Method and system for constructing OLED efficiency roll-off prediction model

Abstract

The present invention relates to the field of model construction technology, specifically to a method and system for constructing an OLED efficiency roll-off prediction model, comprising the following steps: extracting key performance indicators, including brightness, pixel density, temperature and voltage parameters, based on collected OLED device performance-related data, evaluating the correlation of the parameters in predicting OLED efficiency roll-off, and adjusting the data processing flow to construct a set of key performance indicators. In the present invention, by using the collected OLED device performance-related data to extract and optimize key performance indicators, the model can more accurately reflect the efficiency roll-off of OLED, integrate new usage data and analyze time series and usage frequency, improve the ability to identify factors affecting OLED life, and improve the stability of the model under various test conditions through iterative testing, thereby optimizing the reliability of the prediction results, and by collecting and processing multi-source data in real time and using a continuous feedback mechanism to optimize the efficiency of data analysis and processing.

Description

Technical Field

[0001] The present invention relates to the technical field of model construction, and in particular to a method and system for constructing an OLED efficiency roll-off prediction model. Background Art

[0002] Model building technology covers a series of methods that use statistics, mathematics, and computer science principles to create models that describe complex systems or predict future events. The core lies in developing algorithms and frameworks that can process large amounts of data and provide accurate predictions. Model building is an indispensable tool in many fields such as scientific research, engineering applications, and economic forecasting. These models usually require in-depth analysis of data and use machine learning, artificial intelligence, or traditional statistical methods to solve practical problems.

[0003] Among them, the method for constructing an OLED efficiency roll-off prediction model focuses on establishing a prediction model for the performance attenuation of OLED (organic light-emitting diode) devices. OLED technology is widely used in display screens and lighting equipment, but the problem of its efficiency gradually decreasing over time limits its service life and performance. By constructing an effective prediction model, the efficiency roll-off of OLED can be estimated in advance, thereby helping R&D personnel optimize materials and structures and extend the service life of the product. It also provides a scientific basis for quality control of the manufacturing process.

[0004] Existing technologies are not sufficient to adapt to the needs of new data types and complex data structures, lack the ability to integrate new usage data and time series analysis, and find it difficult to accurately capture the performance degradation trends of OLED devices. There is a lack of flexibility in updating and optimizing data processing processes, and there is no effective continuous feedback mechanism, which limits the adaptability and accuracy under new or changing test conditions. This limitation leads to insufficient performance of the prediction model in practical applications, which in turn affects the design and quality control decisions of OLED products and increases R&D costs. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a method and system for constructing an OLED efficiency roll-off prediction model.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for constructing an OLED efficiency roll-off prediction model, comprising the following steps:

[0007] S1: Based on the collected OLED device performance correlation data, extract key performance indicators, including brightness, pixel density, temperature and voltage parameters, evaluate the correlation of parameters in predicting OLED efficiency roll-off, and adjust the data processing flow to build a set of key performance indicators;

[0008] S2: Based on the key performance indicator set, the newly added OLED usage data is integrated, the time series and usage frequency in the data are analyzed, and the identification process of factors affecting the life of OLED is optimized through multi-cycle iterative testing to obtain iterative optimization results;

[0009] S3: Based on the iterative optimization results, collect OLED data from multiple sources, including production batches, usage environment and maintenance records, process and analyze the collected data in real time, optimize data analysis and processing efficiency through a continuous feedback mechanism, and obtain a multi-source data integration framework;

[0010] S4: Based on the multi-source data integration framework, the model is compared with the historical efficiency roll-off data to verify the prediction performance and data processing accuracy of the model, and the model parameters are adjusted again to match the brightness attenuation and life reduction trends of the OLED device to obtain the optimal roll-off prediction model.

[0011] The present invention is improved in that the evaluation step of the correlation of the efficiency roll-off is specifically as follows:

[0012] S111: extracting brightness, pixel density, temperature and voltage parameters based on the collected OLED device performance correlation data to obtain a key parameter data set;

[0013] S112: Based on the key parameter data set, the correlation with OLED efficiency roll-off is analyzed using the formula:

[0014] ;

[0015] Calculate the correlation coefficient between parameters and efficiency roll-off , and get the correlation analysis results, where Represents a single performance parameter value, represents the mean value of performance parameters, represents the efficiency roll-off value, represents the efficiency roll-off mean;

[0016] S113: According to the correlation analysis result, evaluate and determine the correlation between the performance parameter and the efficiency roll-off of the OLED to obtain a parameter evaluation result.

