Intelligent optimization control system for process parameters of photovoltaic encapsulation adhesive film hot press molding

Abstract

The application provides a photovoltaic encapsulating adhesive film hot-press forming process parameter intelligent optimization control system, relates to the photovoltaic material forming processing technical field, and comprises the following steps: extracting multiple process flow nodes, determining multiple process parameters; obtaining a process parameter influence information base; performing multiple hot-press forming simulations to obtain multiple adhesive film quality data; performing regression analysis, constructing a response surface model, performing optimization, and obtaining a first hot-press forming process parameter group; adjusting the response surface model according to the monitoring result to generate a response surface optimization model; performing intelligent optimization control and outputting a second hot-press forming process parameter group. Through the application, the technical problem that product quality is unstable and production efficiency is further affected due to inaccurate process parameter adjustment in the prior art can be solved; through the establishment of the process parameter influence information base and the response surface model, the intelligent optimization control of the process parameters is realized, and the product quality and production efficiency of the photovoltaic encapsulating adhesive film are improved.

Description

Technical Field

[0001] This application relates to the field of photovoltaic material molding and processing technology, and in particular to an intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters. Background Technology

[0002] With the rapid development of the photovoltaic industry, the manufacturing process of photovoltaic modules has increasingly become a key factor in improving photovoltaic efficiency and reducing costs. In the manufacturing process of photovoltaic modules, the thermosetting process of the encapsulating film is a crucial step. Photovoltaic cell encapsulating film (EVA) is a thermosetting adhesive film used to sandwich laminated glass. The main purpose of this process is to ensure a good bond between the photovoltaic cell and the encapsulating material, thereby improving the long-term stability and weather resistance of the module. Existing photovoltaic encapsulating film thermosetting processes typically rely on experience and trial and error, lacking systematic analysis and adjustment methods. The adjustment of process parameters is often imprecise, leading to unstable product quality and affecting the performance and lifespan of the photovoltaic modules.

[0003] In summary, existing technologies suffer from technical problems such as inaccurate adjustment of process parameters, leading to unstable product quality and further affecting production efficiency. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent optimization and control system for the hot pressing process parameters of photovoltaic encapsulation films, in order to solve the technical problem in the prior art that the inaccurate adjustment of process parameters leads to unstable product quality and further affects production efficiency.

[0005] In view of the above problems, this application provides an intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters. The intelligent optimization control system includes: a process node analysis module, used to extract multiple process flow nodes based on the photovoltaic encapsulation film hot pressing process flow, and determine multiple process parameters based on the multiple process flow nodes; an influence information database construction module, used to obtain a process parameter influence information database, which is obtained by analyzing the influence of the multiple process flow nodes based on the multiple process parameters; a hot pressing simulation module, used to perform multiple hot pressing simulations based on the process parameter influence information database to obtain multiple film quality data of the photovoltaic encapsulation film; a regression analysis module, used to perform regression analysis using the multiple film quality data, construct a response surface model, and optimize the response surface model to obtain a first set of hot pressing process parameters; a model adjustment module, used to monitor the execution of the first set of hot pressing process parameters, and adjust the response surface model based on the monitoring results to generate a response surface optimization model; and an intelligent optimization control module, used to perform intelligent optimization control based on the response surface optimization model and output a second set of photovoltaic encapsulation film hot pressing process parameters.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The system employs a process node analysis module to extract multiple process nodes based on the hot-pressing process of photovoltaic encapsulation films, and determines multiple process parameters based on these nodes. An impact information database construction module is used to obtain a database of process parameter impact information, derived from impact analysis of the multiple process nodes based on the aforementioned process parameters. A hot-pressing simulation module performs multiple hot-pressing simulations based on the database to obtain multiple film quality data for the photovoltaic encapsulation films. A regression analysis module uses the film quality data for regression analysis to construct a response surface model, which is then optimized to obtain a first set of hot-pressing process parameters. A model adjustment module monitors the execution of the first set of hot-pressing process parameters and adjusts the response surface model based on the monitoring results to generate an optimized response surface model. An intelligent optimization control module performs intelligent optimization control based on the optimized response surface model, outputting a second set of hot-pressing process parameters for the photovoltaic encapsulation films. In other words, by establishing a process parameter impact information database and a response surface model, intelligent optimization control of process parameters is achieved, improving the product quality and production efficiency of photovoltaic encapsulation films.

