Method and system based on printing process parameter optimization

Abstract

The present invention relates to the technical field of process data optimization, and in particular to a method and system based on printing process parameter optimization. The method comprises the following steps: obtaining printing process area type data; collecting printing area parameters based on the printing process area type data, thereby obtaining first printing process data; calibrating the first parameter of the first printing process data to generate first printing process calibration data; optimizing sensor acquisition based on the first sensor process calibration data to obtain sensor process acquisition optimization data; performing printing process high abnormality analysis on the sensor process acquisition optimization data to generate printing process abnormality data. The present invention realizes the prediction of printing process data by calibrating sensor parameters and establishing a complex nonlinear correlation model, thereby improving the printing quality and production efficiency after optimizing the parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of process data optimization, and in particular to a method and system based on printing process parameter optimization. Background Art

[0002] With the rapid development of electronic technology and computer science, digital printing technology emerged in the second half of the 20th century. Digital printing technology has further improved the quality of printed products, but it has also brought new challenges, such as color management and printing parameter optimization. With the development of artificial intelligence and data analysis technology, printing process parameter optimization methods have gradually become a research hotspot in the printing industry. By collecting and analyzing a large amount of printing data, combined with mathematical models and optimization algorithms, researchers have begun to explore how to optimize printing process parameters to the greatest extent to achieve the optimization of printing quality and improve production efficiency. In recent years, with the widespread application of deep learning and machine learning technologies, printing process parameter optimization methods have made significant progress. By training neural network models, printing companies can automatically optimize various parameters in the printing process, thereby improving the quality of printed products, reducing production costs, and enhancing competitiveness. However, the current printing process optimization is often static and lacks adaptability to dynamic environments. At the same time, the detection and processing of printing process anomalies usually requires a lot of time and manpower, resulting in low printing quality and production efficiency after optimization. Summary of the invention

[0003] Based on this, it is necessary to provide a method and system based on printing process parameter optimization to solve at least one of the above technical problems.

[0004] To achieve the above object, a printing process parameter optimization method is provided, the method comprising the following steps:

[0005] Step S1: Acquire printing process area type data; perform printing area parameter collection based on the printing process area type data, thereby acquiring first printing process data; perform first parameter calibration on the first printing process data, thereby generating first printing process calibration data; perform sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data;

[0006] Step S2: Perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; perform second printing process data screening on the sensor process acquisition optimization data according to the abnormal printing process data to obtain second printing process data; perform complex nonlinear correlation modeling on the second printing process data to generate predicted printing process data;

[0007] Step S3: obtaining printing process production requirement parameters; defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function; performing printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy;

[0008] Step S4: Based on the printing process adaptive control strategy, the first control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the first control parameter and the first control output time; the second control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the second control parameter and the second control output time; the first control output time and the second control output time are weighted for control response difference to obtain the printing process response weight; the first control parameter and the second control parameter are weighted for control performance difference to obtain the printing process performance weight; a comprehensive evaluation of printing optimization is performed based on the printing process performance weight and the printing process response weight, and a printing process optimization report is constructed to evaluate the optimization of printing process parameters.

[0009] The present invention can improve the accuracy and reliability of data by acquiring printing process area type data and sensor process acquisition optimization data, and performing parameter acquisition and calibration, thereby providing reliable basic data for subsequent printing process optimization. By performing high abnormality analysis of the printing process on the sensor process acquisition optimization data, abnormal conditions in the printing process can be detected and identified in a timely manner, helping to eliminate potential problems and improve the stability and consistency of the printing process. By performing complex nonlinear association modeling on the second printing process data and generating predicted printing process data, it can help understand the association relationship and trend in the printing process and provide a reference for formulating a reasonable printing process optimization strategy. By defining the objective function according to the printing process production requirement parameters and performing printing adaptive control based on the printing process optimization data, real-time monitoring and adjustment of the printing process can be achieved, and production efficiency and quality stability can be improved. According to the printing process performance weight and the printing process response weight, a comprehensive evaluation of printing optimization is performed, and a printing process optimization report is generated, which can provide an evaluation and feedback on the optimization effect of the printing process parameters, and provide guidance for further improving and optimizing the printing process. Therefore, the present invention realizes the prediction of the printing process data by calibrating the sensor parameters and establishing a complex nonlinear association model, thereby improving the printing quality and production efficiency after optimizing the parameters.

[0010] In this specification, a printing process parameter optimization system is provided, which is used to execute the above-mentioned printing process parameter optimization method. The printing process parameter optimization system includes:

[0011] A sensor acquisition optimization module is used to obtain printing process area type data; perform printing area parameter acquisition based on the printing process area type data, thereby obtaining first printing process data; perform first parameter calibration on the first printing process data to generate first printing process calibration data; perform sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data;

[0012] A nonlinear association module is used to perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; screen the second printing process data on the sensor process acquisition optimization data according to the abnormal printing process data to obtain the second printing process data; perform complex nonlinear association modeling on the second printing process data to generate predicted printing process data;

[0013] The adaptive control module is used to obtain the required parameters of printing process production; define the objective function according to the required parameters of printing process production to obtain the printing process production objective function; perform printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy;

[0014] The optimization evaluation module is used to output the first control parameter of the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously record the control output time to obtain the first control parameter and the first control output time; output the second control parameter of the optimal printing process iteration data, and synchronously record the control output time to obtain the second control parameter and the second control output time; weight the control response difference between the first control output time and the second control output time to obtain the printing process response weight; weight the control performance difference between the first control parameter and the second control parameter to obtain the printing process performance weight; perform a comprehensive evaluation of printing optimization based on the printing process performance weight and the printing process response weight, and construct a printing process optimization report to evaluate the optimization of printing process parameters.

[0015] The beneficial effect of the present invention is that by collecting and analyzing printing process data, various parameters and variables in the printing process can be understood more accurately, so as to perform effective optimization. Abnormal analysis can detect abnormal conditions in the printing process in a timely manner, so that corresponding strategies can be adopted for adjustment and correction. Through complex nonlinear association modeling, key parameters and variables in the printing process can be predicted, which helps to prepare and adjust in advance. The adaptive control strategy based on printing process optimization data can adjust the control parameters in real time, adapt to different printing conditions and requirements, and improve printing quality and efficiency. By comprehensively evaluating the performance and response weights of the printing process, comprehensive optimization suggestions and improvement plans can be provided to decision makers to promote the continuous improvement of the printing process. Therefore, the present invention realizes the prediction of printing process data by calibrating sensor parameters and establishing a complex nonlinear association model, thereby improving the printing quality and production efficiency after optimizing the parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a step flow chart based on a printing process parameter optimization method;

[0017] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0018] Figure 3 for Figure 2 Detailed implementation steps of step S21;

[0019] Figure 4 for Figure 2 Detailed implementation steps of step S22;

[0020] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0021] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0022] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0023] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0024] To achieve this, please refer to Figures 1 to 4 , a printing process parameter optimization method, the method comprising the following steps:

[0025] Step S1: Acquire printing process area type data; perform printing area parameter collection based on the printing process area type data, thereby acquiring first printing process data; perform first parameter calibration on the first printing process data, thereby generating first printing process calibration data; perform sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data;

[0026] Step S2: Perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; perform second printing process data screening on the sensor process acquisition optimization data according to the abnormal printing process data to obtain second printing process data; perform complex nonlinear correlation modeling on the second printing process data to generate predicted printing process data;

[0027] Step S3: obtaining printing process production requirement parameters; defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function; performing printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy;

[0028] Step S4: Based on the printing process adaptive control strategy, the first control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the first control parameter and the first control output time; the second control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the second control parameter and the second control output time; the first control output time and the second control output time are weighted for control response difference to obtain the printing process response weight; the first control parameter and the second control parameter are weighted for control performance difference to obtain the printing process performance weight; a comprehensive evaluation of printing optimization is performed based on the printing process performance weight and the printing process response weight, and a printing process optimization report is constructed to evaluate the optimization of printing process parameters.

[0029] The present invention can improve the accuracy and reliability of data by acquiring printing process area type data and sensor process acquisition optimization data, and performing parameter acquisition and calibration, thereby providing reliable basic data for subsequent printing process optimization. By performing high abnormality analysis of the printing process on the sensor process acquisition optimization data, abnormal conditions in the printing process can be detected and identified in a timely manner, helping to eliminate potential problems and improve the stability and consistency of the printing process. By performing complex nonlinear association modeling on the second printing process data and generating predicted printing process data, it can help understand the association relationship and trend in the printing process and provide a reference for formulating a reasonable printing process optimization strategy. By defining the objective function according to the printing process production requirement parameters and performing printing adaptive control based on the printing process optimization data, real-time monitoring and adjustment of the printing process can be achieved, and production efficiency and quality stability can be improved. According to the printing process performance weight and the printing process response weight, a comprehensive evaluation of printing optimization is performed, and a printing process optimization report is generated, which can provide an evaluation and feedback on the optimization effect of the printing process parameters, and provide guidance for further improving and optimizing the printing process. Therefore, the present invention realizes the prediction of the printing process data by calibrating the sensor parameters and establishing a complex nonlinear association model, thereby improving the printing quality and production efficiency after optimizing the parameters.

[0030] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a method for optimizing printing process parameters according to the present invention. In this example, the method for optimizing printing process parameters includes the following steps:

[0031] Step S1: Acquire printing process area type data; perform printing area parameter collection based on the printing process area type data, thereby acquiring first printing process data; perform first parameter calibration on the first printing process data, thereby generating first printing process calibration data; perform sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data;

[0032] In an embodiment of the present invention, different types of regions that need to be considered in the printing process are determined, such as the printing head, edge, center, etc. The printing process data of each type of region is collected by means of sensors, monitoring equipment or manual observation, including parameters such as ink flow, printing speed, temperature, humidity, etc. According to each type of region, parameters are collected for each region, and the printing process parameter data of each region is recorded. The collected printing process data is recorded and sorted for subsequent analysis and processing. The key parameters therein are calibrated using the first printing process data to ensure that it accurately reflects the characteristics of the actual printing process, and the first printing process calibration data is generated, including information such as calibration parameter values ​​and calibration process records. According to the first sensor process calibration data, the sensor is optimized and adjusted to improve the precision and accuracy of data acquisition, and the key data and adjustment parameters in the sensor optimization process are recorded to form sensor process acquisition optimization data.

[0033] Step S2: Perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; perform second printing process data screening on the sensor process acquisition optimization data according to the abnormal printing process data to obtain second printing process data; perform complex nonlinear correlation modeling on the second printing process data to generate predicted printing process data;

[0034] In an embodiment of the present invention, the sensor process acquisition optimization data is analyzed by using statistical methods or machine learning techniques to detect and identify abnormal data points, that is, data that deviates greatly from the normal process. According to the abnormal analysis results, the abnormal data points are extracted and recorded to form a printing process abnormal data set, which includes information such as parameter values ​​and timestamps of the abnormal data points. According to the printing process abnormal data, the sensor process acquisition optimization data is screened, the abnormal data points are eliminated or adjusted, and the data that meets the normal process requirements is retained. Based on the screened data, the second printing process data is formed, which includes processed and optimized printing process parameter data. The second printing process data is modeled using complex nonlinear modeling techniques, such as neural networks, support vector machines, or random forests, to capture the complex associations and nonlinear relationships therein. Based on the established model, the printing process is predicted to generate predicted printing process data, which includes the printing process parameter values ​​predicted by the model.

[0035] Step S3: obtaining printing process production requirement parameters; defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function; performing printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy;

[0036] In an embodiment of the present invention, by determining the key parameters that need to be optimized and controlled during the printing process production process, such as printing speed, ink flow, temperature, etc. Real-time data of the printing process production requirement parameters are obtained through monitoring equipment, sensors or manual collection. According to the printing process production requirement parameters, the goals that need to be optimized in the production process are clarified, such as maximizing production speed, minimizing ink consumption, etc. The goals are converted into mathematical expressions, and the printing process production objective function is defined to facilitate target tracking and evaluation in the subsequent optimization control process. The printing process optimization data generated in the previous steps is combined with the real-time collected production requirement parameters to adjust and optimize the printing process production objective function in real time. Based on the optimization data and the objective function, an adaptive control strategy for the printing process is designed, that is, the printing parameters are dynamically adjusted according to the real-time production conditions and target changes to achieve the optimal production effect.

[0037] Step S4: Based on the printing process adaptive control strategy, the first control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the first control parameter and the first control output time; the second control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the second control parameter and the second control output time; the first control output time and the second control output time are weighted for control response difference to obtain the printing process response weight; the first control parameter and the second control parameter are weighted for control performance difference to obtain the printing process performance weight; a comprehensive evaluation of printing optimization is performed based on the printing process performance weight and the printing process response weight, and a printing process optimization report is constructed to evaluate the optimization of printing process parameters.

[0038] In an embodiment of the present invention, the parameter values ​​required for the first control are calculated using the optimal printing process iteration data according to the printing process adaptive control strategy. The first control output time, that is, the time point when the control parameter is output, is recorded. After a certain time interval, the output of the control parameter is recalculated according to the printing process adaptive control strategy to obtain the parameter value of the second control. The second control output time, that is, the time point when the second control parameter is output, is recorded. According to the time points of the first and second control outputs, the difference in control response time is calculated. The difference in control response time is assigned corresponding weights to reflect the importance of control output at different time points. The performance of the first and second control parameter outputs is evaluated, such as printing speed, ink consumption, etc. The difference in control performance is assigned corresponding weights to reflect the impact of different control parameters on printing process performance. A comprehensive evaluation of the printing process is performed in combination with the response weight and performance weight of the control. A printing process optimization report is constructed to evaluate the optimization effect of the printing process parameters and put forward improvement suggestions.