[0017] The present invention is improved in that the step of obtaining the key performance indicator set is specifically as follows:

[0018] S121: adjusting the data processing flow according to the parameter evaluation result, including normalizing the data set and screening parameters associated with efficiency roll-off to obtain an adjusted data processing configuration;

[0019] S122: Based on the adjusted data processing configuration, use the formula:

[0020] ;

[0021] Calculate the evaluation value , and obtain a set of key performance indicators, where It is The weight of each performance parameter, It is The quantitative value of a performance parameter, is the total number of parameters involved in the evaluation.

[0022] The present invention is improved in that the steps of analyzing the time series and the usage frequency are specifically as follows:

[0023] S211: Based on the key performance indicator set, integrating the newly added OLED usage data, including the number of startup times, the duration of each usage, and the shutdown time, to obtain a time-related performance data set;

[0024] S212: Performing time series analysis on the time-related performance data set to determine the periodicity, trend and randomness of the data and obtain a time series analysis result;

[0025] S213: Based on the time series analysis results, the formula is used:

[0026] ;

[0027] Analyze usage frequency , identifying the high-frequency and low-frequency modes used by the device, where It is time interval, is the attenuation coefficient, which indicates the weight of the influence of time interval on the usage frequency. is the number of time intervals.

[0028] The present invention is improved in that the step of obtaining the iterative optimization result is specifically as follows:

[0029] S221: Based on the time series and usage frequency, formulate a multi-cycle iterative test process, continuously adjust test parameters, simulate the response of the OLED device under different usage conditions, and obtain an iterative test configuration;

[0030] S222: Based on the iterative test configuration, perform iterative testing to monitor the performance of the OLED device under a variety of simulated usage conditions, record the response and performance changes of the device, and generate a test result data set;

[0031] S223: Analyze the test result data set, evaluate the impact of various usage modes on the life of the OLED, identify key performance degradation factors, and obtain iterative optimization results.

[0032] The present invention is improved in that the steps of real-time processing and analyzing the collected data are specifically as follows:

[0033] S311: Based on the iterative optimization results, collect OLED data from multiple sources, including production batches, usage environment and maintenance records, extract the data and unify the format to obtain a standardized preliminary data set;

[0034] S312: Based on the standardized preliminary data set, outlier detection and correction, missing value processing and data normalization are performed to obtain a ready data set;

[0035] S313: Analyze the prepared data set using the formula:

[0036] ;

[0037] Identify key trends and patterns, mine data information, and obtain key data indicators ,in, It is The weight of the data item, It is The value of the item data, is the total number of data items.

[0038] The present invention is improved in that the steps of acquiring the multi-source data integration framework are specifically as follows:

[0039] S321: extract key information from the collected data, identify the correlation and potential influencing factors between different data sources, and obtain a data source analysis record;

[0040] S322: Analyze the records based on the data source, and use the formula:

[0041] ;

[0042] Optimize data analysis and processing efficiency to obtain a multi-source data integration framework, where: Represents the analysis indicator, It is The impact factor of each data source, It is The processing results obtained from the data source are is the number of data sources.

[0043] The present invention is improved in that the steps of obtaining the optimal roll-off prediction model are specifically as follows:

[0044] S411: Based on the multi-source data integration framework, collect historical efficiency roll-off data and performance data of current OLED devices, including brightness attenuation and lifespan data, to obtain an integrated OLED data set;

[0045] S412: Based on the integrated OLED data set, perform a baseline test, calculate the error between the model prediction result and the historical data, and obtain a model prediction performance evaluation record;

[0046] S413: Based on the model prediction performance evaluation record, the model parameters are adjusted again using the formula:

[0047] ;

[0048] The model parameters are optimized according to the brightness attenuation and life reduction trend of OLED devices to obtain the optimal roll-off prediction model, where: are the adjusted model parameters, are the existing model parameters, It is The observed value of data points, It is The model prediction value for data points, It is The independent variable has data points, is the learning rate, is the total number of data points.