[0008] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of the intelligent optimization control system for the photovoltaic encapsulation film hot pressing process parameters of this application.

[0011] Figure 2This is a flowchart illustrating the output response surface model in the intelligent optimization control system for the hot pressing process parameters of the photovoltaic encapsulation film in this application.

[0012] Figure labeling: 11 Process node analysis module, 12 Impact information database construction module, 13 Hot pressing molding simulation module, 14 Regression analysis module, 15 Model adjustment module, 16 Intelligent optimization control module. Detailed Implementation

[0013] This application provides an intelligent optimization and control system for the hot pressing process parameters of photovoltaic encapsulation films, solving the technical problem in existing technologies where inaccurate adjustment of process parameters leads to unstable product quality and further affects production efficiency. By establishing a database of process parameter influence information and a response surface model, intelligent optimization and control of process parameters are achieved, improving the product quality and production efficiency of photovoltaic encapsulation films.

[0014] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0015] For examples, please refer to the appendix. Figure 1 This application provides an intelligent optimization and control system for photovoltaic encapsulation film hot pressing process parameters, wherein the intelligent optimization and control system for photovoltaic encapsulation film hot pressing process parameters includes:

[0016] The process node analysis module 11 is used to extract multiple process flow nodes based on the hot pressing molding process of photovoltaic encapsulation film, and determine multiple process parameters based on the multiple process flow nodes.

[0017] Specifically, the process flow nodes in the thermoforming of photovoltaic encapsulating films are identified and determined, including preheating, pressurization, heating, heat preservation, and cooling. Based on the extracted process flow nodes, key process parameters affecting these nodes are determined, including temperature, pressure, time, heating rate, and cooling rate. Preheating involves initially heating the film before formal pressurization to increase its plasticity. Pressurization applies pressure to the film to adhere it to the photovoltaic module. Heating refers to heating the film simultaneously with pressurization to promote adhesion and curing. Heat preservation involves maintaining a certain temperature and time to ensure complete curing of the film. Cooling involves cooling the cured film to room temperature for subsequent processing. By accurately extracting process flow nodes and determining process parameters, the quality of thermoforming photovoltaic encapsulating films can be improved. Reasonable parameter settings ensure that the film achieves optimal plasticity during the forming process, thereby improving the adhesive strength and durability of the product.

[0018] The impact information database construction module 12 is used to obtain a process parameter impact information database, which is obtained by performing impact analysis on the multiple process flow nodes based on the multiple process parameters.

[0019] Specifically, appropriate statistical or machine learning methods are selected to determine the degree and patterns of influence of each process parameter on multiple process flow nodes. Individual or combined impact analyses are performed on each process parameter. The degree of influence of each parameter on the process flow nodes is evaluated. Whether there are interactions between different parameters, i.e., whether the effect of one parameter depends on the value of another, is determined. Mathematical models are built using the collected data to describe the relationship between process parameters and process flow nodes. The analysis results are compiled into a structured database or knowledge base, namely a process parameter influence information database, containing the degree of influence of multiple process parameters on various process flow nodes. The process parameter influence information database includes, but is not limited to, parameter names, parameter ranges, degree of influence, and interactions. Constructing a process parameter influence information database helps to gain a deeper understanding of the impact of different parameters on process flow nodes and predict the process effects under different parameter settings.

[0020] The hot pressing molding simulation module 13 is used to perform multiple hot pressing molding simulations based on the process parameter influence information database to obtain multiple film quality data of photovoltaic encapsulation film.

[0021] Specifically, appropriate simulation tools, such as professional CAD software or custom simulation programs developed for specific processes, are selected to create simulation models that simulate the hot-pressing process of photovoltaic encapsulation films. A database of process parameter influences is used to simulate the process under different combinations of process parameters, recording the process parameters and corresponding film quality indicators for each simulation, such as transparency and the number of bubbles. The number of bubbles inside the film is measured; excessive bubbles will affect the film's performance and the module's appearance. Through multiple hot-pressing simulations, film quality data under different process parameters is obtained, which helps to deeply understand the impact of each parameter on film performance.

[0022] The regression analysis module 14 is used to perform regression analysis using the multiple film quality data, construct a response surface model, and perform optimization through the response surface model to obtain the first hot pressing molding process parameter set.

