] like Figure 1 In this embodiment, a method for laminating a vehicle-mounted curved display module may specifically include: S101: Receive temperature data collected by a sensor and generate an ambient temperature sequence including a timestamp and a temperature value.

[0017] Real-time temperature data is received through the sensor interface, generating a signal sequence containing raw temperature values. A system time stamp is added to the signal sequence to generate an ambient temperature sequence containing both timestamps and temperature values. If the timestamps in the ambient temperature sequence are unevenly spaced, an interpolation algorithm is used to fill in missing time points to generate a temperature sequence with uniform time intervals. A preset filtering algorithm is used to process the temperature sequence with uniform time intervals to eliminate noise and generate a smoothed ambient temperature sequence. The temperature change rate is calculated based on the smoothed ambient temperature sequence to determine the dynamic eigenvalue of the temperature sequence. If the dynamic eigenvalue exceeds a preset range, a reference sequence similar to the current temperature sequence is retrieved from the historical database to obtain a matching environmental parameter set. The matching environmental parameter set is used to adjust the bonding process parameters to generate a preliminary optimized bonding parameter set. The fluctuation amplitude is calculated based on the preliminary optimized bonding parameter set and the real-time bonding strength sequence to determine whether the fluctuation is less than a preset threshold. If the fluctuation amplitude is less than the preset threshold, the current bonding parameter set and the smoothed ambient temperature sequence are saved to the database, updating the historical reference data.

[0018] Exemplarily, a sensor interface receives raw temperature values collected by a temperature sensor at a frequency of once per second, generating a signal sequence containing temperature values. A system time stamp is added to the signal sequence to generate an ambient temperature sequence containing timestamps and temperature values, with timestamps accurate to milliseconds. If the timestamps in the ambient temperature sequence are unevenly spaced, for example, with time intervals greater than 1 second, a linear interpolation algorithm is used to fill in missing time points to obtain a temperature sequence with uniform time intervals. The temperature sequence with uniform time intervals is processed using a preset Kalman filter algorithm to eliminate noise interference and obtain a smoothed ambient temperature sequence. The filtered temperature values are rounded to two decimal places. The temperature change rate is calculated based on the smoothed ambient temperature sequence, for example, by calculating the temperature difference between adjacent time points using a difference method. The dynamic characteristic value of the temperature sequence is determined, for example, if the temperature change rate exceeds 0.5°C / second. If the dynamic characteristic value exceeds a preset range, a reference sequence similar to the current temperature sequence is retrieved from a historical database. For example, a Euclidean distance algorithm is used to match the historical sequence with the highest similarity to obtain a matching environmental parameter set. The bonding process parameters are adjusted based on the matching environmental parameter set, for example, by adjusting the bonding pressure to 0.8 MPa, generating a preliminary optimized bonding parameter set. The fluctuation amplitude is calculated based on the initially optimized fitting parameter set and the real-time fitting strength sequence, for example, using a standard deviation algorithm, and a determination is made as to whether the fluctuation is less than a preset threshold of 0.1. If the fluctuation amplitude is less than the preset threshold, the current fitting parameter set and the smoothed ambient temperature sequence are saved to the database, and the historical reference data is updated, for example, by marking the current parameter set as the optimal parameter set.

[0019] S102. Calculate a predicted bonding strength value using a temperature-strength mapping model based on the ambient temperature sequence. If the predicted value exceeds a preset strength threshold, determine a bonding parameter adjustment coefficient based on regression analysis of historical temperature data and strength data.

[0020] Obtain the ambient temperature sequence and calculate the predicted value of the fitting strength through the temperature-strength mapping model. If the predicted value of the fitting strength exceeds the preset strength threshold, extract the historical temperature data and strength data to obtain the extraction result. According to the extraction result, use the regression analysis method to calculate the fitting parameter adjustment coefficient and determine the adjustment coefficient. Update the current fitting parameter set through the adjustment coefficient to generate an updated parameter set. Obtain the real-time sequence of fitting strength, calculate its real-time fluctuation amplitude, and obtain the fluctuation amplitude value. If the fluctuation amplitude value is less than the preset fluctuation threshold, save the current parameter set and the ambient temperature sequence to the database to obtain a saved record. If the fluctuation amplitude value is greater than or equal to the preset fluctuation threshold, use the machine learning algorithm to iteratively optimize the current parameter set to generate a new parameter set. According to the new parameter set, recalculate the predicted value of the fitting strength through the temperature-strength mapping model to obtain a new predicted value. Repeat steps 2 to 7 with the new predicted value to determine the final parameter set.

