Method for optimizing production process of organic light-emitting device

Abstract

The invention relates to the technical field of organic light-emitting device production process data processing, and provides an organic light-emitting device production process optimization method, which comprises the following steps of: monitoring key parameters in an evaporation cavity in real time, analyzing data in a cleaning treatment process by utilizing a preset statistical algorithm, and optimizing the production process of an organic light-emitting device. And extracting feature data influencing the surface cleanliness and flatness of the substrate. And then, establishing a mathematical model in the vacuum evaporation process, and solving by applying an optimization algorithm to obtain a parameter combination which enables the performance of the device to be optimal. Furthermore, electrical and optical performance and life stability data of the device are collected, and optimization parameters are simulated, calculated and optimized through a data processing model. In the production process, equipment operation data are collected in real time, optimization parameters and real-time data are compared, and the parameters are optimized in real time through a stochastic gradient descent method and stored in a knowledge base. The device performance and the production efficiency are effectively improved, and the production cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing for the production process of organic electroluminescent devices, and particularly to a method for optimizing the production process of organic electroluminescent devices. Background Art

[0002] The production process of organic electroluminescent devices is a combination of functional thin-film processes and surface treatment processes. During the preparation process, first, the ITO conductive glass substrate needs to be strictly cleaned. The cleaning process includes chemical cleaning and ultrasonic cleaning to improve the surface cleanliness and flatness of the ITO conductive glass substrate, so as to optimize hole injection and the film-forming property of organic materials. Through vacuum thermal evaporation technology, organic functional thin films such as a hole injection layer, a hole transport layer, a light-emitting layer, and an electron transport layer, as well as a metal cathode, are sequentially deposited on the ITO substrate. During the evaporation process, the vacuum degree, evaporation rate, and film thickness need to be strictly controlled. In addition, in order to improve the device performance, a buffer layer can also be evaporated between the anode and the light-emitting layer. Finally, the prepared device is encapsulated to prevent the influence of moisture and oxygen on the device performance.

[0003] In the production process of organic electroluminescent devices, small changes in parameters such as the vacuum degree, evaporation rate, and film thickness during the vacuum evaporation process have a significant impact on the device performance. In data processing, the prior art cannot accurately simulate and optimize parameters such as the vacuum degree, evaporation rate, and film thickness during the vacuum evaporation process. To solve this technical pain point, the present invention provides a method for optimizing the production process of organic electroluminescent devices. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a method for optimizing the production process of organic electroluminescent devices, which solves the problem that the prior art cannot accurately simulate and optimize parameters such as the vacuum degree, evaporation rate, and film thickness during the vacuum evaporation process.

[0005] To solve the above technical problems, the specific technical solution of the present invention is as follows: The present invention provides a method for optimizing the production process of organic electroluminescent devices, including: Step S101: Obtain the types of chemical cleaning agents, the concentrations of chemical cleaning agents, the cleaning time of chemical cleaning agents, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data during the cleaning process of the organic electroluminescent conductive glass substrate. Use a preset statistical algorithm to analyze the types of chemical cleaning agents, the concentrations of chemical cleaning agents, the cleaning time of chemical cleaning agents, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data obtained during the cleaning process of the organic electroluminescent conductive glass substrate, and obtain the characteristic data affecting the substrate surface cleanliness and the characteristic data affecting the substrate surface flatness. Step S102: Real-time collect the vacuum degree in the evaporation chamber, the evaporation rate of the organic functional layer material, and the thin film thickness parameter data. Based on the collected vacuum degree in the evaporation chamber, the evaporation rate of the organic functional layer material, and the thin film thickness parameter data, establish a mathematical model of the vacuum evaporation process. Use an optimization algorithm to solve the mathematical model of the vacuum evaporation process, and obtain the combination of the vacuum degree, evaporation rate, and thin film thickness parameters with the optimal device performance.

[0006] Step S103: Collect the electrical performance data, optical performance data, and lifetime stability detection data of the device. Evaluate the device performance according to the collected electrical performance data, optical performance data, and lifetime stability detection data of the device, and obtain the data characteristics affecting the device performance. Simulate and calculate the data characteristics affecting the device performance through the vacuum evaporation process data processing model, and obtain the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters. Step S104: Send the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters to the organic electroluminescent device production equipment, and collect the operation data of the organic electroluminescent device production equipment after executing the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters to obtain the real-time data of the organic electroluminescent device production equipment. Step S105: Compare the real-time data of the organic electroluminescent device production equipment with the parameter combinations of the vacuum degree, evaporation rate, and film thickness that are optimal for device performance. If there is an error in the comparison result, obtain the error data information, extract the data characteristics from the error data information to obtain the error data features, use the stochastic gradient descent method to optimize the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter again to obtain the real-time optimized parameter data for the preparation of organic electroluminescent devices, establish a corresponding relationship between the real-time optimized parameter data for the preparation of organic electroluminescent devices and the error data features, and establish labels to store them in the parameter model knowledge base. Match the real-time error data features in the parameter model knowledge base. If the match is successful, retrieve the corresponding real-time optimized parameter data for the preparation of organic electroluminescent devices in the parameter model knowledge base, and use the retrieved corresponding real-time optimized parameter data for the preparation of organic electroluminescent devices in the parameter model knowledge base as the real-time operating parameters of the organic electroluminescent device production process.

[0007] Further, for the organic electroluminescent device production process optimization method of the present invention, the step S101 includes: Substrate surface cleanliness detection: After ultrasonic cleaning, detect the cleanliness of the substrate surface to obtain cleanliness detection data; Organize the obtained types, concentrations, and cleaning times of chemical cleaning agents, frequencies, powers, and cleaning times of ultrasonic cleaning, and substrate surface cleanliness detection data into a data set; Use a preset statistical algorithm to analyze the data set. The preset statistical algorithms include analysis of variance, regression analysis, and correlation analysis. According to the data analysis results, extract the characteristic data affecting the substrate surface cleanliness and the characteristic data affecting the substrate surface flatness.

[0008] Further, for the organic electroluminescent device production process optimization method of the present invention, the step S102 includes: Based on the collected data, establish a mathematical model of the vacuum evaporation process, and obtain the relationship between the vacuum degree, evaporation rate, and film thickness through the mathematical model of the vacuum evaporation process; Use an optimization algorithm to solve the established mathematical model of the vacuum evaporation process, and determine the optimization objective. The optimization objective is to find the parameter combination of the vacuum degree, evaporation rate, and film thickness that makes the device performance optimal; Verify the optimized parameter combination, and evaluate the effect of the optimized parameter combination in actual production through a preset simulation model; If the verification result meets the preset standard, then use the optimized parameter combination of the vacuum degree, evaporation rate, and film thickness as the final production parameters.