[0049] A system for constructing an OLED efficiency roll-off prediction model, the system comprising:

[0050] The performance index extraction module is based on the collected OLED device performance related data, including brightness, pixel density, temperature and voltage parameters, to evaluate the relevance of the parameters in predicting OLED efficiency roll-off, filter and process the data to reflect the attenuation behavior of OLED and construct a set of key performance indicators;

[0051] The iterative optimization analysis module integrates the newly added OLED usage data based on the key performance indicator set, analyzes the time series and usage frequency, optimizes the identification process of factors affecting the life of OLED through multi-cycle iterative testing, and obtains iterative optimization results;

[0052] The multi-source data processing module collects OLED data from multiple sources, including production batches, usage environment and maintenance records, based on the iterative optimization results, processes and analyzes the collected data in real time, optimizes data analysis and processing efficiency, and obtains a multi-source data integration framework;

[0053] The efficiency verification and adjustment module compares the historical efficiency roll-off data based on the multi-source data integration framework to test the prediction performance and data processing accuracy of the model and generate prediction performance verification results;

[0054] The prediction model building module uses the prediction performance verification results to adjust the model parameters again to match the brightness attenuation and life reduction trends of the OLED device to obtain the optimal roll-off prediction model.

[0055] Compared with the prior art, the advantages and positive effects of the present invention are:

[0056] In the present invention, by using the collected data related to the performance of OLED equipment to extract and optimize key performance indicators, the model can more accurately reflect the efficiency roll-off of OLED, integrate new usage data and analyze the time series and usage frequency, improve the ability to identify factors affecting the life of OLED, and improve the stability of the model under various test conditions through iterative testing, thereby optimizing the reliability of the prediction results, and by collecting and processing data from multiple sources in real time and adopting a continuous feedback mechanism to optimize the efficiency of data analysis and processing, so that the comparison between the model and historical data is more accurate, thereby improving the accuracy of predicting efficiency roll-off. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A flowchart of a method for constructing an OLED efficiency roll-off prediction model proposed by the present invention;

[0058] Figure 2 is a flow chart for evaluating the relevance of efficiency roll-off in the present invention;

[0059] Figure 3 A flowchart for obtaining a key performance indicator set in the present invention;

[0060] Figure 4 It is a flow chart of the analysis of time series and usage frequency in the present invention;

[0061] Figure 5 A flowchart for obtaining iterative optimization results in the present invention;

[0062] Figure 6 A flow chart of the present invention for real-time processing and analysis of collected data;

[0063] Figure 7 It is a flowchart of obtaining the multi-source data integration framework in the present invention;

[0064] Figure 8 This is a flow chart for obtaining the optimal roll-off prediction model in the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0066] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0067] Example: See Figure 1 The present invention provides a technical solution: a method for constructing an OLED efficiency roll-off prediction model, comprising the following steps:

[0068] S1: Based on the collected data related to OLED device performance, extract key performance indicators, including brightness, pixel density, temperature and voltage parameters, evaluate the relevance of the parameters in predicting OLED efficiency roll-off, and adjust the data processing flow to reflect the specific attenuation behavior of OLED and build a set of key performance indicators;

[0069] S2: Based on the key performance indicator set, the newly added OLED usage data is integrated, the time series and usage frequency in the data are analyzed, and the identification process of factors affecting OLED life is optimized through multi-cycle iterative testing to ensure that the efficiency roll-off of OLED can be stably predicted under new test conditions, and the iterative optimization results are obtained;

[0070] S3: Based on the iterative optimization results, collect OLED data from multiple sources, including production batches, usage environment and maintenance records, process and analyze the collected data in real time, optimize data analysis and processing efficiency through a continuous feedback mechanism, and obtain a multi-source data integration framework;

[0071] S4: Based on the multi-source data integration framework, the model’s prediction performance and data processing accuracy are verified by comparing with historical efficiency roll-off data. The model parameters are adjusted again to match the brightness attenuation and life reduction trends of OLED devices to obtain the optimal roll-off prediction model.

[0072] The key performance indicator set includes brightness measurement information, pixel stability information, temperature response information, and voltage change information. The iterative optimization results include time trend analysis results, usage frequency identification results, and stability test results. The optimal roll-off prediction model includes brightness attenuation analysis results, life prediction results, and performance verification results.