[0021] For example, a temperature sequence [25°C, 28°C, 30°C, 27°C] is collected from a sensor and input into a temperature-intensity mapping model y=0.5x+10 (where x is the temperature value and y is the predicted intensity value). This results in a predicted bonding intensity sequence [22.5, 24.0, 25.0, 23.5]. When the predicted value 25.0 exceeds the preset intensity threshold of 24.0, the database retrieves three months of historical temperature-intensity data [(20°C, 18.0), (22°C, 19.5)…(30°C, 24.8)]. Linear regression analysis is used to establish the equation y=0.48x+8.2, resulting in a slope of 0.48 as the adjustment coefficient. This coefficient is then multiplied by the current pressure parameter of 1.2 MPa to update the value to 1.15 MPa. The fit strength sequence [23.1, 23.3, 22.9, 23.0] is monitored in real time. The calculated standard deviation of 0.15 is less than the fluctuation threshold of 0.2, and the parameter set {temperature 27°C, pressure 1.15 MPa} is written to the MySQL database. If the standard deviation of 0.25 exceeds the threshold, the random forest algorithm is used to iteratively optimize the parameters using the mean squared error as the loss function, generating a new parameter set {temperature 26.5°C, pressure 1.18 MPa}. The temperature sequence [26.5°C, 27.0°C] is re-entered into the mapping model to obtain the new predicted value [23.3, 23.5]. The regression analysis and parameter optimization process is repeated until the final parameter set {temperature 26.8°C, pressure 1.16 MPa} is output.

[0022] S103 , obtaining a bonding process parameter combination corresponding to the adjustment coefficient from a preset bonding parameter database, and generating an optimized bonding parameter set.

[0023] The initial bonding parameter set is retrieved from a preset bonding parameter database for bonding process parameter combinations that match the adjustment coefficients. Based on the initial bonding parameter set, a data normalization algorithm is used to normalize the parameter values to obtain a standardized parameter set. A machine learning model is used to extract and optimize the standardized parameter set to generate an optimized bonding parameter set. If the deviation of the optimized bonding parameter set exceeds a preset deviation threshold, a genetic algorithm is used to iteratively adjust the parameter set to obtain an adjusted parameter set. Based on the adjusted parameter set, a control instruction sequence is generated to determine the pressure and colloid viscosity parameter configurations for the bonding equipment. Real-time bonding process data is collected using high-precision sensors to generate a real-time bonding strength sequence. If the fluctuation amplitude of the real-time bonding strength sequence exceeds or equals the preset fluctuation threshold, a time series analysis algorithm is used to extract fluctuation characteristics to obtain a fluctuation feature set. Based on the fluctuation feature set, a reinforcement learning algorithm is used to iteratively optimize the bonding parameter set to generate a new parameter set. The control instruction sequence is updated with the new parameter set to adjust the operating parameters of the bonding equipment and determine the new bonding process configuration.