[0009] Further, in the method for optimizing the production process of the organic electroluminescent device according to the present invention, step S103 includes: Statistically analyze the collected electrical performance data, optical performance data, and life stability detection data, and calculate the average value and standard deviation statistic of each performance parameter; Based on the established mathematical model of the vacuum evaporation process, construct a data processing model, which is used to receive the data characteristics affecting the device performance as input and output the adjustment suggestions for the vacuum degree, evaporation rate, and thin film thickness parameters; Input the extracted data characteristics affecting the device performance into the data processing model, and the data processing model performs simulation calculations according to the input data to obtain the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter; Organize the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter obtained from the simulation calculation to form a data table.

[0010] Further, in the method for optimizing the production process of the organic electroluminescent device according to the present invention, step S104 includes: Set or select to use the received optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter on the production equipment, start the production equipment, and start producing the organic electroluminescent device; During the production process, real-time monitor the operation data of the organic electroluminescent device production equipment, and the operation data of the organic electroluminescent device production equipment includes the vacuum degree, evaporation rate, and thin film thickness parameter; Compare the operation data of the organic electroluminescent device production equipment with the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter to verify whether the production equipment executes the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter as expected.

[0011] Further, in the method for optimizing the production process of the organic electroluminescent device according to the present invention, step S105 includes: The data characteristics extracted from the error data include the change trend, fluctuation range, and occurrence frequency of the error; Preprocess the extracted data characteristics and optimize them using the stochastic gradient descent method; Based on the extracted error data characteristics, construct an optimization model, and the optimization model takes the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter as input; Use the stochastic gradient descent method to train the optimization model, adjust the vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter to minimize the error, and obtain the real-time optimized preparation parameter data of the organic electroluminescent device.

[0012] Establish a corresponding relationship between the production parameter data of the organic electroluminescent device optimized in real time and the error data characteristics.

[0013] Store information such as the parameter data and error data characteristics of the corresponding relationship in the parameter model knowledge base, and establish a label for each record; During the production process, collect the error data characteristics in real time and perform matching in the parameter model knowledge base; If the matching is successful, that is, a record corresponding to the real-time error data characteristics is found, then retrieve the corresponding production parameter data of the organic electroluminescent device optimized in real time in this record; Use the retrieved production parameter data of the organic electroluminescent device optimized in real time as the real-time operating parameters of the production process.

[0014] Furthermore, for the organic electroluminescent device production process optimization method of the present invention, step S105 includes: When using the stochastic gradient descent method to optimize the parameters in the organic electroluminescent device production process, the specific steps are as follows: Receive and set the initial parameters, and set the initial values for the vacuum degree, evaporation rate, and thin film thickness parameters; Receive and set the learning rate, construct a loss function, and according to the optimization goal of the production process, construct a loss function for the relationship between the device performance and the parameters. Calculate the value of the loss function using the parameter values, that is, the performance deviation under the production conditions; Generate randomly selected samples, and based on the randomly selected samples, calculate the gradient of the loss function with respect to the parameters, that is, the partial derivative of the loss value with respect to the parameters; Update the parameters using the learning rate and the gradient value, that is, adjust the parameter values in the opposite direction of the gradient, and repeat the iteration until the preset number of iterations is reached, the loss value converges to an acceptable range, or the stop condition is met.

[0015] Advantages of the present invention: By monitoring key parameters such as the vacuum degree, evaporation rate, and thin film thickness in the evaporation chamber in real time and using an optimization algorithm to precisely control these parameters, the present invention effectively improves the performance of the organic electroluminescent device. By establishing a mathematical model of the vacuum evaporation process, the present invention can accurately simulate the interaction and influence between various parameters during the evaporation process, so as to find the parameter combination that makes the device performance optimal. By directly applying the optimized process parameters to the production equipment, the present invention significantly reduces the number of trial-and-error times during the production process and improves the production efficiency. The real-time data collection and error analysis functions enable problems in the production process to be discovered and solved in a timely manner, avoiding production delays and waste caused by improper parameter settings. By optimizing the production process parameters, the present invention reduces the consumption of raw materials and energy, thereby reducing the production cost. It reduces rework and scrap caused by unqualified device performance, further reducing the production cost.

[0016] By establishing a parameter model knowledge base, the present invention establishes a corresponding relationship between the optimized process parameters and the corresponding error data characteristics, and matches and adjusts the parameters in real time during the production process, ensuring the stability of the production process. The real-time monitoring and data analysis functions enable the timely discovery and handling of abnormal situations in the production process, avoiding production interruptions and quality fluctuations. The present invention provides a new idea and method for the optimization of the production process of organic electroluminescent devices, promoting technological innovation and industrial upgrading in this field.

[0017] In summary, through the application of real-time monitoring, data analysis, model establishment, and optimization algorithms, the present invention realizes the precise control and optimization of key parameters in the production process of organic electroluminescent devices, significantly improving device performance and production efficiency, reducing production costs, and enhancing the stability of the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present invention, the following will briefly introduce the drawings required for the embodiments. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the drawings.

[0019] Figure 1 It is a schematic flow chart of the method for optimizing the production process of organic electroluminescent devices provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with the specific embodiments and corresponding drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. The following will describe the technical solutions provided by each embodiment of the present invention in conjunction with the drawings. In order to better understand the objectives of the present invention, the present invention will be further described below.

[0021] The present invention provides a method for optimizing the production process of organic electroluminescent devices, including: Step S101, obtain the types of chemical cleaning agents, the concentrations of chemical cleaning agents, the cleaning time of chemical cleaning agents, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data during the cleaning process of the organic electroluminescent conductive glass substrate. Use a preset statistical algorithm to analyze the types of chemical cleaning agents, the concentrations of chemical cleaning agents, the cleaning time of chemical cleaning agents, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data obtained during the cleaning process of the organic electroluminescent conductive glass substrate, and obtain the characteristic data affecting the substrate surface cleanliness and the characteristic data affecting the substrate surface flatness.

[0022] Chemical cleaning agent information: Record the types, concentrations, and cleaning times of the chemical cleaning agents used. Different chemical cleaning agents may have different decontamination abilities and corrosivities on the substrate surface, so detailed records are required.

[0023] Ultrasonic cleaning parameters: Record the frequency, power, and cleaning time of ultrasonic cleaning. These parameters directly affect the effect of ultrasonic cleaning and have an important impact on the cleanliness and flatness of the substrate surface.

[0024] Substrate surface cleanliness detection: After ultrasonic cleaning, use appropriate detection means (such as contact angle measurement, surface energy test, or microscopic observation, etc.) to detect the cleanliness of the substrate surface and obtain the cleanliness detection data.

[0025] Organize the collected types, concentrations, cleaning times of chemical cleaning agents, the frequency, power, cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data into a structured data set.

[0026] Preset statistical algorithm selection: Select appropriate statistical algorithms, such as analysis of variance (ANOVA), regression analysis (such as linear regression, multiple regression), and correlation analysis (such as Pearson correlation coefficient), etc., to analyze the organized data set.

[0027] Cleanliness characteristic data: By analyzing the relationship between the types, concentrations, cleaning times of chemical cleaning agents and the substrate surface cleanliness, extract the key characteristic data affecting the substrate surface cleanliness. For example, certain cleaning agents may show better cleaning effects at specific concentrations and cleaning times.