[0039] Preferably, step S1 comprises the following steps:

[0040] Step S11: acquiring printing process area type data; deploying sensors in the area based on the printing process area type data, and collecting printing area parameters using the deployed sensors, thereby acquiring first printing process data, wherein the first printing process data includes printing ink flow, printing speed, printing temperature and printing humidity;

[0041] Step S12: performing data preprocessing on the first printing process data to generate first printing process cleaning data, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization;

[0042] Step S13: performing signal filtering on the first printing process cleaning data to generate first printing process filtering data;

[0043] Step S14: performing first parameter calibration on the first printing process filter data to generate first printing process calibration data; performing sensor acquisition optimization on the sensor based on the first sensor process calibration data to obtain sensor process acquisition optimization data.

[0044] By acquiring the printing process area type data, the present invention can understand the characteristics and requirements of different areas in the printing process, and provide basic data for subsequent sensor deployment and parameter collection. Data preprocessing ensures the quality and accuracy of the first printing process data, including cleaning data, filling missing values ​​and standardizing data, thereby providing a reliable data basis for subsequent analysis and processing. Signal filtering helps to remove noise and interference in the first printing process data, improve the stability and reliability of the data, and make subsequent analysis more accurate. By performing parameter calibration and sensor acquisition optimization on the first printing process data, the deployment and acquisition strategy of the sensor can be optimized, and the efficiency and accuracy of data acquisition can be improved, thereby providing strong support for the optimization and improvement of the printing process.

[0045] In an embodiment of the present invention, different types of regions in the printing process are determined, such as the printing head, bottom, edge, etc. Then, according to these region types, sensors are deployed in the printing device to collect corresponding printing process parameter data in each region. The deployed sensors are used to collect printing region parameters to obtain first printing process data, including parameters such as printing ink flow, printing speed, printing temperature and printing humidity. Data preprocessing is performed on the acquired first printing process data. Data preprocessing includes data cleaning, that is, removing outliers and correcting outlier data points; filling missing data values, that is, filling or estimating missing data; data standardization, that is, normalizing or standardizing data for subsequent analysis and processing. Signal filtering is performed on the preprocessed first printing process data to remove noise and smooth data. Signal filtering can use common filtering algorithms, such as moving average filtering, median filtering, etc., to improve the quality and stability of data and generate first printing process filtered data. The first printing process data after filtering is used to calibrate the first parameter. Parameter calibration can use statistical analysis or modeling methods to determine the relationship between various parameters, and generate the first printing process calibration data based on the existing data. At the same time, the sensor is optimized for sensor acquisition based on the first sensor process calibration data to improve the accuracy and stability of the sensor data, thereby obtaining sensor process acquisition optimization data.

[0046] Preferably, step S14 includes the following steps:

[0047] Step S141: performing data error analysis on the first printing process filter data to generate first printing process error data;

[0048] Step S142: performing calibration curve fitting using the first printing process error data to generate a first printing process calibration curve;

[0049] Step S143: calibrating the actual parameters of the first printing process filter data according to the first printing process calibration curve to generate first printing process calibration data;

[0050] Step S144: performing acquisition optimization on the sensor based on the first printing process calibration data to obtain an optimized calibration sensor, wherein the parameter acquisition optimization includes sensor position optimization, sampling frequency optimization and gain optimization;

[0051] Step S145: Optimize and collect printing area parameters by optimizing and calibrating the sensor, thereby generating sensor process collection optimization data.

[0052] The present invention can identify and quantify the source of errors in the data by performing data error analysis on the first printing process filter data, provide a reference basis for subsequent calibration and improvement, and improve the accuracy and reliability of the data. By using the first printing process error data to perform calibration curve fitting, a relationship model between parameters can be established to provide a basis for actual parameter calibration, thereby more accurately adjusting the printing process parameters. According to the first printing process calibration curve, the actual parameter calibration of the filter data is performed, that is, the relationship between the actual measured value and the theoretical value is calibrated, thereby improving the accuracy and reliability of the sensor measurement. By performing sensor acquisition optimization based on the first printing process calibration data, including position optimization, sampling frequency optimization and gain optimization, the performance of the sensor and the data acquisition efficiency can be improved, and the parameter control in the printing process can be further optimized. By optimizing the calibration sensor to perform optimized acquisition of printing area parameters, more accurate and reliable printing process data can be obtained, providing support for improving printing quality and improving production efficiency.

[0053] In an embodiment of the present invention, error analysis is performed on the first printing process data after filtering. By comparing the difference between the filtered data and the actual value, the error distribution and characteristics of the data are analyzed to generate the first printing process error data. The calibration curve is fitted using the first printing process error data. A mathematical model or a curve fitting algorithm can be used to fit the error data into a calibration curve to describe the regularity of the error changing with the parameter. According to the generated calibration curve, the actual parameters of the first printing process filter data are calibrated. The filter data is converted into the actual parameter value by methods such as curve interpolation or reverse calculation to generate the first printing process calibration data. The first printing process calibration data is used to optimize the sensor acquisition. Optimizing parameter acquisition includes optimizing the sensor position, that is, adjusting the sensor position to improve the accuracy of data acquisition; optimizing the sampling frequency, that is, optimizing the frequency of data acquisition to adapt to the actual process requirements; and optimizing the gain, that is, adjusting the gain parameter of the sensor to optimize the sensitivity and range of the data. The optimized calibration sensor is used to optimize the acquisition of printing area parameters. By acquiring the optimized sensor data, more accurate and stable printing process data is obtained, thereby generating sensor process acquisition optimization data.

[0054] Preferably, step S2 comprises the following steps:

[0055] Step S21: performing printing process high abnormality analysis on the sensor process collection optimization data to generate printing process abnormality data;

[0056] Step S22: screening the sensor process collection optimization data for second printing process data according to the printing process abnormality data to obtain second printing process data;

[0057] Step S23: extracting multi-dimensional printing features from the second printing process data to generate printing process multi-dimensional feature data;

[0058] Step S24: Perform complex nonlinear correlation modeling on the multi-dimensional feature data of the printing process to generate predicted printing process data.

[0059] The present invention can identify and record abnormal conditions in the printing process by performing high abnormality analysis on the sensor process acquisition optimization data, help diagnose potential problems and improvement directions, and improve the stability and reliability of the printing process. The sensor process acquisition optimization data is screened according to the printing process abnormality data to obtain the second printing process data, which helps to take corresponding treatment measures for the abnormal situation and optimize the printing process flow. Multi-dimensional printing feature extraction is performed on the second printing process data, and the characteristics and changes in the printing process can be analyzed from multiple angles, providing a richer data foundation for subsequent modeling and prediction. By performing complex nonlinear correlation modeling on the multi-dimensional feature data of the printing process, the potential complex relationships and laws in the data can be revealed, and then the changing trends and problems in the printing process can be predicted, helping to optimize the printing process and improve printing quality.