[0024] Exemplarily, a combination of bonding process parameters that matches the adjustment coefficient is retrieved from a preset bonding parameter database. For example, when the adjustment coefficient is 1.2, the database is queried to obtain a combination of a pressure parameter range of 20-30 MPa and a colloid viscosity parameter range of 50-80 cP to obtain an initial bonding parameter set. According to the initial bonding parameter set, the parameter values are normalized using the Z-score normalization algorithm. For example, the pressure parameter 25 MPa is converted to a standard value of 0.5, and the colloid viscosity parameter 65 cP is converted to a standard value of 0.3 to obtain a standardized parameter set. The standardized parameter set is feature extracted and optimized by a machine learning model. For example, the parameter weights are analyzed using a random forest algorithm, and the pressure parameter weight is adjusted to 0.6 and the colloid viscosity parameter weight is adjusted to 0.4 to generate an optimized bonding parameter set. If the deviation value of the optimized bonding parameter set is greater than the preset deviation threshold of 0.1, a genetic algorithm is used to iteratively adjust the parameter set. For example, a new combination of a pressure parameter of 28 MPa and a colloid viscosity parameter of 70 cP is generated by a cross-mutation operation to obtain an adjusted parameter set. Based on the adjusted parameter set, a control instruction sequence is generated. For example, the pressure control instruction is set to 28 MPa ± 0.5 MPa and the colloid viscosity control instruction is set to 70 cP ± 2 cP, thereby determining the pressure and colloid viscosity parameter configurations for the bonding equipment. Real-time bonding process data is collected using high-precision sensors, for example, bonding strength data is collected every 100 ms to generate a real-time bonding strength sequence. If the fluctuation amplitude of the real-time bonding strength sequence is greater than or equal to a preset fluctuation threshold of 5%, a time series analysis algorithm is used to extract fluctuation characteristics. For example, an ARIMA model is used to identify periodic fluctuations with a frequency of 0.1 Hz, thereby generating a fluctuation feature set. Based on the fluctuation feature set, a reinforcement learning algorithm is used to iteratively optimize the bonding parameter set. For example, a Q-learning algorithm is used to adjust the pressure parameter to 29 MPa and the colloid viscosity parameter to 75 cP, thereby generating a new parameter set. The control instruction sequence is updated using the new parameter set. For example, the pressure control instruction is adjusted to 29 MPa ± 0.5 MPa and the colloid viscosity control instruction is adjusted to 75 cP ± 2 cP. The operating parameters of the bonding equipment are adjusted to determine the new bonding process configuration.

[0025] S104: Input the optimized bonding parameter set into the control system, adjust the pressure and colloid viscosity parameters of the bonding equipment, and use high-precision sensors to monitor and generate a real-time sequence of bonding strength.

[0026] The bonding process parameter combination corresponding to the adjustment coefficient is obtained from the preset bonding parameter database to generate an optimized bonding parameter set. The optimized bonding parameter set is received by the control system, and the pressure and colloid viscosity parameters of the bonding equipment are adjusted. High-precision sensors are used to collect strength data during the bonding process in real time to generate a real-time bonding strength sequence. The fluctuation amplitude is calculated based on the real-time bonding strength sequence to obtain a fluctuation amplitude value. If the fluctuation amplitude value is greater than or equal to the preset fluctuation threshold, a machine learning algorithm is initiated to iteratively optimize the bonding parameter set to generate a new parameter set. The new parameter set is received by the control system, and the pressure and colloid viscosity parameters of the bonding equipment are updated. High-precision sensors are used to collect bonding strength data again to generate an updated real-time bonding strength sequence. The fluctuation amplitude is recalculated based on the updated real-time bonding strength sequence to determine whether the fluctuation amplitude is less than the preset fluctuation threshold. If the fluctuation amplitude is less than the preset fluctuation threshold, the current parameter set is stored in the bonding parameter database, and the process parameter combination in the database is updated.

[0027] Exemplarily, a combination of bonding process parameters corresponding to an adjustment coefficient of 0.85 is extracted from a preset bonding parameter database to generate an optimized bonding parameter set, including a pressure parameter of 2.5 MPa and a colloid viscosity parameter of 1200 cP. The parameter set is received by the control system, and the pressure of the bonding equipment is adjusted to 2.5 MPa and the colloid viscosity to 1200 cP. A high-precision sensor is used to collect strength data during the bonding process in real time at a frequency of 100 Hz to generate a real-time bonding strength sequence, and the sequence data is [12.3N, 12.5N, 12.4N, 12.6N, 12.2N]. The fluctuation amplitude is calculated based on the real-time bonding strength sequence, and the standard deviation algorithm is used to obtain a fluctuation amplitude value of 0.15N. If the fluctuation amplitude value is greater than the preset fluctuation threshold of 0.1N, the random forest model in the machine learning algorithm is started to iteratively optimize the bonding parameter set to generate a new parameter set, including a pressure parameter of 2.6 MPa and a colloid viscosity parameter of 1150 cP. The control system receives the new parameter set, updating the laminating equipment's pressure to 2.6 MPa and the colloid viscosity to 1150 cP. A high-precision sensor is used to collect laminating strength data again, generating an updated real-time laminating strength sequence: [12.4 N, 12.4 N, 12.5 N, 12.4 N, 12.5 N]. The fluctuation amplitude is recalculated based on the updated real-time laminating strength sequence. Using the standard deviation algorithm, the fluctuation amplitude is 0.05 N, which is determined to be less than the preset fluctuation threshold of 0.1 N. The current parameter set is stored in the laminating parameter database, and the process parameter combination in the database is updated to 2.6 MPa pressure and 1150 cP colloid viscosity.