[0073] See also Figure 2 , the evaluation steps for the correlation in terms of efficiency roll-off are as follows:

[0074] S111: extracting brightness, pixel density, temperature and voltage parameters based on the collected OLED device performance correlation data to obtain a key parameter data set;

[0075] In the initial stage of OLED device performance evaluation, a variety of sensors, such as photoelectric sensors and voltage sensors, are used to monitor the device's operating status in real time and collect key parameters such as brightness, pixel density, temperature and voltage. The system performs preliminary processing on the raw data, such as removing noise and outliers, and then uses data processing technology to further analyze the data, such as using statistical methods to calculate the mean and standard deviation of each parameter. The purpose is to extract the most critical parameters for performance evaluation from massive data, provide a basis for subsequent data analysis, and ensure that the obtained key parameter data set has high accuracy and reliability.

[0076] S112: Based on the key parameter data set, analyze the correlation with OLED efficiency roll-off using the formula:

[0077] ;

[0078] Calculate the correlation coefficient between parameters and efficiency roll-off , and get the correlation analysis results, where Represents a single performance parameter value, represents the mean value of performance parameters, represents the efficiency roll-off value, represents the efficiency roll-off mean;

[0079] The sample data of brightness is ,average value , the sample data of efficiency roll-off is ,average value , substitute into the formula to calculate :

[0080] First compute the sum of the products of the differences:

[0081] ;

[0082] ;

[0083] ;

[0084] Then calculate the sum of the squares of the differences in x:

[0085] ;

[0086] The sum of squared differences is:

[0087] ;

[0088] Finally calculate the R value:

[0089] ;

[0090] This result shows that the correlation between brightness and efficiency roll-off is very weak, indicating that the effect of brightness changes on OLED efficiency roll-off is not significant and the role of other parameters needs to be considered.

[0091] S113: evaluating and determining the correlation between the performance parameter and the efficiency roll-off of the OLED according to the correlation analysis result, and obtaining a parameter evaluation result;

[0092] After the key parameter data set is obtained, the extracted parameters such as brightness, pixel density, temperature and voltage are analyzed in detail, and statistical analysis methods, such as Pearson correlation coefficient calculation, are used to evaluate the relationship between the parameters and the OLED efficiency roll-off. During this process, the strength of the correlation between each parameter and the efficiency roll-off is monitored, and hypothesis testing is performed to confirm statistical significance. This analysis helps identify the performance parameters that most significantly affect efficiency, and the parameter evaluation results obtained will directly affect future OLED improvements and quality control measures.

[0093] See also Figure 3 , the steps for obtaining the key performance indicator set are as follows:

[0094] S121: adjusting the data processing flow according to the parameter evaluation result, including normalizing the data set and screening parameters associated with efficiency roll-off to obtain an adjusted data processing configuration;

[0095] In the process of OLED device performance evaluation, precise adjustment and selection of parameters are crucial. The process begins with the normalization of all collected data, including converting the data into a format with a unified scale for effective comparison. In addition, based on previous research and analysis results related to OLED efficiency roll-off, specific parameters such as brightness, temperature and voltage are identified as highly correlated. The parameters are screened through data mining techniques such as principal component analysis and cluster analysis to ensure that only the most influential parameters are included in the analysis. After the selection and verification of the parameters are completed, each selected parameter is further tested to verify its actual impact on the attenuation of OLED performance. After completing the steps, the adjusted data processing configuration is formed, which not only includes a list of selected parameters, but also sets the specific application method of each parameter in data processing, ensuring that the resulting data set can provide accurate and reliable input for the next step of performance evaluation.

[0096] S122: Based on the adjusted data processing configuration, use the formula:

[0097] ;

[0098] Calculate the evaluation value , and obtain a set of key performance indicators, where It is The weight of each performance parameter, It is The quantitative value of a performance parameter, is the total number of parameters involved in the evaluation;

[0099] There are three parameters, brightness ( =300), temperature ( =350), voltage ( = 15), the weights are determined as , , , the calculation process of the formula is:

[0100] ;

[0101] The result shows that after integrating the three parameters, the evaluation value of the key performance indicator set is 258, which reflects the comprehensive performance evaluation of current OLED devices under the influence of specific performance parameters and guides subsequent performance optimization and parameter adjustment strategies.

[0102] See also Figure 4 , the analysis steps of time series and usage frequency are as follows:

[0103] S211: Based on the key performance indicator set, the newly added OLED usage data, including the number of startup times, the duration of each usage, and the shutdown time, are integrated to obtain a time-correlated performance data set;

[0104] Based on the existing set of key performance indicators, the latest collected OLED usage data is integrated, including the specific time records of each power on and off and the duration of each use. The data is extracted from the equipment operation log to ensure the integrity and accuracy of the data. The data is further processed, including data cleaning and verification, removal of outliers and completion of missing data, to generate a time-correlated performance data set. The data set will be used for further time series analysis to explore equipment usage patterns and potential maintenance needs.