[0028] Flatness characteristic data: By analyzing the relationship between the frequency, power, cleaning time of ultrasonic cleaning and the substrate surface flatness, extract the characteristic data affecting the substrate surface flatness. Improper settings of the power and frequency of ultrasonic waves may cause microscopic unevenness on the substrate surface, affecting the deposition quality of subsequent thin films.

[0029] Result verification: Verify the extracted feature data to ensure that it can accurately reflect the cleanliness and flatness of the substrate surface. Verification can be carried out through comparative experiments or actual production data.

[0030] Output: A set of feature data affecting the cleanliness of the substrate surface. A set of feature data affecting the flatness of the substrate surface.

[0031] Step S102, Collect in real time the vacuum degree, evaporation rate of the organic functional layer material, and thin film thickness parameter data in the evaporation chamber. Based on the collected vacuum degree, evaporation rate of the organic functional layer material, and thin film thickness parameter data in the evaporation chamber, establish a mathematical model of the vacuum evaporation process, and use an optimization algorithm to solve the mathematical model of the vacuum evaporation process to obtain the combination of vacuum degree, evaporation rate, and thin film thickness parameters with the optimal device performance. Real-time monitoring: During the evaporation process, monitor in real time the key parameters such as the vacuum degree, evaporation rate of the organic functional layer material, and thin film thickness in the evaporation chamber. This can be achieved through high-precision sensors and an online monitoring system.

[0032] Data storage: Store the collected data in chronological order or batches for subsequent analysis and processing.

[0033] Mathematical model establishment: Model selection: According to the physical principles and empirical knowledge of the evaporation process, select a suitable mathematical model to describe the relationship between the vacuum degree, evaporation rate, and thin film thickness. This may include mass conservation equations, fluid dynamics equations, etc.

[0034] Model construction: Based on the collected data, use statistical methods or physical principles to construct a mathematical model of the vacuum evaporation process. The model should be able to accurately reflect the interaction and influence between various parameters during the evaporation process.

[0035] Application of optimization algorithm: Optimization goal setting: Clearly define the optimization goal, that is, to find the combination of vacuum degree, evaporation rate, and thin film thickness parameters that makes the device performance (such as brightness, efficiency, lifespan, etc.) optimal.

[0036] Optimization algorithm selection: According to the characteristics of the mathematical model and the complexity of the optimization goal, select a suitable optimization algorithm for solution. This may include genetic algorithms, particle swarm optimization algorithms, simulated annealing algorithms, etc.

[0037] Parameter solution: Use the selected optimization algorithm to solve the mathematical model to obtain the combination of vacuum degree, evaporation rate, and thin film thickness parameters that makes the device performance optimal.

[0038] Result verification: Simulation verification: Verify the optimized parameter combinations through a preset simulation model and evaluate their effects in actual production. The simulation model should be able to simulate the evaporation process and predict device performance.

[0039] Experimental verification (optional): If conditions permit, actual production experiments can be conducted to further verify the accuracy of the optimization results.

[0040] Output: The parameter combinations of the optimal vacuum degree, evaporation rate, and film thickness for device performance. The optimized parameter combinations will serve as the basis for evaluating device performance and optimizing the production process in subsequent steps (such as step S103). If the verification results do not meet the preset standards, the mathematical model or optimization algorithm needs to be adjusted and the solution and verification need to be performed again.

[0041] In step S103, collect the electrical performance data, optical performance data, and life stability detection data of the device. Evaluate the device performance based on the collected electrical performance data, optical performance data, and life stability detection data of the device to obtain the data characteristics affecting the device performance. Simulate and calculate the data characteristics affecting the device performance through the vacuum evaporation process data processing model to obtain the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter; Electrical performance data: Collect electrical performance data such as current, voltage, and resistance of the device. These data reflect the electrical characteristics of the device.

[0042] Optical performance data: Collect optical performance data such as luminous brightness, chromaticity coordinates, and luminous efficiency of the device. These data reflect the optical characteristics of the device.

[0043] Life stability data: Obtain the life stability data of the device through accelerated aging tests or other methods, including half-life, working life, etc.

[0044] Device performance evaluation: Data analysis: Conduct statistical analysis on the collected electrical performance data, optical performance data, and life stability detection data, and calculate statistical quantities such as the average value and standard deviation of each performance parameter to evaluate the overall performance of the device.

[0045] Performance evaluation: Evaluate the device performance according to the preset performance indicators and evaluation criteria. If the device performance does not meet the requirements, the factors affecting the performance need to be further analyzed.

[0046] Extraction of data characteristics affecting device performance: Feature identification: By analyzing the device performance data, identify the key factors or data characteristics affecting the device performance. These characteristics may be related to parameters such as vacuum degree, evaporation rate, and film thickness.

[0047] Feature extraction: Use data analysis methods (such as principal component analysis, factor analysis, etc.) to extract the main data features that affect the device performance, providing a basis for subsequent optimization calculations.

[0048] Application of the data processing model for the vacuum evaporation process: Model construction: Based on the physical principles of the vacuum evaporation process and the established mathematical model, construct a data processing model. This model should be able to receive the data features that affect the device performance as inputs and output adjustment suggestions for the vacuum degree, evaporation rate, and film thickness parameters.

[0049] Simulation calculation: Input the data features that affect the device performance extracted into the data processing model. The model performs simulation calculations based on the input data to obtain the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters.

[0050] Result output: Organize the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters obtained from the simulation calculations into a data table or report for use in subsequent steps. The optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters. The optimized parameters will be used as the basis for setting the operating parameters of the organic electroluminescent device production equipment in subsequent steps (such as step S104). If the optimized parameters cannot achieve the expected effect in actual production, it is necessary to return to step S102 or re - perform data collection and analysis to adjust the mathematical model and optimization algorithm.

[0051] Step S104: Send the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters to the organic electroluminescent device production equipment, and collect the operating data of the organic electroluminescent device production equipment after executing the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters to obtain the real - time data of the organic electroluminescent device production equipment; Parameter sending and equipment setting: Send the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters to the organic electroluminescent device production equipment through the control system or data transmission interface. Set or select the received optimized parameters on the production equipment to ensure that the equipment can produce according to the optimized parameters.

[0052] Start production: Start the production equipment to start producing organic electroluminescent devices. During the production process, the equipment will perform process operations such as vacuum evaporation according to the optimized parameters.

[0053] Real - time data collection: During the production process, monitor the operating data of the organic electroluminescent device production equipment in real - time. These data include but are not limited to key parameters such as vacuum degree, evaporation rate, and film thickness. Use devices such as sensors and data acquisition cards to transmit the real - time operating data to the data processing system or storage medium for subsequent analysis.

[0054] Compare the collected real-time operation data with the optimized parameters to verify whether the production equipment has executed the optimized parameters as expected.

[0055] If there is a deviation between the real-time operation data and the optimized parameters, it is necessary to analyze the cause of the deviation and consider whether it is necessary to fine-tune or re-optimize the optimized parameters.