[0028] S105: If the fluctuation amplitude of the real-time sequence of the bonding strength is less than a preset fluctuation threshold, save the current parameter set and the ambient temperature sequence to a database.

[0029] Acquire real-time sequence data of bonding strength, and generate a real-time bonding strength sequence through high-precision sensor acquisition. Based on the real-time bonding strength sequence, calculate the sequence fluctuation amplitude and obtain the fluctuation amplitude value. If the fluctuation amplitude value is less than the preset fluctuation threshold, trigger the parameter saving process and determine that the saving condition is met. Acquire the current bonding parameter set and ambient temperature sequence, and extract the current parameters and ambient temperature data collected by the sensor through the control system. Save the current parameter set and ambient temperature sequence to the database through the database interface to complete data storage. Based on the saved ambient temperature sequence, use the temperature-intensity mapping model to calculate the predicted bonding strength value and obtain the predicted value. If the predicted value exceeds the preset intensity threshold, perform regression analysis based on the historical temperature data and intensity data to determine the bonding parameter adjustment coefficient. Input the optimized bonding parameter set through the control system interface to adjust the pressure and colloid viscosity parameters of the bonding equipment. Use high-precision sensors to monitor the adjusted bonding process, generate a new real-time bonding strength sequence, and obtain updated sequence data.

[0030] For example, a high-precision sensor collects fit strength data at a sampling frequency of 100 Hz, generating a real-time series of 1000 data points. A sliding window standard deviation algorithm is used to calculate the series fluctuation amplitude, with a window size of 50 data points. If the standard deviation falls below a preset threshold of 0.05 MPa, the fluctuation amplitude is considered to meet the requirements. After triggering the save process, the current pressure parameters (1.2 MPa ± 0.1 MPa) and colloid viscosity parameters (3500 cP) are read from the PLC control system. The ambient temperature series is simultaneously collected (with a sampling interval of 10 seconds and an accuracy of ±0.5°C). The parameter set and temperature series are written to the process_parameters table in a MySQL database via the JDBC interface, with the timestamp as the primary key. A temperature-strength mapping model (polynomial regression R² ≥ 0.95) is invoked, and the input temperature series [25.3°C, 25.1°C, 25.4°C] outputs a predicted strength value of 1.8 MPa. If the predicted value exceeds the threshold of 1.5 MPa, the ridge regression model (α = 0.5) is trained using the last 30 days of historical data. The resulting pressure adjustment coefficient is 0.92 and the viscosity coefficient is 1.05. The adjusted parameters (pressure 1.1 MPa, viscosity 3675 cP) are sent to the device controller via the Modbus protocol. The sensor continuously monitors the intensity sequence under these new parameters, generating an updated data stream containing [1.53 MPa, 1.55 MPa, 1.52 MPa].

[0031] S106. If the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, the fitting parameter set is iteratively optimized using a machine learning algorithm, and a new parameter set is generated and the process returns to step 4 for execution.

[0032] Real-time fluctuation data for the bonding parameter set is obtained. The fluctuation amplitude is quantified by calculating the standard deviation to mean ratio of each parameter value in the parameter set. If the fluctuation amplitude is greater than or equal to a preset fluctuation threshold, a machine learning algorithm is used to extract features from the bonding parameter set and identify the key influencing factors. Based on these key influencing factors, an iterative optimization model based on gradient boosting is constructed to generate a new bonding parameter set. The control system receives the new bonding parameter set and adjusts the pressure and colloid viscosity parameters of the bonding equipment to obtain the updated equipment operating status. Real-time bonding strength sequence data is collected using a high-precision sensor. The real-time fluctuation amplitude is calculated through time series analysis to determine the bonding strength stability. If the fluctuation amplitude of the real-time bonding strength sequence is less than the preset fluctuation threshold, the current bonding parameter set and the ambient temperature sequence are extracted to generate a structured data record. This structured data record is saved to the database via a database interface, and the historical optimization record of the parameter set is updated. Based on the historical optimization records in the database, the correlation pattern between the parameter set and the ambient temperature sequence is analyzed to obtain trend characteristics of the optimized parameters. This trend characteristic is used to update the training dataset of the machine learning algorithm, and the model parameters are iteratively optimized to generate a more accurate bonding parameter set.