[0105] S212: Perform time series analysis on the time-related performance data set to determine the periodicity, trend, and randomness of the data and obtain a time series analysis result;

[0106] A time series analysis is performed on the time-correlated performance data set integrated from OLED devices. First, the data set is analyzed for trend and periodicity. The autoregressive moving average (ARMA) model and seasonally decomposed time series analysis (STL) are used to identify periodicity and trends in the data. The analysis includes calculating the mean, standard deviation, autocorrelation function, and partial autocorrelation function of the data. Through this method, the randomness and periodic changes in the data set are determined, thereby obtaining a comprehensive time series analysis result. The result will provide a basis for further analysis of the frequency of OLED use and help optimize the operation and maintenance strategy of the equipment.

[0107] S213: Based on the results of time series analysis, the formula is used:

[0108] ;

[0109] Analyze usage frequency , identifying the high-frequency and low-frequency modes used by the device, where It is time interval, is the attenuation coefficient, which indicates the weight of the influence of time interval on the usage frequency. is the number of time intervals;

[0110] There are three time intervals: Hour, Hour, Hours, attenuation coefficient , then the formula is calculated as:

[0111] ;

[0112] Substitute the values ​​into:

[0113] ;

[0114] The results show that after comprehensively considering the attenuation effects of time intervals and usage frequency, the comprehensive usage frequency index of the equipment is 1.062, which helps to formulate more effective equipment operation and maintenance plans, thereby improving the use efficiency and life of the equipment.

[0115] See also Figure 5 , the specific steps for obtaining iterative optimization results are:

[0116] S221: Based on the time series and usage frequency, formulate a multi-cycle iterative test process, continuously adjust the test parameters, simulate the OLED device response under different usage conditions, and obtain the iterative test configuration;

[0117] Based on in-depth analysis of time series and usage frequency, a multi-cycle iterative testing process is developed. This process aims to simulate different usage conditions to observe the response of OLED devices. By adjusting test parameters such as temperature, usage frequency, brightness, etc., the extreme usage of the device is simulated. This is to test the stability and durability of the device under different operating conditions, thereby identifying the key variables that lead to performance degradation, including specific parameter settings, test intervals, and expected monitoring points, to ensure the comprehensiveness of the test. The generated iterative test configuration will be directly applied to the next actual testing phase, providing a basis for the optimization of OLED devices.

[0118] S222: Based on the iterative test configuration, perform iterative testing to monitor the performance of the OLED device under a variety of simulated usage conditions, pay special attention to those conditions identified as high-frequency usage patterns, record the device's response and performance changes, and generate a test result data set;

[0119] Based on the developed iterative test configuration, iterative testing was performed, focusing on analyzing the performance of OLED devices under conditions identified as high-frequency usage modes. The device's response and performance changes under the conditions were monitored, such as the continuity of screen brightness, power consumption, and response time. Detailed performance data was recorded and the data was then aggregated into a test result data set, which included all relevant performance indicators from the start to the end of the test, providing an information source for further data analysis and performance evaluation.

[0120] S223: Analyze the test result data set, evaluate the impact of various usage modes on the life of OLED, identify key performance degradation factors, and obtain iterative optimization results;

[0121] A comprehensive analysis of the result data set obtained from the iterative testing is conducted. The analysis aims to evaluate the specific impact of different usage patterns on the life of OLED devices and identify key attenuation factors that affect device performance and life, such as the impact of temperature changes on screen performance and the impact of power-on and power-off frequency on overall life. Through the evaluation of these factors, iterative optimization results are obtained, including improvements to device design and usage recommendations.