[0060] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0061] Step S21: performing printing process high abnormality analysis on the sensor process collection optimization data to generate printing process abnormality data;

[0062] In an embodiment of the present invention, an abnormality in the sensor process acquisition optimization data is analyzed by selecting a suitable anomaly detection algorithm. Common anomaly detection algorithms include statistical-based methods (such as mean, standard deviation), distance-based methods (such as K nearest neighbor algorithm, isolation forest algorithm) and machine learning-based methods (such as support vector machine, random forest), etc. Anomaly analysis of the sensor process acquisition optimization data and identification of abnormal printing process data involves setting thresholds or training models to determine abnormal data points, and marking or classifying abnormal data. The identified printing process abnormality data is recorded, including relevant timestamps, sensor locations and other information. These records will be helpful for subsequent analysis and processing.

[0063] Step S22: screening the sensor process collection optimization data for second printing process data according to the printing process abnormality data to obtain second printing process data;

[0064] In an embodiment of the present invention, the sensor process acquisition optimization data is screened, including deleting abnormal data points, correcting abnormal data, or filling abnormal data using an interpolation method. The second printing process data is generated according to the sensor process acquisition optimization data after the abnormal data is screened. These data include parameters such as printing ink flow, printing speed, printing temperature, and printing humidity. After screening, the quality and reliability of the data are guaranteed. The generated second printing process data is quality evaluated to check the integrity, accuracy, and consistency of the data. This can be done through statistical analysis, data visualization, and the like.

[0065] Step S23: extracting multi-dimensional printing features from the second printing process data to generate printing process multi-dimensional feature data;

[0066] In an embodiment of the present invention, appropriate features are selected from the second printing process data for extraction. These features may include parameters such as printing speed, ink flow, temperature, humidity, and the correlation and interaction features between them. According to the selected features, a suitable feature extraction method is used to extract multidimensional printing features from the second printing process data. Common feature extraction methods include statistical feature extraction (such as mean, standard deviation, kurtosis, skewness, etc.), frequency domain feature extraction (such as power spectrum density, frequency components, etc.), time domain feature extraction (such as time series statistical features, autocorrelation functions, etc.), etc. The extracted features are combined to form a multidimensional feature vector to form multidimensional feature data of the printing process. These feature vectors can reflect multiple aspects of the printing process, including quality, stability, efficiency, etc.

[0067] Step S24: Perform complex nonlinear correlation modeling on the multi-dimensional feature data of the printing process to generate predicted printing process data.

[0068] In an embodiment of the present invention, the integrity and accuracy of the data are ensured by using the multidimensional feature data of the printing process generated in step S23 as the input data of the model. A suitable complex nonlinear association modeling method is selected to construct a prediction model. Common methods include neural network models, support vector machine models, decision tree models, etc., which can capture the complex relationship between features. The input multidimensional feature data of the printing process is trained using the selected model. During the training process, the parameters of the model are optimized so that the model can best fit the input data and can accurately predict the printing process parameters. The trained model is evaluated to check the performance and generalization ability of the model. Cross-validation, error analysis and other methods can be used to evaluate the accuracy and stability of the model. The model is optimized and adjusted according to the evaluation results to improve the prediction ability and robustness of the model, including adjusting the structure of the model, parameter tuning, feature engineering and other methods. The trained model is used to predict the new multidimensional feature data of the printing process to generate predicted printing process data.

[0069] Preferably, step S21 includes the following steps:

[0070] Step S211: Screening abnormal parameters of sensor process acquisition optimization data to obtain abnormal process acquisition data;

[0071] Step S212: performing abnormal data collection timing analysis on the abnormal process collection data to generate an abnormal data collection timestamp;

[0072] Step S213: performing abnormal collection time interval analysis on the abnormal process collection data according to the abnormal data collection timestamp to generate the abnormal collection time interval;

[0073] Step S214: Compare the abnormal collection time interval with the preset abnormal time interval threshold. When the abnormal collection time interval is greater than or equal to the preset abnormal time interval threshold, integrate the timestamps within the corresponding abnormal collection time interval and mark it as a high abnormality frequent period.

[0074] Step S215: performing printing process abnormality analysis on the abnormal process collection data based on the high abnormality frequency period to generate printing process abnormality data, wherein the printing process abnormality analysis includes printing process abnormality type analysis, printing process abnormality location analysis and printing process abnormality degree analysis.

[0075] The present invention can effectively screen out the parameters that cause abnormal situations by screening the sensor process acquisition optimization data for abnormal parameters, and provide key data for subsequent abnormal analysis. Performing time series analysis on abnormal process acquisition data and generating abnormal data acquisition timestamps can help determine the time point and order of abnormal occurrence, and provide a time series perspective for subsequent abnormal analysis. Performing time interval analysis on abnormal acquisition data based on abnormal data acquisition timestamps can help identify the frequency and interval of abnormal occurrence, and provide a basis for subsequent abnormal frequent period marking. Comparing the abnormal acquisition time interval with the preset abnormal time interval threshold and marking the high abnormal frequent period can help focus on the time period that causes the abnormality, and further refine the scope and focus of the abnormal analysis. Performing printing process abnormality analysis based on the high abnormal frequent period, including abnormal type, location and degree analysis, can deeply understand the nature and scope of influence of the abnormal situation, and provide guidance for formulating corresponding improvement measures.

[0076] As an example of the present invention, refer to Figure 3 As shown, in this example, step S21 includes:

[0077] Step S211: Screening abnormal parameters of sensor process acquisition optimization data to obtain abnormal process acquisition data;

[0078] In an embodiment of the present invention, abnormal parameters are identified by performing abnormal detection on each parameter in the sensor process acquisition optimization data. Abnormal parameters are manifested as data deviating from the normal range, large fluctuations or instability. According to the identification results of abnormal parameters, the sensor process acquisition optimization data is screened, and the data points corresponding to the abnormal parameters are eliminated or marked. This can be screened using a threshold-based method, a statistical method, or a machine learning method. The data points that have been screened by the abnormal parameters are combined into an abnormal process acquisition data set. These data sets contain the sensor process acquisition optimization data corresponding to the abnormal parameters, reflecting the characteristics of abnormal conditions in the printing process.

[0079] Step S212: performing abnormal data collection timing analysis on the abnormal process collection data to generate an abnormal data collection timestamp;

[0080] In an embodiment of the present invention, by performing time series analysis on abnormal process collection data, the time series characteristics of the data, such as fluctuations, trend changes, etc., are explored. This can be achieved by drawing a time series diagram, calculating time series statistics, analyzing periodicity, and the like. Based on the time series analysis, an abnormal data collection timestamp is generated. The timestamp indicates the time point when the abnormal data collection occurs, and can accurately record the time of occurrence of the abnormal event, which is helpful for subsequent time series analysis and determination of the cause of the abnormality. According to the abnormal data collection timestamp, the abnormal event is divided into different time periods or intervals. This helps to perform more detailed analysis and location of the abnormal event, and identify information such as the start time, duration, and end time of the abnormal event.