[0056] The real-time operation data of the organic electroluminescent device production equipment includes key parameters such as vacuum degree, evaporation rate, and film thickness. The real-time operation data will be used as an important basis for evaluating the optimization effect and adjusting the optimized parameters in subsequent steps (such as step S105). If the real-time operation data indicates that the optimization effect is not ideal, it is necessary to return to step S102 or S103 to adjust and optimize the mathematical model, optimization algorithm, or data processing model.

[0057] Step S105: Compare the real-time data of the organic electroluminescent device production equipment with the parameter combination of the optimal vacuum degree, evaporation rate, and film thickness of the device performance. If there is an error in the comparison result, obtain the error data information, extract the data characteristics of the error data to obtain the error data characteristics, use the stochastic gradient descent method to optimize the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter again to obtain the real-time optimized parameter data for preparing the organic electroluminescent device, establish a corresponding relationship between the real-time optimized parameter data for preparing the organic electroluminescent device and the error data characteristics, and establish a label to store it in the parameter model knowledge base. Match the real-time error data characteristics in the parameter model knowledge base. If the match is successful, retrieve the corresponding real-time optimized parameter data for preparing the organic electroluminescent device in the parameter model knowledge base, and use the retrieved corresponding real-time optimized parameter data for preparing the organic electroluminescent device in the parameter model knowledge base as the real-time operation parameters of the organic electroluminescent device production process.

[0058] Real-time data comparison: Compare the real-time data (including vacuum degree, evaporation rate, film thickness, etc.) of the organic electroluminescent device production equipment with the optimal parameter combination of the device performance. Calculate the error value between the real-time data and the optimal parameter combination, and evaluate whether the operation status of the production equipment and the device performance meet the expectations.

[0059] Error data analysis: When there is an error in the comparison result, analyze the error data information in detail, including characteristics such as the magnitude, direction, and distribution of the error. Determine whether the error is caused by factors such as production equipment failure, raw material change, and environmental fluctuation.

[0060] Error data feature extraction: Use data analysis methods (such as statistical analysis, machine learning algorithms, etc.) to extract the features of the error data. Extract data features that can reflect the essence of the error, such as the mean, variance, and trend of the error.

[0061] Parameter re - optimization: Use the stochastic gradient descent method (or other optimization algorithms) to re - optimize the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter. Take the error data characteristics as the optimization objective or constraint condition, and adjust the parameter values to reduce the error and improve the device performance.

[0062] Establish a corresponding relationship between the organic electroluminescent device fabrication parameter data obtained by real - time optimization and the error data characteristics. Assign a unique label to each corresponding relationship, and store the label and the corresponding parameter data in the parameter model knowledge base.

[0063] During the production process, collect the error data characteristics in real - time and perform matching in the parameter model knowledge base. If the matching is successful, retrieve the corresponding real - time optimized organic electroluminescent device fabrication parameter data in the parameter model knowledge base. Use the retrieved parameter data as the real - time operating parameters of the organic electroluminescent device production process, and adjust the production equipment to optimize the device performance.

[0064] Real - time optimize the organic electroluminescent device fabrication parameter data. The parameter model knowledge base contains the corresponding relationship and label between the error data characteristics and the optimized parameters. The real - time operating parameters will be directly applied to the production process of the organic electroluminescent device to improve the device performance and production efficiency. If the matching is unsuccessful or the error persists, it is necessary to return to steps S102 to S104 to review and adjust the processes of data collection, performance evaluation, parameter optimization, etc. Regularly update the parameter model knowledge base to include new corresponding relationships between error data characteristics and optimized parameters, and improve the matching accuracy and optimization effect.

[0065] Specifically, for the organic electroluminescent device production process optimization method described in the present invention, step S101 includes: Substrate surface cleanliness detection: After ultrasonic cleaning, detect the cleanliness of the substrate surface to obtain cleanliness detection data; Organize the obtained types, concentrations, and cleaning times of chemical cleaning agents, frequencies, powers, and cleaning times of ultrasonic cleaning, as well as the substrate surface cleanliness detection data into a data set; Use a preset statistical algorithm to analyze the data set. The preset statistical algorithms include analysis of variance, regression analysis, and correlation analysis. According to the data analysis results, extract the characteristic data affecting the substrate surface cleanliness and the characteristic data affecting the substrate surface flatness.

[0066] Through the substrate surface cleanliness detection, combined with the parameters of chemical cleaning and ultrasonic cleaning, use a statistical algorithm to analyze the data set, and extract the characteristic data affecting the substrate surface cleanliness and flatness, providing a basis for the subsequent optimization of the organic electroluminescent device production process.

[0067] Substrate surface cleanliness detection: After ultrasonic cleaning, use appropriate detection tools (such as optical microscopes, electron microscopes, or surface contamination detectors) to detect the cleanliness of the substrate surface. Obtain cleanliness detection data, including but not limited to indicators such as the type, quantity, distribution of surface residues, and surface roughness.

[0068] Cleaning parameter recording: Record the type, concentration, and cleaning time of the chemical cleaning agent. This information is crucial for understanding the cleaning effect because different combinations of cleaning agents, concentrations, and times may have different effects on the substrate surface. Record the frequency, power, and cleaning time of ultrasonic cleaning. The parameter settings of ultrasonic cleaning directly affect the cleaning efficiency and the physical properties of the substrate surface.

[0069] Dataset collation: Collate the obtained type, concentration, and cleaning time of the chemical cleaning agent, the frequency, power, and cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data into a dataset.

[0070] Data analysis: Use preset statistical algorithms to analyze the dataset. Statistical algorithms include but are not limited to analysis of variance, regression analysis, and correlation analysis.

[0071] Analysis of variance: Used to compare the effects of different cleaning agents, concentrations, cleaning times, and ultrasonic parameters on the substrate surface cleanliness, and to determine which factors have a significant impact on cleanliness.

[0072] Regression analysis: Establish a mathematical model between the cleanliness detection data and the cleaning parameters, and use it to predict the cleanliness of the substrate under different cleaning conditions.

[0073] Correlation analysis: Analyze the correlation between the cleaning parameters and the substrate surface cleanliness and flatness, and identify the key parameters that affect the substrate surface characteristics.

[0074] Characteristic data extraction: According to the data analysis results, extract the characteristic data that affect the substrate surface cleanliness. These characteristic data may include specific types of cleaning agents, concentration ranges, cleaning times, and combinations of ultrasonic frequencies and powers. At the same time, extract the characteristic data that affect the substrate surface flatness. Flatness is crucial for the performance of organic light-emitting devices, so identifying the key parameters that affect flatness is crucial for optimizing the production process. Output the set of characteristic data that affect the substrate surface cleanliness and flatness.

[0075] The extracted characteristic data will be used as the basis for optimizing the cleaning process parameters and selecting appropriate substrate treatment methods in subsequent steps (such as step S102). If the characteristic data indicates that the current cleaning process is insufficient, it is necessary to adjust the type, concentration, cleaning time, or ultrasonic parameters of the cleaning agent, and repeat step S101 to verify the improvement effect.