[0122] See also Figure 6 ,The steps for real-time processing and analysis of the collected data are as follows:

[0123] S311: Based on the iterative optimization results, collect OLED data from multiple sources, including production batches, usage environment and maintenance records, extract the data and unify the format to obtain a standardized preliminary data set;

[0124] When collecting OLED data from multiple sources, information is obtained from different channels such as production batches, usage environment and maintenance records, and data is entered. During the entry, the data format is checked to see if it meets the predefined standards, such as date format, data type, etc., to ensure the consistency and processability of the data. Once the data passes the preliminary formatting and standardization process, it is stored as a preliminary data set to provide a basis for subsequent processing and analysis. The process includes synchronizing data from various data sources and standardizing it for effective data integration and utilization;

[0125] S312: Based on the standardized preliminary data set, outlier detection and correction, missing value processing and data normalization are performed to obtain a ready data set;

[0126] When entering the further data processing stage, the collected standardized data sets are quality checked, including the detection and correction of outliers, which involves statistical analysis methods to identify and process data points that deviate from the normal range. Next, missing value processing is carried out, and interpolation or prediction model-based methods are used to fill in missing data. The data normalization process ensures that data from different sources are at the same measurement standard when compared and comprehensively analyzed, ensuring the accuracy and consistency of the data set.

[0127] S313: Analyze the prepared data set using the formula:

[0128] ;

[0129] Identify key trends and patterns, mine data information, and obtain key data indicators ,in, It is The weight of the data item, It is The value of the item data, is the total number of data items;

[0130] There are the following data items and their weights, the data values ​​are , the weight is , substituting into the formula we get:

[0131] ;

[0132] The results show that after weighting the importance of each data item, the key data indicators obtained are 119, and the indicators comprehensively reflect the weighted contribution of all data points.

[0133] See also Figure 7 ,The acquisition steps of the multi-source data integration framework are as follows:

[0134] S321: extract key information from the collected data, identify the correlation and potential influencing factors between different data sources, and obtain data source analysis records;

[0135] Extract key information from the collected data, analyze the multi-source characteristics of the data, including the statistical dimensions of production batches, changing parameters of the usage environment, and key indicators in maintenance records, use screening rules to preliminarily classify the data, such as by time series or source type, and further input the grouped data into the correlation analysis process to analyze the potential impact of each data source on the overall data set, identify the correlation and related characteristics between different data sources, and finally record the above analysis results in the data source analysis record.

[0136] S322: Based on the data source analysis records, through the continuous feedback mechanism, the formula is used:

[0137] ;

[0138] Optimize data analysis and processing efficiency to obtain a multi-source data integration framework, where: Represents the analysis indicator, It is The impact factor of each data source, It is The processing results obtained from the data source are is the number of data sources;

[0139] There are 3 data sources, and their processing results are , , The corresponding impact factors are , , , the total number of data sources is , substitute into the formula to calculate:

[0140] ;

[0141] The calculation results in:

[0142] ;

[0143] ;

[0144] The results show that after comprehensively considering the weights of each data source, the analysis indicators ,The indicators comprehensively reflect the dynamic contribution of multi-source data to ,analysis efficiency and processing capability.

[0145] See also Figure 8 , the steps to obtain the optimal roll-off prediction model are as follows:

[0146] S411: Based on the multi-source data integration framework, collect historical efficiency roll-off data and current OLED device performance data, including brightness decay and lifetime data, to obtain an integrated OLED data set;

[0147] By integrating data from different sources, we actively collect historical and current performance data on OLED devices, including but not limited to brightness decay rate and device life. The data collection covers records from testing, actual usage feedback, and production monitoring. The data is cleaned, sorted, and de-redundant to ensure data quality and consistency. Discrete data items are further integrated together to form a comprehensive OLED data set, which will be used to drive model training and verification in the future to ensure that the resulting data set is not only complete but also highly relevant, providing a data foundation for the model.

[0148] S412: Based on the integrated OLED data set, perform a baseline test, calculate the error between the model prediction result and the historical data, and obtain a model prediction performance evaluation record;

[0149] After acquiring the integrated OLED data set, a baseline performance test is immediately conducted. The process involves an in-depth analysis of the data, mainly evaluating the performance of the model by comparing the model's prediction results with known historical data, including mean square error calculation, trend line analysis, and deviation percentage evaluation, to ensure that each step is correct. A performance evaluation record is generated based on the test results. The record illustrates the accuracy and reliability of the model in predicting OLED device performance, providing a basis for subsequent model optimization and parameter adjustment.