[0081] Step S213: performing abnormal collection time interval analysis on the abnormal process collection data according to the abnormal data collection timestamp to generate the abnormal collection time interval;

[0082] In an embodiment of the present invention, the time interval between adjacent abnormal data points is calculated according to the abnormal data collection timestamp. The time interval can be used to measure the duration of abnormal events and the interval between abnormal events. The calculated abnormal collection time interval is statistically analyzed, including statistics such as the mean, median, maximum, and minimum. This helps to understand the duration distribution of abnormal events and the interval between abnormal events. The distribution of abnormal collection time intervals is analyzed to explore the duration distribution characteristics of abnormal events and the interval distribution characteristics between abnormal events. The distribution of time intervals can be displayed by drawing charts such as histograms and box plots. Anomaly detection is performed on the abnormal collection time interval to identify abnormal time interval data points. This can be achieved using statistical methods, machine learning methods, or special anomaly detection algorithms.

[0083] Step S214: Compare the abnormal collection time interval with the preset abnormal time interval threshold. When the abnormal collection time interval is greater than or equal to the preset abnormal time interval threshold, integrate the timestamps within the corresponding abnormal collection time interval and mark it as a high abnormality frequent period.

[0084] In an embodiment of the present invention, a threshold value of the abnormal time interval is set according to actual conditions, that is, a preset abnormal time interval threshold value. The calculated abnormal collection time interval is compared with the preset abnormal time interval threshold value. If the abnormal time interval is greater than or equal to the preset abnormal time interval threshold value, it is considered that there are high-frequency abnormalities in the time period. The timestamps in the corresponding abnormal collection time interval are integrated and marked as high abnormal frequency periods. This can be achieved by classifying the timestamps into the corresponding time periods and adding marks to indicate that an abnormal event occurred in the time period. The abnormal timestamps marked as high abnormal frequency periods are recorded and appropriately stored and managed.

[0085] Step S215: performing printing process abnormality analysis on the abnormal process collection data based on the high abnormality frequency period to generate printing process abnormality data, wherein the printing process abnormality analysis includes printing process abnormality type analysis, printing process abnormality location analysis and printing process abnormality degree analysis.

[0086] In an embodiment of the present invention, by performing an abnormality type analysis on the abnormal process collection data during a period of high abnormality frequency, different types of printing process abnormalities are identified, such as abnormal ink flow, abnormal temperature, abnormal humidity, etc. Statistical analysis, pattern recognition, or expert experience are used to determine the type of abnormality and its characteristics for subsequent analysis and processing. The location and area of ​​the abnormal process collection data during the printing process are analyzed to determine the specific location of the abnormal event, which involves different parts of the printing machine or different process areas. Combining information such as equipment layout and printing process flow to determine the location where the abnormality occurs helps to accurately locate and eliminate the cause of the abnormality. The abnormal degree of abnormality is analyzed on the abnormal process collection data to evaluate the severity and impact of the abnormal event. Quantitative indicators or qualitative descriptions can be used to evaluate the degree of abnormality, such as the degree of deviation, duration, frequency, and other indicators of abnormal data.

[0087] Preferably, step S22 includes the following steps:

[0088] Step S221: removing abnormal data from the sensor process acquisition optimization data according to the printing process abnormal data to obtain the sensor process acquisition normal data;

[0089] Step S222: performing direct feature extraction on normal data collected by the sensor process to obtain direct feature data of process printing;

[0090] Step S223: performing indirect feature conversion on the process printing direct feature data to generate process printing indirect feature data;

[0091] Step S224: perform indirect feature pattern recognition on the process printing indirect feature data using signal processing technology to generate process printing indirect feature pattern recognition data; perform numerical simulation on the process printing indirect feature pattern recognition data to generate second printing process data.

[0092] The present invention eliminates abnormal data from sensor process acquisition optimization data according to abnormal data of printing process to obtain normal data of sensor process acquisition, thereby eliminating interference from abnormal data and ensuring the accuracy and reliability of subsequent analysis. Direct feature extraction is performed on normal data of sensor process acquisition to obtain direct feature data of process printing, which can capture key features in the process and provide a basis for subsequent indirect feature conversion. Indirect feature conversion is performed on direct feature data of process printing to generate indirect feature data of process printing, which helps to convert raw data into more representative and interpretable features and improve the efficiency and accuracy of pattern recognition. Indirect feature pattern recognition is performed on indirect feature data of process printing using signal processing technology to generate indirect feature pattern recognition data of process printing, and then numerical simulation is performed to generate second printing process data. By performing pattern recognition on indirect feature data using signal processing technology, potential abnormal process patterns or features can be identified, providing clues for subsequent analysis and improvement. Numerical simulation is performed on pattern recognition data to simulate data of the second printing process, which helps to predict and optimize the effect of the printing process and provide improvement suggestions.

[0093] As an example of the present invention, refer to Figure 4 As shown, in this example, step S22 includes:

[0094] Step S221: removing abnormal data from the sensor process acquisition optimization data according to the printing process abnormal data to obtain the sensor process acquisition normal data;

[0095] In an embodiment of the present invention, by using the printing process abnormality data generated in step S215, abnormal data points, i.e., abnormal process acquisition data, are identified. These data points are affected by various abnormal factors in the printing process and need to be eliminated to ensure the accuracy of subsequent analysis. According to the results of abnormal data identification, abnormal data points are eliminated from the sensor process acquisition optimization data. The elimination can be performed manually or automatically, such as a threshold-based method, anomaly detection algorithm, etc. The sensor process acquisition optimization data after eliminating the abnormal data is used as a normal data set. These data sets do not contain abnormal data points, reflecting the state of the process parameters under normal conditions in the printing process.

[0096] Step S222: performing direct feature extraction on normal data collected by the sensor process to obtain direct feature data of process printing;

[0097] In the embodiment of the present invention, the direct features of process printing are extracted, and these features should be able to fully reflect the state and performance of the printing process. Specific direct features include ink flow, printing speed, temperature, humidity, etc. Feature extraction is performed on the normal data collected by the sensor process, and various statistics of each feature in the time series are calculated based on the pre-selected features, such as mean, standard deviation, maximum value, minimum value, etc. This can be achieved through basic statistical analysis methods. The extracted features are combined into feature vectors, and each feature vector represents the state of the printing process within a period of time. These feature vectors constitute a process printing direct feature data set, and the generated process printing direct feature data are recorded and stored.

[0098] Step S223: performing indirect feature conversion on the process printing direct feature data to generate process printing indirect feature data;

[0099] In an embodiment of the present invention, by selecting a feature conversion method, such as principal component analysis (PCA), linear discriminant analysis (LDA), t-distribution neighborhood embedding (t-SNE), etc., the direct feature data of process printing is standardized to ensure that each feature has the same scale and range. This can be achieved through methods such as mean normalization and standard deviation normalization. Using the selected feature conversion method, the standardized direct feature data of process printing is converted and calculated, and mapped to a new feature space. This new feature space has a higher dimension, or better captures the structural information of the original features. The converted feature data is used as indirect feature data of process printing, that is, a new feature representation is generated. These indirect feature data can capture information that is difficult to capture in the original feature space, which helps to improve the data expression ability and model performance.