[0033] Exemplarily, the real-time fluctuation amplitude data of the fitting parameter set is obtained, and the calculation method of the standard deviation to mean ratio is adopted. For example, the fluctuation amplitude of the pressure parameter is calculated to be 0.15, and the fluctuation amplitude of the colloid viscosity parameter is 0.12. The overall fluctuation amplitude quantification value is obtained as 0.14. If the fluctuation amplitude is greater than the preset threshold value of 0.10, the random forest algorithm is used to extract features of the fitting parameter set, and the influence weights of pressure, colloid viscosity and ambient temperature are analyzed to be 0.45, 0.35 and 0.20 respectively, and pressure is determined to be the key influencing factor. According to the key influencing factors, an XGBoost iterative optimization model is constructed, and the learning rate is set to 0.01, the number of iterations is 100, and a new fitting parameter set is generated. The pressure is adjusted to 12.5MPa and the colloid viscosity is adjusted to 850cP. The new parameter set is received by the control system, the pressure of the fitting equipment is adjusted to 12.5MPa, the colloid viscosity is adjusted to 850cP, and the equipment operation status is recorded as stable. A high-precision pressure sensor was used to collect fitting strength data at a frequency of 100Hz. The real-time fluctuation amplitude was calculated using a sliding window with a window size of 50 data points, resulting in a fluctuation amplitude of 0.08. If the fluctuation amplitude was less than the preset threshold of 0.10, the current parameter set and the ambient temperature of 25.3°C were extracted to generate a structured data record in JSON format. An INSERT operation was performed through the MySQL database interface, and the structured data was saved to the optimization_history table, with the updated history record ID set to 1024. Based on the historical records in the database, the Pearson correlation coefficient was used to analyze the association pattern between the parameter set and the ambient temperature, resulting in a correlation coefficient of -0.65 between pressure and temperature, and a correlation coefficient of -0.42 between colloid viscosity and temperature. The training dataset of the random forest algorithm was updated based on trend features, 100 new sample data were added, and the model was retrained to generate an optimized fitting parameter set. The pressure was adjusted to 12.3MPa and the colloid viscosity was adjusted to 860cP.

[0034] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for laminating a vehicle-mounted curved display module, characterized in that: include: Get the temperature data collected by the sensor and generate an ambient temperature sequence containing timestamps and temperature values; Determining a predicted bonding strength value according to the ambient temperature sequence; If the predicted bonding strength value exceeds a preset strength threshold, a bonding parameter adjustment coefficient is determined based on historical data; a bonding process parameter combination corresponding to the adjustment coefficient is obtained from a preset database to generate an optimized bonding parameter set; the optimized bonding parameter set is input into a control system, bonding equipment parameters are adjusted, and a real-time bonding strength sequence is generated through monitoring; if the fluctuation amplitude of the real-time bonding strength sequence is less than a preset fluctuation threshold, the bonding parameter set and the ambient temperature sequence are saved in a database; If the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, the bonding parameter set is iteratively optimized using an optimization algorithm to generate a new bonding parameter set and readjust the bonding device parameters.

2. The method according to claim 1, wherein The method of acquiring temperature data collected by a sensor and generating an ambient temperature sequence including a timestamp and a temperature value includes: collecting ambient temperature data in real time through a temperature sensor; determining a timestamp corresponding to each data point in the ambient temperature data; generating an ambient temperature sequence in a time series format based on the ambient temperature data and the timestamp, wherein the ambient temperature sequence includes temperature values corresponding to multiple consecutive time points; preprocessing the ambient temperature sequence to filter out abnormal temperature values; determining a time interval and a temperature change trend of the sequence based on the preprocessed ambient temperature sequence; and obtaining a temperature data sequence that can be used for subsequent analysis by verifying the integrity of the ambient temperature sequence.

3. The method according to claim 1, wherein Determining the predicted bonding strength value based on the ambient temperature sequence includes: obtaining a preset temperature-strength mapping model; inputting the ambient temperature sequence into the temperature-strength mapping model; calculating the predicted bonding strength value corresponding to each time point through the temperature-strength mapping model; generating a bonding strength prediction sequence based on the bonding strength prediction value; smoothing the bonding strength prediction sequence to eliminate short-term fluctuation interference; and determining whether the predicted value exceeds a preset strength threshold based on the smoothed bonding strength prediction sequence.