[0023] Specifically, regression analysis was performed on data obtained from multiple hot-pressing simulation experiments. Regression analysis is a statistical method used to study the relationships between variables. Regression functions were used to perform regression analysis on the pre-processed film quality standard data. Multiple regression coefficients were obtained through regression analysis, reflecting the degree of influence of different process parameters on film quality. Parameter response correlation analysis was performed using the obtained regression coefficients, and a response surface plot was drawn to show the influence of different combinations of process parameters on film quality. The response surface model was validated using the response surface plot. If the model passed validation, a validation token was generated, and the validated response surface model, which describes the influence of different combinations of process parameters on film quality, was output. After constructing the response surface model, the system optimized the model, i.e., found the process parameter combination that achieves the best film quality, i.e., the first hot-pressing process parameter set. The optimization process included using optimization algorithms (such as gradient descent, genetic algorithms, etc.) to adjust parameter values ​​to minimize or maximize specific quality indicators. For example, suppose there is a set of data containing three process parameters (temperature, pressure, and time) and one response variable (film transparency). A response surface methodology (RSM) model was constructed using regression analysis to predict transparency under different parameter combinations. Then, an optimization algorithm was used to refine the model, ultimately yielding the following parameter set: temperature 85°C, pressure 1.8 MPa, and time 6 minutes. By constructing and optimizing the RSM model through regression analysis, the optimal combination of process parameters was accurately identified to achieve the best film quality, which helps improve the hot-pressing quality and production efficiency of photovoltaic encapsulation films.

[0024] The model adjustment module 15 is used to monitor the first hot pressing process parameter group, adjust the response surface model according to the monitoring results, and generate an optimized response surface model.

[0025] Specifically, based on the hot pressing molding process flow, key process parameters requiring monitoring are determined, and monitoring points are set at critical locations in the process flow to ensure comprehensive coverage of the entire process. Sensors and instruments are used to collect process parameter data at the monitoring points in real time, as well as monitoring response variable values ​​directly related to the process parameters, such as film transparency and bubble count. Through monitoring, the first response variable value is obtained. Trend analysis is performed based on the first response variable value to identify abnormal fluctuation data, i.e., response variable values ​​that do not conform to expectations or standard ranges. Abnormal fluctuation data is added to the monitoring results. The monitoring results are a data set recording process parameters, response variable values, and their changing trends. Correlation analysis is performed between the abnormal fluctuation data and the first set of hot pressing molding process parameters to determine which parameters have a strong correlation with the abnormal fluctuation data. A response surface model is used to predict the first set of hot pressing molding process parameters, generating predicted response values. Based on the correlation coefficient, the predicted response values ​​are compared with the actual abnormal fluctuation data, and error analysis is performed to generate response error values. The response surface model is diagnosed based on the response error values, identifying multiple problem areas that may be areas with large prediction errors in the model. The response surface model is refitted for the identified problem areas, and the model parameters are optimized. If the model performance is significantly improved, the optimized response surface model is output. The response surface model is continuously optimized by monitoring the first set of hot pressing process parameters and adjusting the model based on the monitoring results to better adapt it to the actual production environment.

[0026] The intelligent optimization control module 16 is used to perform intelligent optimization control based on the response surface optimization model and output the second hot pressing process parameter set of the photovoltaic encapsulation film.

[0027] Specifically, intelligent optimization control is achieved using a response surface methodology (RSM) model to predict the film quality under different combinations of process parameters, and the process parameters are automatically adjusted based on these predictions. Through multiple iterations and adjustments, the combination of process parameters that maximizes product performance or minimizes cost is found, outputting a second set of hot-pressing process parameters for photovoltaic encapsulation films, including optimal combinations of temperature, pressure, and time, for use in actual production. By utilizing the RSM model for intelligent optimization control, real-time parameter adjustments are achieved during the production process to adapt to changes in production conditions and product quality requirements, which helps improve the hot-pressing quality and production efficiency of photovoltaic encapsulation films.

[0028] Furthermore, the regression analysis module 14 in the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters is also used for:

[0029] The response surface model is used to preprocess the multiple film quality data to generate multiple film quality standard data. A regression function is introduced to perform regression analysis on the multiple film quality standard data to obtain multiple regression fitting coefficients. The multiple regression fitting coefficients are used to perform parameter response correlation analysis and draw a response surface plot. The response surface plot is used to validate the response surface model, and the response surface model is output according to the validation token.