[0076] Specifically, for the method for optimizing the production process of the organic electroluminescent device of the present invention, step S102 includes: Based on the collected data, establish a mathematical model of the vacuum evaporation process, and obtain the relationship between the vacuum degree, evaporation rate, and film thickness through the mathematical model of the vacuum evaporation process; Use an optimization algorithm to solve the established mathematical model of the vacuum evaporation process, determine the optimization objective, and the optimization objective is to find the parameter combination of the vacuum degree, evaporation rate, and film thickness that makes the device performance optimal; Verify the optimized parameter combination, and evaluate the effect of the optimized parameter combination in actual production through a preset simulation model; If the verification result meets the preset standard, then use the optimized parameter combination of the vacuum degree, evaporation rate, and film thickness as the final production parameters.

[0077] Establish a mathematical model of the vacuum evaporation process: Based on the data collected during the vacuum evaporation process (such as vacuum degree, evaporation rate, film thickness, and related process parameters), use mathematical methods and physical principles to establish a mathematical model describing the vacuum evaporation process. The model should be able to accurately reflect the interaction relationship between the vacuum degree, evaporation rate, and film thickness, as well as their influence on the device performance.

[0078] Determine the optimization objective: Clearly define the optimization objective as finding the parameter combination of the vacuum degree, evaporation rate, and film thickness that makes the device performance optimal. The device performance may include indicators such as luminous efficiency, color purity, and stability. Select appropriate performance indicators as the optimization objective according to specific requirements.

[0079] Use an optimization algorithm to solve the mathematical model: Select a suitable optimization algorithm (such as genetic algorithm, particle swarm optimization algorithm, simulated annealing algorithm, etc.) to solve the established mathematical model. Search the parameter space through the optimization algorithm to find the parameter combination of the vacuum degree, evaporation rate, and film thickness that makes the optimization objective reach the optimal.

[0080] Verify the optimized parameter combination: Use the preset simulation model, input the optimized parameter combination into the simulation model, and evaluate its effect in actual production. The simulation model should be able to simulate the vacuum evaporation process and predict the performance indicators of the device.

[0081] Compare the simulation results with the preset standard to evaluate whether the optimized parameter combination meets the requirements. If the verification result meets the preset standard, that is, the device performance indicators predicted by the simulation reach or exceed the expectations, then use the optimized parameter combination of the vacuum degree, evaporation rate, and film thickness as the final production parameters. If the verification result does not meet the preset standard, it is necessary to return to step 3, adjust the optimization algorithm or mathematical model, and re-optimize and verify.

[0082] Apply the finally determined production parameters to actual production, monitor the performance indicators of the device, and collect production data. Continuously optimize and adjust the production parameters based on the production data and device performance indicators to improve production efficiency and device performance.

[0083] Output the optimized combination of vacuum degree, evaporation rate, and film thickness parameters. The optimized parameter combination will serve as the basis for production equipment settings, real-time data collection, and comparison in subsequent steps (such as steps S103, S104, etc.). If it is found in actual production that the device performance does not meet the expectations, it is necessary to return to step S102 to review and adjust the mathematical model, optimization algorithm, or simulation model.

[0084] Specifically, for the method for optimizing the production process of an organic electroluminescent device according to the present invention, step S103 includes: Conduct statistical analysis on the collected electrical performance data, optical performance data, and life stability detection data, and calculate the average value and standard deviation statistic of each performance parameter; Based on the established mathematical model of the vacuum evaporation process, construct a data processing model. The data processing model is used to receive the data characteristics affecting the device performance as input and output adjustment suggestions for the vacuum degree, evaporation rate, and film thickness parameters; Input the extracted data characteristics affecting the device performance into the data processing model. The data processing model performs simulation calculations based on the input data to obtain the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter; Organize the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter obtained from the simulation calculation to form a data table.

[0085] By conducting statistical analysis on the collected electrical performance data, optical performance data, and life stability detection data, and combining with the established mathematical model of the vacuum evaporation process, construct a data processing model to provide adjustment suggestions for the vacuum degree, evaporation rate, and film thickness parameters, thereby optimizing the performance of the organic electroluminescent device.

[0086] Conduct statistical analysis on the collected electrical performance data (such as current density, voltage, luminous efficiency, etc.), optical performance data (such as emission wavelength, color purity, brightness, etc.), and life stability detection data (such as life time, attenuation rate, etc.).

[0087] Calculate the average value of each performance parameter to understand the overall performance level. Calculate the standard deviation statistic of each performance parameter to evaluate the fluctuation range and stability of the performance.

[0088] Construct a data processing model: Based on the established mathematical model of the vacuum evaporation process and combined with the results of performance data analysis, construct a data processing model. The data processing model should be able to receive data characteristics affecting device performance (such as the average value, standard deviation, etc. of performance parameters) as inputs. The data processing model should be able to output adjustment suggestions for vacuum degree, evaporation rate, and thin film thickness parameters to optimize device performance.

[0089] Perform simulation calculations on input data characteristics: Input the extracted data characteristics affecting device performance (such as the average value, standard deviation, etc. of performance parameters) into the data processing model. The data processing model performs simulation calculations based on the input data, considering the relationships between vacuum degree, evaporation rate, thin film thickness parameters, and device performance. Through simulation calculations, obtain optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters to improve the performance of the device.

[0090] Organize the optimized parameters into a data table: Organize the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters obtained from the simulation calculations. Form a data table to facilitate parameter setting and adjustment in subsequent steps. The data table should include information such as the optimized parameter values, corresponding performance expectations, and possible adjustment ranges.

[0091] In actual applications, apply the optimized parameters to production trials, and conduct performance tests on the trial-produced devices. Compare the performance of the trial-produced devices with the expected performance to verify the effectiveness of the optimization suggestions. If the performance of the trial-produced devices meets or exceeds the expectations, adopt the optimized parameters as the formal production parameters. If the performance of the trial-produced devices does not meet the expectations, it is necessary to return to Step 2 or Step 3 to review and adjust the data processing model or input data characteristics. Output a data table of the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters. A performance test report of the trial-produced devices and an evaluation report of the optimization suggestions. The data table of the optimized parameters will serve as the basis for production equipment adjustment, real-time data collection, and comparison in subsequent steps (such as Step S104, S105, etc.). If it is found in actual production that the device performance does not meet the expectations, it is necessary to return to Step S103 to review and adjust the data processing model, input data characteristics, or simulation calculation process.

[0092] Specifically, for the method for optimizing the production process of the organic electroluminescent device described in the present invention, Step S104 includes: Set or select and use the received optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters on the production equipment, start the production equipment, and begin to produce organic electroluminescent devices; During the production process, real-time monitor the operating data of the organic electroluminescent device production equipment. The operating data of the organic electroluminescent device production equipment includes vacuum degree, evaporation rate, and thin film thickness parameters; Compare the operation data of the organic electroluminescent device production equipment with the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters to verify whether the production equipment executes the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters as expected.