[0100] Step S224: perform indirect feature pattern recognition on the process printing indirect feature data using signal processing technology to generate process printing indirect feature pattern recognition data; perform numerical simulation on the process printing indirect feature pattern recognition data to generate second printing process data.

[0101] In the embodiment of the present invention, by extracting feature patterns from indirect feature data of process printing, signal processing technology (such as wavelet transform, spectrum analysis, etc.) or pattern recognition algorithm (such as support vector machine, neural network, etc.) is used to extract pattern information in the data. The extracted feature patterns are identified and classified using the selected pattern recognition algorithm, and the process printing indirect feature data are divided into different pattern categories. According to the pattern recognition results, a corresponding pattern category tag is added to each data point to form a process printing indirect feature pattern recognition data set. Numerical simulation modeling is performed using the obtained process printing indirect feature pattern recognition data. This can be done using methods such as physical models, statistical models or machine learning models. The parameters in the model are estimated and optimized to fit the observed process printing indirect feature pattern recognition data to the greatest extent. Based on the model and the estimated parameters, numerical simulation is performed to generate second printing process data, which reflects the printing process state and performance under simulation conditions. The specific second printing process data is related to the printing process, but cannot be directly measured by means such as sensors. These data are obtained by analyzing, processing or inferring the direct printing process data, or are data indirectly related to printing quality, product performance, etc. For example, the color accuracy of the printed product, the uniformity of ink distribution, the surface smoothness of the printed product, etc.

[0102] Preferably, step S24 includes the following steps:

[0103] Step S241: performing data dimensionality reduction on the printing process multidimensional feature data to generate a printing process multidimensional feature vector;

[0104] Step S242: constructing a printing process association network model for the printing process multi-dimensional feature vector based on nonlinear modeling technology to generate a printing process association network model;

[0105] Step S243: importing the multi-dimensional feature vector of the printing process into the printing process association network model to capture the nonlinear effect of the data and generate printing process optimization data.

[0106] The present invention can reduce data dimensions and improve the efficiency of data processing and modeling by performing data dimensionality reduction on multidimensional characteristic data of printing process, while retaining the main information, which is helpful to better understand and explain the characteristics of printing process. Based on nonlinear modeling technology, a printing process association network model is constructed for the multidimensional characteristic vector of printing process, which can capture the complex correlation relationship and nonlinear effect in printing process, and provide a more accurate model basis for optimization. Based on nonlinear modeling technology, a printing process association network model is constructed for the multidimensional characteristic vector of printing process, which can capture the complex correlation relationship and nonlinear effect in printing process, and provide a more accurate model basis for optimization.

[0107] In an embodiment of the present invention, a suitable data dimensionality reduction method is selected according to the characteristics and requirements of the data. Common dimensionality reduction methods include principal component analysis (PCA), linear discriminant analysis (LDA), t-distribution neighborhood embedding (t-SNE), etc. The multidimensional feature data of the printing process is input into the selected dimensionality reduction method to perform data dimensionality reduction processing. This can be achieved by linear transformation, projection, etc., mapping high-dimensional data into low-dimensional space. Features are extracted from the reduced data to generate a multidimensional feature vector of the printing process. These feature vectors have lower dimensions, but still retain the main features of the original data. Suitable nonlinear modeling techniques, such as neural networks, support vector machines, decision trees, etc., are selected to construct a printing process association network model. The multidimensional feature vector of the printing process is used as input, and the printing process association network model is constructed using the selected nonlinear modeling technology. In the process of model construction, it is necessary to determine the structure, parameters and training method of the model. The multidimensional feature vector of the printing process is used as input and input into the constructed printing process association network model. In the printing process association network model, nonlinear modeling technology is used to capture the complex relationship and nonlinear effect between data. The model processes and analyzes the input features to generate printing process optimization data. Based on the output of the model, printing process optimization data is generated. These data reflect the correlation and nonlinear effects between printing process parameters and can be used for process optimization and adjustment.

[0108] Preferably, step S3 comprises the following steps:

[0109] Step S31: Obtaining printing process production requirement parameters;

[0110] Step S32: defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function;

[0111] Step S33: optimizing and iterating the printing process production objective function based on the printing process optimization data until the optimal printing process iteration data is generated;

[0112] Step S34: using the printing process optimization iteration data to perform printing adaptive control on the printing process, and generating a printing process adaptive control strategy.

[0113] The present invention can clarify the goals and requirements required in the production process by obtaining the required parameters of the printing process production, and provide an accurate reference for subsequent optimization and control. By defining the objective function, the printing process production requirements are converted into a mathematical model, the objectives required to be optimized in the optimization process are clarified, and guidance is provided for subsequent optimization iterations. Through the iterative optimization process, the printing process parameters can be continuously adjusted so that the objective function is gradually optimized and close to the optimal solution. This can improve printing quality, reduce costs, and improve production efficiency. Through optimization iteration, optimal printing process iteration data can be obtained, which represent the optimal parameter configuration under given production requirements and can be used to guide the actual printing process. The printing process adaptive control monitors and adjusts the printing parameters in real time to keep them consistent with the optimized iteration data, thereby realizing automatic control and regulation of the printing process. Through the printing process adaptive control, real-time optimization and adjustment of the printing process can be achieved, ensuring that the printing parameters are always kept in the best state, and improving the stability and consistency of the printing quality.

[0114] In an embodiment of the present invention, the production requirements of the printing process are determined, and the requirements include printing quality, cost, production capacity, manufacturing time and other aspects. Collect and record the parameters of the printing process production requirements, such as printing quality indicators, process parameter ranges, operating constraints, etc. According to the parameters of the printing process production requirements, determine the goals that need to be optimized, such as minimizing ink consumption, maximizing printing speed, optimizing color accuracy, etc. According to the selected optimization target, it is converted into a mathematical model to establish a printing process production objective function. This function will be used as an evaluation criterion for the optimization algorithm. According to the actual situation, select a suitable optimization algorithm, such as a genetic algorithm, a particle swarm algorithm, a simulated annealing algorithm, etc. Initialize the initial value of the parameter according to the scope and constraints of the printing process. Use the selected optimization algorithm to iteratively optimize the printing process parameters to minimize or maximize the objective function. Set convergence conditions, such as stopping the iteration when the maximum number of iterations is reached or the objective function changes less than the set threshold. Record the optimal printing process parameter configuration as the optimal printing process iteration data.