[0030] Specifically, film quality data is obtained from multiple hot-pressing simulation experiments. The data undergoes cleaning and standardization, including data cleaning, normalization, and transformation, to ensure data quality and consistency. After processing, multiple film quality standard data conforming to certain standards and quality requirements are obtained. A regression function is introduced to perform regression analysis on the multiple film quality standard data. A multivariate regression function is constructed, and the multiple film quality standard data are input into the regression function to find the quantitative relationship between process parameters and film quality, obtaining multiple regression fitting coefficients that describe the relationship between independent variables (such as process parameters) and dependent variables (such as film quality). Using the regression fitting coefficients and process parameter data, the response variable values ​​under different parameter combinations are analyzed to determine multiple response vectors. These response vectors are mapped into a matrix, forming multiple response matrices. The rows of each matrix represent different combinations of process parameters, and the columns represent different response variables. It is determined whether the columns of these response matrices are linearly independent to ensure that the data used does not produce multicollinearity. If the columns of the response matrices are linearly independent, response surface plots are constructed based on these matrices. Response surface plots present the relationship between independent variables (process parameters) and dependent variables (response variables) in the form of 3D surface plots or contour plots.

[0031] Multiple process parameters are predicted using response surface methodology (RSM) and standard encapsulation film quality data to obtain predicted values ​​for the process parameter response variables. These predicted values ​​are then used as indices for retrieval on the RSM. The corresponding response variable values ​​are retrieved based on the indices, and a trend graph of the response variables as a function of the process parameters is constructed. Based on the response trend, peaks and troughs of multiple parameters are identified, representing the range of key process parameters affecting encapsulation film quality. These extreme points are used to cross-validate the RSM model. If the model performs well during validation (i.e., the predicted values ​​are highly consistent with the actual values), a validation token is generated. The validated RSM model is output to predict encapsulation film quality under different combinations of process parameters. Analyzing and predicting the impact of different process parameter combinations on encapsulation film quality allows for optimization of process parameters, improving the hot-pressing quality and production efficiency of photovoltaic encapsulation films.

[0032] Furthermore, the intelligent optimization and control system for photovoltaic encapsulation film hot pressing process parameters also includes a fitting coefficient acquisition module, used for:

[0033] Based on the multiple film quality standard data, a random selection is made, and the regression function is constructed according to the data selection results: ;in, For multiple regression fitting coefficients corresponding to multiple film quality standard data, Let i be the i-th independent variable in multiple film quality standard data. Let j be the j-th independent variable in multiple film quality standard data. This is the intercept term, representing the expected value of the response variable when all independent variables are 0. and The corresponding linear coefficients represent right The linear effect, It is a quadratic term. Is with The corresponding coefficients represent right The impact, Let i be an interactive term, where i ≠ j. Is with The corresponding coefficients represent and The interaction between them The impact, The number of independent variables is the total number of multiple film quality standard data; the multiple film quality standard data are synchronized to the regression function for regression analysis, and the multiple regression fit coefficients are output.

[0034] Specifically, a subset of sample data was randomly selected from multiple film quality standard data to construct a multiple regression function: Where y represents multiple regression fit coefficients. and Let i and j represent the i-th and j-th independent variables in multiple film quality standard data, respectively. The intercept term represents the expected value of the response variable when all independent variables are 0. These are linear coefficients, representing the independent variable. The direct impact on the response variable y; These are quadratic coefficients, representing the independent variable. The effect on the response variable y; These are the interaction term coefficients, representing the independent variables. and The interaction between the parameters affects the response variable y. p represents the number of independent variables, i.e., the total number of independent variables in the film quality standard data. Multiple film quality standard data are input into a regression function for regression analysis, yielding multiple regression fit coefficients. By synchronizing multiple film quality standard data with the regression function for regression analysis, multiple regression fit coefficients are output, thus accurately describing the relationship between process parameters and film quality.

[0035] Furthermore, the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters also includes a response surface plotting module, used for:

[0036] Response variable analysis is performed based on the multiple regression fitting coefficients and the multiple process parameters to determine multiple response vectors; correlation mapping is performed on the multiple response vectors to construct multiple response matrices; based on the multiple response matrices, it is determined whether multiple columns of the multiple response matrices are linearly independent; if the multiple columns of the multiple response matrices are linearly independent, the response surface plot is constructed based on the multiple response matrices.