4. The method according to claim 1, wherein The method of determining the fitting parameter adjustment coefficient based on historical data includes: obtaining historical temperature data and corresponding historical fitting strength data from a historical database; establishing a regression analysis model based on the historical temperature data and the historical fitting strength data; determining the correlation between temperature and fitting strength through the regression analysis model; calculating the fitting strength deviation based on the ambient temperature sequence and the correlation; determining the fitting parameter adjustment coefficient based on the fitting strength deviation; and obtaining the adjustment coefficient for parameter optimization by verifying the stability of the fitting parameter adjustment coefficient.

5. The method according to claim 1, wherein The step of obtaining a bonding process parameter combination corresponding to the adjustment coefficient from a preset database to generate an optimized bonding parameter set includes: obtaining a preset bonding parameter database; matching a corresponding bonding process parameter combination in the bonding parameter database according to the bonding parameter adjustment coefficient; determining a pressure parameter and a colloid viscosity parameter corresponding to the adjustment coefficient by table lookup; generating an optimized bonding parameter set according to the pressure parameter and the colloid viscosity parameter; formatting the optimized bonding parameter set to adapt to control system input requirements; and obtaining a parameter set that can be used for equipment adjustment by verifying the integrity of the optimized bonding parameter set.

6. The method according to claim 1, wherein The step of inputting the optimized bonding parameter set into a control system and adjusting bonding equipment parameters includes: transmitting the optimized bonding parameter set to a bonding equipment control system; parsing the pressure parameters and colloid viscosity parameters in the optimized bonding parameter set through the control system; adjusting the pressure control module of the bonding equipment according to the pressure parameters; adjusting the colloid supply module of the bonding equipment according to the colloid viscosity parameters; verifying the accuracy of the bonding equipment parameter adjustment through a real-time feedback mechanism; and executing bonding process operations according to the adjusted bonding equipment parameters.

7. The method according to claim 1, wherein The monitoring generates a real-time sequence of bonding strength, including: collecting strength data during the bonding process in real time through a high-precision sensor; generating a real-time sequence of bonding strength containing a timestamp based on the strength data; denoising the real-time sequence of bonding strength to filter out environmental interference; calculating the fluctuation amplitude of the sequence based on the real-time sequence of bonding strength after denoising; determining the changing trend of the fluctuation amplitude through statistical analysis; and generating a real-time sequence of bonding strength for subsequent judgment based on the changing trend.

8. The method according to claim 1, wherein If the fluctuation amplitude of the real-time sequence of the fitting strength is less than a preset fluctuation threshold, the fitting parameter set and the ambient temperature sequence are saved to the database, including: obtaining the preset fluctuation threshold; judging whether the fluctuation amplitude meets the requirement by comparing the fluctuation amplitude of the real-time sequence of the fitting strength with the preset fluctuation threshold; if the fluctuation amplitude is less than the preset fluctuation threshold, generating a saving instruction; according to the saving instruction, storing the optimized fitting parameter set and the ambient temperature sequence to the database; confirming that the saving operation is successful by verifying the integrity of the stored data; and updating the records of the historical database according to the saved data.

9. The method according to claim 1, wherein The iterative optimization of the fitting parameter set using an optimization algorithm includes: obtaining a preset machine learning algorithm; inputting the real-time sequence of fitting strength and the optimized fitting parameter set into the machine learning algorithm; analyzing the correlation between the fitting strength fluctuation and the parameter set through the machine learning algorithm; adjusting the pressure parameters and colloid viscosity parameters in the optimized fitting parameter set based on the correlation; generating a new fitting parameter set through iterative calculation; and obtaining a parameter set that can be used to readjust device parameters by verifying the validity of the new fitting parameter set.

10. The method according to claim 1, wherein If the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, a new bonding parameter set is generated and the bonding equipment parameters are readjusted, including: judging whether optimization is needed by comparing the fluctuation amplitude of the real-time bonding strength sequence with the preset fluctuation threshold; if the fluctuation amplitude is greater than or equal to the preset fluctuation threshold, triggering the execution of the optimization algorithm; generating a new bonding parameter set through the optimization algorithm; transmitting the new bonding parameter set to the bonding equipment control system; according to the new bonding parameter set, readjusting the pressure parameters and colloid viscosity parameters of the bonding equipment; and verifying the effectiveness of the readjusted bonding equipment parameters through real-time monitoring.

Images