[0115] Preferably, step S4 comprises the following steps:

[0116] Step S41: outputting a first control parameter for the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously recording the control output time to obtain the first control parameter and the first control output time; performing printing performance evaluation on the first control parameter to generate first control performance data;

[0117] Step S42: outputting a second control parameter for the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously recording the control output time to obtain the second control parameter and the second control output time; performing printing performance evaluation on the second control parameter to generate second control performance data;

[0118] Step S43: weighting the control response difference between the first control output time and the second control output time to obtain a printing process response weight; weighting the control performance difference between the first control performance data and the second control performance data to obtain a printing process performance weight;

[0119] Step S44: Perform a comprehensive evaluation of printing optimization based on the printing process performance weight and the printing process response weight, and construct a printing process optimization report to evaluate the optimization of printing process parameters.

[0120] The present invention generates a first control parameter according to the optimal printing process iteration data and an adaptive control strategy, and performs a printing performance evaluation on the first control parameter, so as to understand the performance and effect of the first control parameter in actual printing. By generating a second control parameter according to the optimal printing process iteration data and an adaptive control strategy, and performing a printing performance evaluation on the second control parameter, the performance and effect of the second control parameter in actual printing can be understood. By weighting the control response difference between the first control output time and the second control output time, and weighting the control performance difference between the first control performance data and the second control performance data, the printing process response weight and the printing process performance weight can be determined for subsequent comprehensive evaluation of printing optimization. By performing a comprehensive evaluation of printing optimization according to the printing process performance weight and the printing process response weight, the response and performance differences of the printing process can be comprehensively considered to obtain a comprehensive evaluation result of the printing process. Generating a printing process optimization report can provide decision makers and relevant personnel with the ability to evaluate the effect and impact of printing process parameter optimization.

[0121] In the embodiment of the present invention, the first control parameter is output according to the printing process adaptive control strategy, that is, certain parameters in the printing process are adjusted to achieve optimization. The output time of the first control parameter is recorded synchronously for use in the subsequent evaluation of the control response. The adjusted printing process is evaluated for performance, including evaluations of printing quality, production efficiency, cost, etc., thereby generating first control performance data. Similar to step S41, the second control parameter is output according to the printing process adaptive control strategy, and the printing process is adjusted again. The output time of the second control parameter is recorded synchronously for subsequent control response difference weighting. The adjusted printing process is evaluated for performance and second control performance data is generated. According to the difference between the first control output time and the second control output time, the control responses of the two adjustments are weighted to reflect the sensitivity and efficiency of the control. According to the difference between the first control performance data and the second control performance data, the control performance of the two adjustments is weighted to reflect the changes in printing quality and production efficiency. According to the control response weight and the control performance weight, the comprehensive performance weight of the printing process is calculated. Combined with the weight information, the optimization effect of the printing process is comprehensively evaluated to generate a printing process optimization report. The report includes optimization effects, improvement suggestions, etc., which are used to evaluate the effect and effectiveness of the optimization of printing process parameters.

[0122] In this specification, a printing process parameter optimization system is provided, which is used to execute the above-mentioned printing process parameter optimization method. The printing process parameter optimization system includes:

[0123] A sensor acquisition optimization module is used to obtain printing process area type data; perform printing area parameter acquisition based on the printing process area type data, thereby obtaining first printing process data; perform first parameter calibration on the first printing process data to generate first printing process calibration data; perform sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data;

[0124] A nonlinear association module is used to perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; screen the second printing process data on the sensor process acquisition optimization data according to the abnormal printing process data to obtain the second printing process data; perform complex nonlinear association modeling on the second printing process data to generate predicted printing process data;

[0125] The adaptive control module is used to obtain the required parameters of printing process production; define the objective function according to the required parameters of printing process production to obtain the printing process production objective function; perform printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy;

[0126] The optimization evaluation module is used to output the first control parameter of the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously record the control output time to obtain the first control parameter and the first control output time; output the second control parameter of the optimal printing process iteration data, and synchronously record the control output time to obtain the second control parameter and the second control output time; weight the control response difference between the first control output time and the second control output time to obtain the printing process response weight; weight the control performance difference between the first control parameter and the second control parameter to obtain the printing process performance weight; perform a comprehensive evaluation of printing optimization based on the printing process performance weight and the printing process response weight, and construct a printing process optimization report to evaluate the optimization of printing process parameters.

[0127] The beneficial effect of the present invention is that by collecting and analyzing printing process data, various parameters and variables in the printing process can be understood more accurately, so as to perform effective optimization. Abnormal analysis can detect abnormal conditions in the printing process in a timely manner, so that corresponding strategies can be adopted for adjustment and correction. Through complex nonlinear association modeling, key parameters and variables in the printing process can be predicted, which helps to prepare and adjust in advance. The adaptive control strategy based on printing process optimization data can adjust the control parameters in real time, adapt to different printing conditions and requirements, and improve printing quality and efficiency. By comprehensively evaluating the performance and response weights of the printing process, comprehensive optimization suggestions and improvement plans can be provided to decision makers to promote the continuous improvement of the printing process. Therefore, the present invention realizes the prediction of printing process data by calibrating sensor parameters and establishing a complex nonlinear association model, thereby improving the printing quality and production efficiency after optimizing the parameters.