[0037] Specifically, regression analysis is used to obtain fitting coefficients, which are then combined with process parameters to analyze the response variables. The response variables refer to performance indicators related to film quality, such as transparency and the number of bubbles. By analyzing the relationship between the regression fitting coefficients and process parameters, the response variable values ​​under different combinations of process parameters are predicted, forming multiple response vectors. Each vector represents a specific combination of process parameters and its corresponding response variable value. By mapping the response vectors to a matrix, the impact of different combinations of process parameters on the response variables is visually displayed. The response matrix is ​​a two-dimensional array, where rows represent different combinations of process parameters and columns represent different response variables. Statistical methods (such as principal component analysis and singular value decomposition) are used to determine whether the columns of the response matrix are linearly independent. Linear independence means that no single vector in a set can be represented as a linear combination of the other vectors. In statistics and data analysis, linear independence is crucial for building reliable and effective models. By determining whether multiple columns of multiple response matrices are linearly independent, it is ensured that the data used does not suffer from multicollinearity, thereby improving the accuracy and reliability of the model. If multiple columns of the response matrix are linearly independent, a response surface plot is constructed accordingly. Response surface methodology (RSM) plots are typically drawn in a three-dimensional coordinate system, where two independent variables (such as temperature and pressure) are plotted on the x and y axes, and a third independent variable (such as time) is plotted on the z-axis. Each point represents a specific combination of process parameters, while the color or height of the point indicates the corresponding response variable value (such as the transparency or adhesive strength of the adhesive film). Accurate analysis and prediction of the impact of different process parameter combinations on adhesive film quality, along with linear independence assessments, help avoid multicollinearity issues and improve the accuracy and reliability of the model.

[0038] Further details are attached. Figure 2 As shown, the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters also includes a response surface model output module, used for:

[0039] The response surface graph is used in conjunction with the multiple film quality standard data to predict the multiple process parameters, thereby obtaining predicted values ​​of the process parameter response variables. These predicted values ​​are then used as indexes to traverse the response surface graph for retrieval, and a response change trend is constructed based on the retrieval results. Extreme points are identified based on the response change trend to determine the peak and valley values ​​of multiple parameters. The response surface model is then cross-validated based on the peak and valley values ​​of the multiple parameters. If the validation passes, a validation token is generated, and the response surface model is output.

[0040] Specifically, response surface methodology (RSM) is used in conjunction with standard film quality data to predict the values ​​of response variables under different combinations of process parameters. The predicted values ​​of the process parameter response variables are used as indices to traverse the RSM. For each predicted value, the corresponding point on the RSM is located, and the response variable value at that point is recorded. The trend of the response variable is determined, i.e., the direction of change (increase or decrease) in film quality under a specific combination of process parameters. Points where the response variable reaches its maximum or minimum value are identified in the trend graph; these points are called extreme points. Extreme points can be peaks (points where the response variable reaches its maximum value) or valleys (points where the response variable reaches its minimum value). These extreme points are used to cross-validate the response surface model. Cross-validation is a statistical method used to evaluate the predictive performance of the model. The dataset is divided into a training set and a validation set. The training set is used to build the model, and the validation set is used to test the predictive accuracy of the model. When cross-validation passes, meaning the model's prediction accuracy on the validation set reaches a predetermined standard, it indicates the model has good predictive ability. A validation token is generated, signifying the model has passed validation and can be used for process optimization and decision-making in actual production. Subsequently, the validated response surface model is output to predict film quality under different combinations of process parameters. By predicting response change trends, identifying extreme points, and cross-validating the response surface model, the range of key process parameters affecting film quality is determined, ensuring the accuracy and reliability of the response surface model and enabling its use for process optimization and decision-making in actual production.