[0093] Apply the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters to the production equipment, start producing organic electroluminescent devices, and monitor the operation data of the production equipment in real time to verify whether the production equipment executes the optimized parameters as expected and ensure the stability of the production process and the consistency of device performance.

[0094] Set or select the optimized parameters: According to the data sheet of the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters provided in step S103, set or select the corresponding parameter values on the production equipment.

[0095] Start the production equipment: After confirming that the production equipment has been correctly set with the optimized parameters, start the production equipment to start producing organic electroluminescent devices. Monitor the startup process of the production equipment to ensure that the equipment operates normally without any abnormal conditions.

[0096] Monitor the operation data in real time: During the production process, use the monitoring system of the production equipment or external data acquisition equipment to monitor the operation data of the organic electroluminescent device production equipment in real time. The operation data should include key parameters such as vacuum degree, evaporation rate, and thin film thickness, as well as other relevant data that may affect device performance (such as temperature, pressure, etc.).

[0097] Compare the operation data monitored in real time with the optimized vacuum degree parameters, evaporation rate parameters, and thin film thickness parameters. Analyze the differences between the data and evaluate whether the production equipment executes the optimized parameters as expected. If the data is consistent or the differences are within an acceptable range, it indicates that the production equipment operates normally and the optimized parameters are effectively executed.

[0098] If significant differences or abnormal conditions are found during the data comparison process, immediately stop the production equipment and check the reasons for equipment failures or incorrect parameter settings. According to the inspection results, perform necessary repairs or adjustments on the production equipment to ensure that the equipment can operate normally according to the optimized parameters. After the problem is solved, restart the production equipment and continue to monitor the operation data until it is confirmed that the equipment operates stably and the optimized parameters are effectively executed.

[0099] Record the real-time monitoring data, comparison results, and any abnormal handling situations during the production process. Analyze these data to evaluate the stability of the production process, the consistency of device performance, and the effectiveness of optimized parameters. Based on the analysis results, provide data support and improvement suggestions for the subsequent optimization of the production process. Output the real-time monitoring data record during the production process. The real-time monitoring data and comparison verification results will be used as the basis for product quality assessment, production process optimization, and equipment maintenance in subsequent steps (such as steps S105, S106, etc.). If it is found during the production process that the equipment cannot operate according to the optimized parameters or the device performance does not meet the expectations, it is necessary to return to step S104 or previous steps to review and adjust the equipment settings, parameter optimization, or monitoring methods.

[0100] Specifically, for the method for optimizing the production process of the organic electroluminescent device according to the present invention, step S105 includes: The data features extracted from the error data include the change trend, fluctuation range, and occurrence frequency of the error; Preprocess the extracted data features and optimize them using the stochastic gradient descent method; Based on the extracted error data features, construct an optimization model. The optimization model takes the optimized vacuum degree parameter, evaporation rate parameter, and film thickness parameter as inputs; Use the stochastic gradient descent method to train the optimization model, adjust the vacuum degree parameter, evaporation rate parameter, and film thickness parameter to minimize the error, and obtain the real-time optimized preparation parameter data of the organic electroluminescent device.

[0101] Establish a corresponding relationship between the real-time optimized preparation parameter data of the organic electroluminescent device and the error data features.

[0102] Store information such as the parameter data of the corresponding relationship and the error data features in the parameter model knowledge base, and establish a label for each record; During the production process, real-time collect the error data features and match them in the parameter model knowledge base; If the match is successful, that is, a record corresponding to the real-time error data features is found, then retrieve the corresponding real-time optimized preparation parameter data of the organic electroluminescent device in the record; Use the retrieved real-time optimized preparation parameter data of the organic electroluminescent device as the real-time operating parameters of the production process.

[0103] By analyzing the error data in the production process, extracting data features, constructing an optimization model, and using the stochastic gradient descent method to train the model, the preparation parameters of organic light-emitting devices are optimized in real time to improve production efficiency and device performance. At the same time, a corresponding relationship is established between the optimized parameter data and the error data features and stored in the parameter model knowledge base to provide a basis for real-time parameter adjustment in subsequent production processes.

[0104] Extract key data features from the error data in the production process, including the change trend, fluctuation range, and occurrence frequency of errors. Preprocess the extracted error data features, including data cleaning, normalization, denoising, etc., to improve the quality and usability of the data. Use the stochastic gradient descent method to preliminarily optimize the data features to prepare for constructing an optimization model.

[0105] Construct an optimization model: Based on the extracted error data features, construct an optimization model. This model takes the optimized vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter as inputs and outputs with the goal of minimizing errors. The optimization model should be able to consider the interaction relationships between parameters and their impacts on device performance.

[0106] Train the optimization model: Use the stochastic gradient descent method to train the optimization model. By continuously adjusting the vacuum degree parameter, evaporation rate parameter, and thin film thickness parameter, minimize the errors output by the model. During the training process, monitor the convergence situation and performance indicators of the model to ensure the effectiveness and stability of the model.

[0107] Obtain the real-time optimized preparation parameter data: After training, the optimization model can output the real-time optimized preparation parameter data of organic light-emitting devices.

[0108] Establish a corresponding relationship and store it in the knowledge base: Establish a corresponding relationship between the real-time optimized preparation parameter data of organic light-emitting devices and the error data features. Store information such as the parameter data of the corresponding relationship and the error data features in the parameter model knowledge base, and establish labels for each record for easy subsequent retrieval and matching.

[0109] During the production process, real-time collect error data features and perform matching in the parameter model knowledge base. During the matching process, consider the similarities and differences of error data features to ensure the accuracy and reliability of the matching.

[0110] If the matching is successful, that is, a record corresponding to the real-time error data features is found, then retrieve the corresponding real-time optimized preparation parameter data of organic light-emitting devices in that record. The retrieved parameter data should be verified and confirmed to ensure its effectiveness and usability.

[0111] Use the retrieved real-time optimized organic electroluminescent device fabrication parameter data as the real-time operating parameters of the production process.

[0112] Real-time adjust the parameter settings of the production equipment to ensure that the production process proceeds according to the optimized parameters, improving the device performance and production efficiency. Output the real-time optimized organic electroluminescent device fabrication parameter data. The parameter model knowledge base contains information such as parameter data with corresponding relationships, error data characteristics, etc. Records of the real-time adjusted production process operating parameters.

[0113] The real-time optimized fabrication parameter data will be used as the basis for adjusting the production equipment, evaluating product quality, and optimizing the production process in subsequent steps (such as steps S106, S107, etc.).

[0114] If it is found during the production process that the real-time optimized parameters cannot effectively improve the device performance or production efficiency, it is necessary to return to step S105 or previous steps to review and adjust the error data feature extraction, optimization model construction, or training method.