[0150] S413: Based on the model prediction performance evaluation record, the model parameters are adjusted again using the formula:

[0151] ;

[0152] The model parameters are optimized according to the brightness attenuation and life reduction trend of OLED devices to obtain the optimal roll-off prediction model, where: are the adjusted model parameters, are the existing model parameters, It is The observed value of data points, It is The model prediction value for data points, It is The independent variables of each data point include factors such as time and temperature. is the learning rate, which determines the sensitivity of parameter updates, is the total number of data points;

[0153] The known parameters are as follows: , , the observed value and predicted value of each data point are and , the corresponding independent variable , the total number of data points , calculated by substituting the values ​​into the formula:

[0154] ;

[0155] ;

[0156] This result shows that by adjusting the model parameters, the new parameters The increase from 0.5 to 0.53 indicates that the model’s adaptability to the data has improved and it can more accurately predict the performance roll-off of OLED devices.

[0157] The system for building the OLED efficiency roll-off prediction model includes:

[0158] The performance index extraction module is based on the collected OLED device performance related data, including brightness, pixel density, temperature and voltage parameters, to evaluate the relevance of the parameters in predicting OLED efficiency roll-off, filter and process the data to reflect the attenuation behavior of OLED and construct a set of key performance indicators;

[0159] The iterative optimization analysis module integrates the newly added OLED usage data based on the key performance indicator set, analyzes the time series and usage frequency, and optimizes the identification process of factors affecting OLED life through multi-cycle iterative testing to obtain iterative optimization results;

[0160] Based on the iterative optimization results, the multi-source data processing module collects OLED data from multiple sources, including production batches, usage environment and maintenance records, processes and analyzes the collected data in real time, optimizes data analysis and processing efficiency, and obtains a multi-source data integration framework;

[0161] The efficiency verification and adjustment module is based on a multi-source data integration framework, and compares it with historical efficiency roll-off data to test the model's prediction performance and data processing accuracy, and generate prediction performance verification results;

[0162] The prediction model building module uses the prediction performance verification results to adjust the model parameters again to match the brightness attenuation and life reduction trends of OLED devices to obtain the optimal roll-off prediction model.

[0163] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for constructing an OLED efficiency roll-off prediction model, characterized in that: The following steps are involved: Based on the collected data related to OLED device performance, key performance indicators are extracted, including brightness, pixel density, temperature and voltage parameters, the relevance of the parameters in predicting OLED efficiency roll-off is evaluated, and the data processing flow is adjusted to construct a set of key performance indicators; Based on the key performance indicator set, the newly added OLED usage data is integrated, the time series and usage frequency in the data are analyzed, and the identification process of factors affecting the life of OLED is optimized through multi-cycle iterative testing to obtain iterative optimization results; Based on the iterative optimization results, OLED data from multiple sources are collected, including production batches, usage environments, and maintenance records, and the collected data are processed and analyzed in real time. The data analysis and processing efficiency is optimized through a continuous feedback mechanism to obtain a multi-source data integration framework; The steps of real-time processing and analysis of the collected data are specifically as follows: Based on the iterative optimization results, collect OLED data from multiple sources, including production batches, usage environment and maintenance records, extract the data and unify the format to obtain a standardized preliminary data set; Based on the standardized preliminary data set, outlier detection and correction, missing value processing and data normalization are performed to obtain a ready data set; The prepared data set is analyzed using the formula: ； Identify key trends and patterns, mine data information, and obtain key data indicators ,in, It is The weight of the data item, It is The value of the item data, is the total number of data items; The steps for obtaining the multi-source data integration framework are specifically as follows: Extract key information from the collected data, identify the correlation and potential influencing factors between different data sources, and obtain data source analysis records; Based on the analysis records of the data source, through a continuous feedback mechanism, the formula is adopted: ； Optimize data analysis and processing efficiency to obtain a multi-source data integration framework, where: Represents the analysis indicator, It is The impact factor of each data source, It is The processing results obtained from the data source are is the number of data sources; Based on the multi-source data integration framework, the model prediction performance and data processing accuracy are tested by comparing with the historical efficiency roll-off data, and the model parameters are adjusted again to match the brightness attenuation and life reduction trends of the OLED device to obtain the optimal roll-off prediction model; The steps for obtaining the optimal roll-off prediction model are specifically as follows: Based on the multi-source data integration framework, historical efficiency roll-off data and current OLED device performance data, including brightness decay and lifetime data, are collected to obtain an integrated OLED data set; Based on the integrated OLED data set, a baseline test is performed to calculate the error between the model prediction result and the historical data to obtain a model prediction performance evaluation record; Based on the model prediction performance evaluation record, the model parameters are adjusted again using the formula: ； The model parameters are optimized according to the brightness attenuation and life reduction trend of OLED devices to obtain the optimal roll-off prediction model, where: are the adjusted model parameters, are the existing model parameters, It is The observed value of data points, It is The model prediction value for data points, It is The independent variable has data points, is the learning rate, is the total number of data points.