[0041] Furthermore, the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters also includes a response surface model optimization module, used for:

[0042] The optimization objective is set by combining the peak and trough values ​​of multiple parameters using the response surface model. The combined gradient of the multiple process parameters is calculated using gradient descent combined with the response surface model to determine multiple partial derivatives, which correspond to the multiple process parameters. The gradient direction is determined based on the multiple partial derivatives and the optimization objective. Following the gradient direction, the multiple partial derivatives are sequentially compared with a preset gradient threshold. Process parameters with partial derivatives less than the preset gradient threshold are extracted and added to the candidate parameter group, while process parameters with partial derivatives greater than or equal to the preset gradient threshold are extracted and added to the elimination parameter group. This process is iterated until all process parameters have been traversed. The elimination parameter group is deleted, and the candidate parameter groups are combined according to the hot pressing process flow to output the first hot pressing process parameter group.

[0043] Specifically, response surface methodology (RSM) is used to analyze the impact of peak and trough values ​​of multiple parameters on film quality. Based on the analysis results, optimization objectives are set. Optimization objectives typically maximize or minimize the values ​​of response variables (such as film transparency, bubble count, etc.), maximizing peak values ​​or minimizing trough values ​​to achieve optimal film quality or cost efficiency. Based on the RSM and extreme points, specific parameter values ​​or ranges need to be achieved during the optimization process to realize optimal film quality in actual production. Gradient descent is used to optimize process parameters. Gradient descent is a commonly used optimization algorithm that iteratively calculates the gradient (i.e., partial derivatives) of parameters and adjusts parameter values ​​according to the gradient direction to minimize the objective function (such as the prediction error of the response variable). Combined with the RSM, the partial derivatives of each process parameter with respect to the response variable are calculated, reflecting the parameter's sensitivity to the response variable.

[0044] Based on the optimization objective and the signs of the calculated partial derivatives, the gradient direction for parameter adjustment is determined, identifying which parameters should be increased or decreased to approach the optimization objective. If the optimization objective is to maximize the response variable (such as the transparency of the film), parameters with positive partial derivatives should have their values ​​increased, while those with negative partial derivatives should have their values ​​decreased. Following this gradient direction, each partial derivative is compared sequentially with a preset gradient threshold. The preset gradient threshold is used to distinguish the magnitude of the influence of process parameters on the response variable. Process parameters with partial derivatives less than the threshold are considered to have a smaller impact on the response variable and are added to the candidate parameter group. Conversely, process parameters with partial derivatives greater than or equal to the threshold are considered to have a larger impact on the response variable and are added to the elimination parameter group. The elimination parameter group is deleted because these parameters have a smaller impact on the response variable and do not require special attention during optimization. The parameters in the candidate parameter group are combined according to the hot pressing process flow to form one or more possible optimal process parameter combinations, outputting the first hot pressing process parameter group. By setting optimization goals and clarifying the optimization direction, the gradient descent method and the calculation of partial derivatives help to quickly find the optimal combination of process parameters. The parameter screening and iteration process helps to avoid unnecessary adjustments to parameters with minor impact, reduce the number of experiments, and accelerate the product development and process optimization process.

[0045] Furthermore, the model adjustment module 15 in the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters is also used for:

[0046] Multiple monitoring points are set up based on the hot pressing process flow. Data monitoring is performed on the first hot pressing process parameter group according to the multiple monitoring points to obtain the first response variable value. The first response variable value has a corresponding relationship with the first hot pressing process parameter group. Trend analysis is performed based on the first response variable value to determine abnormal fluctuation data. The abnormal fluctuation data is added to the monitoring results.

[0047] Specifically, the hot pressing process flow is analyzed, and multiple monitoring points are set to ensure coverage of key stages such as preheating, pressurization, heating, heat preservation, and cooling. Data is monitored for the first set of hot pressing process parameters according to these monitoring points, collecting process parameters from actual production, such as temperature, pressure, and time, as well as monitoring response variables directly related to these parameters, such as film transparency, bubble count, adhesive strength, and thickness. Through monitoring, the system obtains the values ​​of the first set of response variables. These values ​​correspond to the settings of the first set of hot pressing process parameters, reflecting the film quality under specific parameter combinations. The response variable data undergoes preprocessing such as cleaning and noise reduction to improve the accuracy of the analysis. Statistical methods (such as time series analysis and trend line fitting) are used to identify the trends of the response variables. By analyzing the changing trends of the response variable values, abnormal fluctuations are identified, i.e., those response variable values ​​that do not conform to the expected or standard range. The identified abnormal fluctuation data is added to the monitoring results. The monitoring results are a data set recording process parameters, response variable values, and their changing trends.