[0115] Specifically, for the organic electroluminescent device production process optimization method of the present invention, step S105 includes: When using the stochastic gradient descent method to optimize the parameters in the organic electroluminescent device production process, the specific steps are as follows: Receive and set initial parameters, set initial values for the vacuum degree, evaporation rate, and thin film thickness parameters; Receive and set the learning rate, construct a loss function, and according to the optimization goal of the production process, construct a loss function for the relationship between device performance and parameters. Calculate the value of the loss function using the parameter values, that is, the performance deviation under the production conditions; Generate randomly selected samples, and based on the randomly selected samples, calculate the gradient of the loss function with respect to the parameters, that is, the partial derivative of the loss value with respect to the parameters; Update the parameters using the learning rate and gradient value, that is, adjust the parameter values in the opposite direction of the gradient, and repeat the iteration until the preset number of iterations is reached, the loss value converges to an acceptable range, or the stop condition is met.

[0116] Through the stochastic gradient descent method, optimize the parameters such as the vacuum degree, evaporation rate, and thin film thickness in the organic electroluminescent device production process to minimize the deviation between the device performance and the production conditions (i.e., the loss function value), thereby improving the device performance and production efficiency.

[0117] Receive and set initial parameters: Set initial values for parameters such as the vacuum degree, evaporation rate, and thin film thickness. These initial values can be based on experience, historical data, or pre-experiments.

[0118] Receive and set the learning rate: Set the learning rate, which is an important parameter controlling the step size of parameter updates. A too large learning rate may lead to instability in the optimization process, while a too small learning rate may result in an overly slow convergence speed. Select an appropriate learning rate value according to the complexity of the problem and the sensitivity of the parameters.

[0119] Construct the loss function: Based on the optimization goal of the production process, construct a loss function that represents the relationship between device performance and parameters. The loss function should be able to reflect the deviation between device performance and production conditions.

[0120] Calculate the value of the loss function using the parameter values, i.e., the performance deviation under the current production conditions. This can be obtained through experimental data, simulation results, or theoretical calculations.

[0121] Generate randomly selected samples: During the optimization process, to improve efficiency and reduce computational complexity, randomly selected samples are usually used to calculate the gradient. Generate a set of randomly selected samples, which can be randomly drawn from historical data or generated through a certain sampling strategy.

[0122] Calculate the gradient of the loss function with respect to the parameters: Based on the randomly selected samples, calculate the gradient of the loss function with respect to the parameters, i.e., the partial derivative of the loss value with respect to the parameters. This can be obtained through methods such as numerical differentiation, analytical differentiation, or automatic differentiation. The gradient reflects the sensitivity of the loss function to parameter changes and is an important basis for parameter updates.

[0123] Update the parameters using the learning rate and gradient values: Adjust the parameter values in the opposite direction of the gradient, i.e., update the parameters in the opposite direction of the gradient to reduce the value of the loss function. The update formula is usually: new parameter value = old parameter value - learning rate * gradient value.

[0124] Repeat the iteration: Repeat the above steps, i.e., calculate the loss function value, generate randomly selected samples, calculate the gradient, and update the parameters, until a preset number of iterations is reached, the loss value converges to an acceptable range, or a stopping condition is met. During the iteration process, the change of the loss function value can be monitored to evaluate the progress and effectiveness of the optimization process.

[0125] Judge the stopping condition: Set the stopping condition, such as the loss value converges below a certain threshold, the number of iterations reaches the upper limit, or the parameter update amount is less than a certain threshold, etc. When the stopping condition is met, end the optimization process and output the optimized parameter values. Output the optimized parameter values such as vacuum degree, evaporation rate, film thickness, etc. Records of the change of the loss function value during the optimization process. Information on the number of iterations and the satisfaction of the stopping condition.

[0126] The optimized parameter values will be used as production process parameters in subsequent steps (such as step S106, production implementation, etc.) to guide the actual production process. If the optimization result does not meet the expectations or further improvement of device performance is required, it is possible to return to step S105 or previous steps to review and adjust the initial parameters, learning rate, loss function, or iteration process.

[0127] Aiming at the problem that the prior art cannot accurately simulate and optimize parameters such as vacuum degree, evaporation rate, and film thickness in the vacuum evaporation process, the present invention provides a method for optimizing the production process of organic electroluminescent devices. This method realizes the precise control and optimization of key parameters in the vacuum evaporation process through real-time monitoring, data analysis, model establishment, and the application of optimization algorithms, thereby improving device performance and production efficiency.

[0128] Obtain the types, concentrations, cleaning times of chemical cleaning agents during the cleaning process of organic electroluminescent conductive glass substrates, the frequencies, powers, cleaning times of ultrasonic cleaning, and the substrate surface cleanliness detection data after ultrasonic cleaning. Use preset statistical algorithms (such as analysis of variance, regression analysis, correlation analysis) to analyze these data and extract the characteristic data affecting the substrate surface cleanliness and flatness. Real-time collect the vacuum degree in the evaporation chamber, the evaporation rate of the organic functional layer material, and the film thickness parameter data. Based on these data, establish a mathematical model of the vacuum evaporation process to describe the relationship between the vacuum degree, evaporation rate, and film thickness. Use optimization algorithms (such as genetic algorithm, particle swarm optimization algorithm, simulated annealing algorithm) to solve the model and find the combination of vacuum degree, evaporation rate, and film thickness parameters that makes the device performance optimal.

[0129] Collect the electrical performance data, optical performance data, and lifetime stability detection data of the device to evaluate the device performance. Extract the data characteristics affecting the device performance and perform simulation calculations through the vacuum evaporation process data processing model to obtain the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters. Send the optimized parameters to the organic electroluminescent device production equipment and collect the real-time operation data of the equipment after executing these parameters. Compare the real-time operation data with the optimized parameters to verify whether the production equipment executes as expected. Compare the real-time operation data with the parameter combination with the optimal device performance. If there is an error, extract the error data characteristics (such as the change trend, fluctuation range, occurrence frequency of the error). Use the stochastic gradient descent method to optimize the optimized parameters again to obtain the real-time optimized preparation parameter data. Establish a corresponding relationship between the optimized parameter data and the error data characteristics and store them in the parameter model knowledge base. During the production process, real-time collect the error data characteristics and perform matching in the knowledge base. If the matching is successful, retrieve the corresponding real-time optimized parameters as the real-time operation parameters of the production process.

[0130] Through the application of real-time monitoring, data analysis, model establishment, and optimization algorithms, the present invention realizes the precise control and optimization of the key parameters in the vacuum evaporation process of the organic electroluminescent device production process. This method effectively solves the problem that the prior art cannot accurately simulate and optimize parameters such as vacuum degree, evaporation rate, and film thickness, improves device performance and production efficiency, reduces production costs, and has significant economic and social benefits.