[0128] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0129] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A printing process parameter optimization method, characterized in that: The following steps are involved: Step S1: Acquire printing process area type data; Collect printing area parameters based on the printing process area type data, thereby obtaining first printing process data; Performing first parameter calibration on the first printing process data to generate first printing process calibration data; Performing sensor acquisition optimization on the sensor based on the first sensor process calibration data, thereby obtaining sensor process acquisition optimization data; Step S1 includes the following steps: Step S11: acquiring printing process area type data; deploying sensors in the area based on the printing process area type data, and collecting printing area parameters using the deployed sensors, thereby acquiring first printing process data, wherein the first printing process data includes printing ink flow, printing speed, printing temperature and printing humidity; Step S12: performing data preprocessing on the first printing process data to generate first printing process cleaning data, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization; Step S13: performing signal filtering on the first printing process cleaning data to generate first printing process filtering data; Step S14: performing first parameter calibration on the first printing process filter data to generate first printing process calibration data; performing sensor acquisition optimization on the sensor based on the first sensor process calibration data to obtain sensor process acquisition optimization data; Step S2: Perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; perform second printing process data screening on the sensor process acquisition optimization data according to the abnormal printing process data to obtain second printing process data; perform complex nonlinear correlation modeling on the second printing process data to generate predicted printing process data; Step S2 includes the following steps: Step S21: Perform printing process high abnormality analysis on the sensor process collection optimization data to generate printing process abnormality data; Step S21 includes the following steps: Step S211: Screening abnormal parameters of sensor process acquisition optimization data to obtain abnormal process acquisition data; Step S212: performing abnormal data collection timing analysis on the abnormal process collection data to generate an abnormal data collection timestamp; Step S213: performing abnormal collection time interval analysis on the abnormal process collection data according to the abnormal data collection timestamp to generate the abnormal collection time interval; Step S214: Compare the abnormal collection time interval with the preset abnormal time interval threshold. When the abnormal collection time interval is greater than or equal to the preset abnormal time interval threshold, integrate the timestamps within the corresponding abnormal collection time interval and mark it as a high abnormality frequent period. Step S215: performing printing process abnormality analysis on the abnormal process collection data based on the high abnormality frequency period to generate printing process abnormality data, wherein the printing process abnormality analysis includes printing process abnormality type analysis, printing process abnormality location analysis and printing process abnormality degree analysis; Step S22: Screening the sensor process collection optimization data for second printing process data according to the printing process abnormality data to obtain second printing process data; Step S22 includes the following steps: Step S221: removing abnormal data from the sensor process acquisition optimization data according to the printing process abnormal data to obtain the sensor process acquisition normal data; Step S222: performing direct feature extraction on normal data collected by the sensor process to obtain direct feature data of process printing; Step S223: performing indirect feature conversion on the process printing direct feature data to generate process printing indirect feature data; Step S224: performing indirect feature pattern recognition on the process printing indirect feature data using signal processing technology to generate process printing indirect feature pattern recognition data; performing numerical simulation on the process printing indirect feature pattern recognition data to generate second printing process data; Step S23: extracting multi-dimensional printing features from the second printing process data to generate printing process multi-dimensional feature data; Step S24: Perform complex nonlinear correlation modeling on the multi-dimensional feature data of the printing process to generate predicted printing process data; Step S24 includes the following steps: Step S241: performing data dimensionality reduction on the printing process multidimensional feature data to generate a printing process multidimensional feature vector; Step S242: constructing a printing process association network model for the printing process multi-dimensional feature vector based on nonlinear modeling technology to generate a printing process association network model; Step S243: importing the printing process multidimensional feature vector into the printing process association network model to capture the data nonlinear effect and generate predicted printing process data; Step S3: obtaining printing process production requirement parameters; defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function; performing printing adaptive control on the printing process production objective function based on the predicted printing process data to generate a printing process adaptive control strategy; Step S4: Based on the printing process adaptive control strategy, the first control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the first control parameter and the first control output time; the second control parameter is output for the optimal printing process iteration data, and the control output time is synchronously recorded to obtain the second control parameter and the second control output time; the first control output time and the second control output time are weighted for the control response difference to obtain the printing process response weight; the first control parameter and the second control parameter are weighted for the control performance difference to obtain the printing process performance weight; a comprehensive evaluation of printing optimization is performed based on the printing process performance weight and the printing process response weight, and a printing process optimization report is constructed to evaluate the optimization of printing process parameters.

2. The printing process parameter optimization method according to claim 1, characterized in that: Step S14 includes the following steps: Step S141: performing data error analysis on the first printing process filter data to generate first printing process error data; Step S142: performing calibration curve fitting using the first printing process error data to generate a first printing process calibration curve; Step S143: calibrating the actual parameters of the first printing process filter data according to the first printing process calibration curve to generate first printing process calibration data; Step S144: performing acquisition optimization on the sensor based on the first printing process calibration data to obtain an optimized calibration sensor, wherein the parameter acquisition optimization includes sensor position optimization, sampling frequency optimization and gain optimization; Step S145: Optimize and collect printing area parameters by optimizing and calibrating the sensor, thereby generating sensor process collection optimization data.

3. The printing process parameter optimization method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: Obtaining printing process production requirement parameters; Step S32: defining the objective function according to the printing process production requirement parameters to obtain the printing process production objective function; Step S33: optimizing and iterating the printing process production objective function based on the printing process optimization data until the optimal printing process iteration data is generated; Step S34: using the printing process optimization iteration data to perform printing adaptive control on the printing process, and generating a printing process adaptive control strategy.

4. The printing process parameter optimization method according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: outputting a first control parameter for the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously recording the control output time to obtain the first control parameter and the first control output time; performing printing performance evaluation on the first control parameter to generate first control performance data; Step S42: outputting a second control parameter for the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously recording the control output time to obtain the second control parameter and the second control output time; performing printing performance evaluation on the second control parameter to generate second control performance data; Step S43: weighting the control response difference between the first control output time and the second control output time to obtain a printing process response weight; weighting the control performance difference between the first control performance data and the second control performance data to obtain a printing process performance weight; Step S44: Perform a comprehensive evaluation of printing optimization based on the printing process performance weight and the printing process response weight, and construct a printing process optimization report to evaluate the optimization of printing process parameters.

5. A printing process parameter optimization system, characterized in that: Used to execute the printing process parameter optimization method according to claim 1, the printing process parameter optimization system comprises: A sensor acquisition optimization module is used to obtain printing process area type data, including printing head, edge, and center; based on the printing process area type data, perform printing area parameter acquisition to obtain first printing process data, wherein the first printing process data includes printing ink flow, printing speed, printing temperature, and printing humidity; perform first parameter calibration on the first printing process data to generate first printing process calibration data, including calibration parameter values ​​and calibration process records; based on the first sensor process calibration data, perform sensor acquisition optimization on the sensor to obtain sensor process acquisition optimization data, wherein the sensor acquisition optimization is to optimize the deployment and acquisition strategy of the sensor; The nonlinear association module is used to perform high abnormality analysis of the printing process on the sensor process acquisition optimization data to generate abnormal printing process data; screen the second printing process data on the sensor process acquisition optimization data according to the abnormal printing process data to obtain the second printing process data; perform complex nonlinear association modeling on the second printing process data to generate predicted printing process data; wherein the step of performing complex nonlinear association modeling on the second printing process data includes: Perform printing process abnormality analysis on the sensor process collection optimization data to generate printing process abnormality data; Screening the sensor process collection optimization data for second printing process data according to the printing process abnormality data to obtain second printing process data; Perform multi-dimensional printing feature extraction on the second printing process data to generate multi-dimensional feature data of the printing process Performing data dimension reduction on the multi-dimensional feature data of the printing process to generate a multi-dimensional feature vector of the printing process; Based on nonlinear modeling technology, a printing process correlation network model is constructed for the multi-dimensional feature vector of the printing process to generate a printing process correlation network model; Import the multi-dimensional feature vector of printing process into the printing process association network model to capture the nonlinear effect of data and generate predicted printing process data; The adaptive control module is used to obtain the required parameters of printing process production; define the objective function according to the required parameters of printing process production to obtain the printing process production objective function; perform printing adaptive control on the printing process production objective function based on the printing process optimization data to generate a printing process adaptive control strategy; The optimization evaluation module is used to output the first control parameter of the optimal printing process iteration data based on the printing process adaptive control strategy, and synchronously record the control output time to obtain the first control parameter and the first control output time; output the second control parameter of the optimal printing process iteration data, and synchronously record the control output time to obtain the second control parameter and the second control output time; weight the control response difference between the first control output time and the second control output time to obtain the printing process response weight; weight the control performance difference between the first control parameter and the second control parameter to obtain the printing process performance weight; perform a comprehensive evaluation of printing optimization based on the printing process performance weight and the printing process response weight, and construct a printing process optimization report to evaluate the optimization of printing process parameters.

Images