2. The method for constructing an OLED efficiency roll-off prediction model according to claim 1, characterized in that: The steps for evaluating the relevance of the efficiency roll-off are specifically as follows: Based on the collected OLED device performance correlation data, brightness, pixel density, temperature and voltage parameters are extracted to obtain a key parameter data set; Based on the key parameter data set, the correlation with OLED efficiency roll-off is analyzed using the formula: ； Calculate the correlation coefficient between parameters and efficiency roll-off , and get the correlation analysis results, where Represents a single performance parameter value, represents the mean value of performance parameters, represents the efficiency roll-off value, represents the efficiency roll-off mean; According to the correlation analysis result, the correlation between the performance parameter and the efficiency roll-off of the OLED is evaluated and determined to obtain a parameter evaluation result.

3. The method for constructing an OLED efficiency roll-off prediction model according to claim 2, characterized in that: The steps for obtaining the key performance indicator set are specifically as follows: According to the parameter evaluation results, adjusting the data processing flow includes normalizing the data set and screening parameters associated with efficiency roll-off to obtain an adjusted data processing configuration; Based on the adjusted data processing configuration, the formula is adopted: ； Calculate the evaluation value , and obtain a set of key performance indicators, where It is The weight of each performance parameter, It is The quantitative value of a performance parameter, is the total number of parameters involved in the evaluation.

4. The method for constructing an OLED efficiency roll-off prediction model according to claim 1, characterized in that: The analysis steps of the time series and usage frequency are specifically as follows: Based on the key performance indicator set, newly added OLED usage data, including the number of startup times, the duration of each usage, and the shutdown time, are integrated to obtain a time-correlated performance data set; Performing time series analysis on the time-correlated performance data set to determine the periodicity, trend and randomness of the data and obtain a time series analysis result; Based on the time series analysis results, the formula is adopted: ； Analyze usage frequency , identifying the high-frequency and low-frequency modes used by the device, where It is time interval, is the attenuation coefficient, which indicates the weight of the influence of time interval on the usage frequency. is the number of time intervals.

5. The method for constructing an OLED efficiency roll-off prediction model according to claim 1, characterized in that: The steps for obtaining the iterative optimization results are specifically as follows: Based on the time series and usage frequency, a multi-cycle iterative test process is formulated, and test parameters are continuously adjusted to simulate the response of OLED devices under different usage conditions to obtain an iterative test configuration; Based on the iterative test configuration, perform iterative testing to monitor the performance of the OLED device under a variety of simulated usage conditions, record the response and performance changes of the device, and generate a test result data set; The test result data set is analyzed to evaluate the impact of various usage modes on the life of the OLED, identify key performance degradation factors, and obtain iterative optimization results.

6. A system for constructing an OLED efficiency roll-off prediction model, characterized in that: According to the method for constructing an OLED efficiency roll-off prediction model according to any one of claims 1 to 5, the system comprises: The performance index extraction module is based on the collected OLED device performance related data, including brightness, pixel density, temperature and voltage parameters, to evaluate the relevance of the parameters in predicting OLED efficiency roll-off, filter and process the data to reflect the attenuation behavior of OLED and construct a set of key performance indicators; The iterative optimization analysis module integrates the newly added OLED usage data based on the key performance indicator set, analyzes the time series and usage frequency, optimizes the identification process of factors affecting the life of OLED through multi-cycle iterative testing, and obtains iterative optimization results; The multi-source data processing module collects OLED data from multiple sources, including production batches, usage environment and maintenance records, based on the iterative optimization results, processes and analyzes the collected data in real time, optimizes data analysis and processing efficiency, and obtains a multi-source data integration framework; The efficiency verification and adjustment module compares the historical efficiency roll-off data based on the multi-source data integration framework to test the prediction performance and data processing accuracy of the model and generate prediction performance verification results; The prediction model building module uses the prediction performance verification results to adjust the model parameters again to match the brightness attenuation and life reduction trends of the OLED device to obtain the optimal roll-off prediction model.

Images