Claims

1. A method for optimizing the production process of an organic electroluminescent device, characterized in that: include: Step S101, obtaining the type of chemical cleaning agent, the concentration of chemical cleaning agent, the cleaning time of chemical cleaning agent, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data after ultrasonic cleaning during the cleaning process of the organic electroluminescent conductive glass substrate, and using a preset statistical algorithm to analyze the type of chemical cleaning agent, the concentration of chemical cleaning agent, the cleaning time of chemical cleaning agent, the frequency of ultrasonic cleaning, the power of ultrasonic cleaning, the cleaning time of ultrasonic cleaning, and the substrate surface cleanliness detection data after ultrasonic cleaning, to obtain characteristic data affecting the substrate surface cleanliness and characteristic data affecting the substrate surface flatness; Step S102, collecting the vacuum degree in the evaporation chamber, the evaporation rate of the organic functional layer material, and the film thickness parameter data in real time, establishing a mathematical model of the vacuum evaporation process based on the collected vacuum degree in the evaporation chamber, the evaporation rate of the organic functional layer material, and the film thickness parameter data, and solving the mathematical model of the vacuum evaporation process using an optimization algorithm to obtain the vacuum degree, evaporation rate, and film thickness parameter combination with the best device performance; Step S103, collecting electrical performance data, optical performance data and life stability test data of the device, evaluating the device performance according to the collected electrical performance data, optical performance data and life stability test data of the device, obtaining data features that affect the device performance, and simulating and calculating the data features that affect the device performance through a vacuum evaporation process data processing model to obtain optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters; Step S104, sending the optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters to the organic electroluminescent device production equipment, collecting the operation data of the organic electroluminescent device production equipment after executing the optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters, and obtaining the real-time data of the organic electroluminescent device production equipment; Step S105, compare the real-time data of the organic electroluminescent device production equipment with the vacuum degree, evaporation rate, and film thickness parameter combination with the best device performance. If there is an error in the comparison result, obtain error data information, extract data features from the error data information, obtain error data features, use the stochastic gradient descent method to optimize the optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters again, and obtain real-time optimized organic electroluminescent device preparation parameter data, establish a corresponding relationship between the real-time optimized organic electroluminescent device preparation parameter data and the error data features, establish a label and store it in the parameter model knowledge base, match the real-time error data features in the parameter model knowledge base, and if the match is successful, retrieve the corresponding real-time optimized organic electroluminescent device preparation parameter data in the parameter model knowledge base, and use the retrieved real-time optimized organic electroluminescent device preparation parameter data in the parameter model knowledge base as the real-time operating parameters of the organic electroluminescent device production process.

2. The method for optimizing the production process of an organic electroluminescent device according to claim 1, characterized in that: The step S101 includes: Cleanliness test of the substrate surface: after ultrasonic cleaning, the cleanliness test of the substrate surface is performed to obtain cleanliness test data; The acquired chemical cleaning agent type, concentration, cleaning time, ultrasonic cleaning frequency, power, cleaning time, and substrate surface cleanliness detection data are organized into a data set; The data set is analyzed using a preset statistical algorithm, which includes variance analysis, regression analysis, and correlation analysis. Based on the data analysis results, characteristic data that affects the cleanliness of the substrate surface and characteristic data that affects the flatness of the substrate surface are extracted.

3. The method for optimizing the production process of an organic electroluminescent device according to claim 1, characterized in that: The step S102 includes: Based on the collected data, a mathematical model of the vacuum evaporation process is established, and the relationship between the vacuum degree, evaporation rate and film thickness is obtained through the mathematical model of the vacuum evaporation process; The established mathematical model of vacuum evaporation process is solved by using optimization algorithm to determine the optimization target, which is to find the parameter combination of vacuum degree, evaporation rate and film thickness that makes the device performance optimal; Verify the optimized parameter combination and evaluate the effect of the optimized parameter combination in actual production through the preset simulation model; If the verification results meet the preset standards, the optimized combination of vacuum degree, evaporation rate and film thickness parameters will be used as the final production parameters.

4. The method for optimizing the production process of an organic electroluminescent device according to claim 1, characterized in that: The step S103 includes: Conduct statistical analysis on the collected electrical performance data, optical performance data and life stability test data, and calculate the average value and standard deviation statistics of each performance parameter; Based on the established mathematical model of vacuum evaporation process, a data processing model is constructed. The data processing model is used to receive data features that affect device performance as input and output adjustment suggestions for vacuum degree, evaporation rate, and film thickness parameters; The extracted data features affecting device performance are input into the data processing model, and the data processing model performs simulation calculations based on the input data to obtain optimized vacuum degree parameters, evaporation rate parameters, and film thickness parameters; The optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters obtained through simulation calculation are sorted out to form a data table.

5. The method for optimizing the production process of an organic electroluminescent device according to claim 1, characterized in that: The step S104 includes: Setting or selecting to use the received optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters on the production equipment, starting the production equipment, and starting to produce the organic electroluminescent device; During the production process, real-time monitoring of the operating data of the organic electroluminescent device production equipment, including vacuum degree, evaporation rate, and film thickness parameters; Compare the operating data of the organic electroluminescent device production equipment with the optimized vacuum parameters, evaporation rate parameters and film thickness parameters to verify whether the production equipment executes the optimized vacuum parameters, evaporation rate parameters and film thickness parameters as expected.

6. The method for optimizing the production process of an organic electroluminescent device according to claim 1, characterized in that: The step S105 includes: The data features extracted from the error data include the error change trend, fluctuation range and occurrence frequency; Preprocess the extracted data features and optimize them using stochastic gradient descent method; Based on the extracted error data features, an optimization model is constructed, and the optimization model takes the optimized vacuum degree parameters, evaporation rate parameters and film thickness parameters as input; The optimization model is trained using the stochastic gradient descent method, and the vacuum degree parameters, evaporation rate parameters, and film thickness parameters are adjusted to minimize the error, thereby obtaining real-time optimized organic electroluminescent device preparation parameter data; Establishing a corresponding relationship between the real-time optimized organic electroluminescent device preparation parameter data and the error data characteristics; The parameter data and error data characteristic information of the corresponding relationship are stored in the parameter model knowledge base, and a label is established for each record; During the production process, error data features are collected in real time and matched in the parameter model knowledge base; If the match is successful, that is, a record corresponding to the real-time error data feature is found, then the corresponding real-time optimized organic electroluminescent device preparation parameter data in the record is retrieved; The retrieved real-time optimized organic electroluminescent device preparation parameter data is used as the real-time operating parameters of the production process.

7. The method for optimizing the production process of an organic electroluminescent device according to claim 6, characterized in that: The step S105 includes: When using the stochastic gradient descent method to optimize the parameters in the production process of organic electroluminescent devices, the specific steps are as follows: Receive and set initial parameters, and set initial values ​​for vacuum degree, evaporation rate, and film thickness parameters; Receive and set the learning rate, construct the loss function, construct the loss function of the relationship between device performance and parameters according to the optimization goal of the production process, and use the parameter value to calculate the value of the loss function, that is, the performance deviation under production conditions; Generate randomly selected samples, and based on the randomly selected samples, calculate the gradient of the loss function with respect to the parameters, that is, the partial derivative of the loss value with respect to the parameters; Use the learning rate and gradient value to update the parameters, that is, adjust the parameter value in the opposite direction of the gradient, and repeat the iteration until the preset number of iterations is reached, the loss value converges to an acceptable range, or the stopping condition is met.

Images