[0048] In a specific example, the following key process steps and parameters were identified as monitoring points: preheating temperature, pressure, hot pressing time, and hot pressing temperature. Data monitoring was conducted on the first set of hot pressing process parameters, and the response variable values ​​for each monitoring point were recorded: No. 1: Preheating temperature 120℃, pressure 1MPa, hot pressing time 1 minute, hot pressing temperature 150℃, transparency 90%; No. 2: Preheating temperature 150℃, pressure 1.5MPa, hot pressing time 1.5 minutes, hot pressing temperature 160℃, transparency 92%; No. 3: Preheating temperature 130℃, pressure 2MPa, hot pressing time 2 minutes, hot pressing temperature 170℃, transparency 94%; No. 4: Preheating temperature 125℃, pressure 1MPa, hot pressing time 2 minutes, hot pressing temperature 160℃, transparency 93%; No. 5: Preheating temperature 120℃, pressure 1.5MPa, hot pressing time 1.5 minutes, hot pressing temperature 150℃, transparency 91%. By observing the transparency data, a sudden increase in the transparency value of sample number 3 was discovered. This was identified as an abnormal fluctuation and added to the monitoring results. Monitoring allows for the identification of abnormal fluctuations, enabling timely detection of problems in the production process, such as improper process parameter settings, equipment malfunctions, and environmental changes. Based on this, necessary adjustments can be made to improve the hot-pressing quality and production efficiency of photovoltaic encapsulation films.

[0049] Furthermore, the model adjustment module 15 in the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters is also used for:

[0050] Correlation analysis is performed between the abnormal fluctuation data and the first hot-pressing process parameter set to generate multiple correlation coefficients; the response surface model is used to predict the first hot-pressing process parameter set to generate predicted response values; based on the multiple correlation coefficients, error analysis is performed between the predicted response values ​​and the abnormal fluctuation data to generate response error values; model diagnosis is performed on the response surface model based on the response error values ​​to identify and determine multiple problem areas; the response surface model is refitted according to the multiple problem areas, and the fitting results are evaluated to output the optimized response surface model.

[0051] Specifically, statistical methods were used to perform correlation analysis between abnormal fluctuation data and the first set of hot-pressing process parameters. Correlation analysis is a statistical method used to study the strength and direction of the relationship between two or more variables. Through correlation analysis, the correlation coefficient between each process parameter and the abnormal fluctuation data was calculated, reflecting the degree of association between the parameter and the abnormal fluctuation. The first set of hot-pressing process parameters was input into a response surface model for prediction, and the model calculated the corresponding predicted response values ​​based on the parameter values. Based on multiple correlation coefficients, the predicted response values ​​were compared with the actual abnormal fluctuation data. Error analysis is a statistical method used to assess the difference between the predicted values ​​and the actual values. Through error analysis, the error between the predicted response values ​​and the actual abnormal fluctuation data was calculated, reflecting the accuracy of the prediction.

[0052] By analyzing response error values, the model is diagnosed to identify regions where predictions are inaccurate, i.e., problem areas. These problem areas may be caused by inaccurate model parameters, incomplete data, or inapplicable model assumptions. The response surface model is then refitted according to these problem areas, adjusting model parameters. Refitting includes adjusting model parameters, adding or removing variables from the model, and modifying the model structure. The fitting process utilizes optimization algorithms and statistical methods to optimize model parameters with the goal of minimizing prediction error. After refitting, the fitting results are evaluated, including calculating model error, comparing predicted and actual values, and performing cross-validation. If the evaluation results show a significant improvement in model performance, the optimized response surface model is output. This optimized model is used to more accurately predict film quality under different combinations of process parameters. Through model diagnosis, problem areas causing abnormal fluctuations are accurately identified, and the model is refitted and optimized based on these problem areas to improve the accuracy of model predictions.

[0053] In summary, the intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters provided in this application has the following technical effects:

[0054] The system employs a process node analysis module to extract multiple process nodes based on the hot-pressing process of photovoltaic encapsulation films, and determines multiple process parameters based on these nodes. An impact information database construction module is used to obtain a database of process parameter impact information, derived from impact analysis of the multiple process nodes based on the aforementioned process parameters. A hot-pressing simulation module performs multiple hot-pressing simulations based on the database to obtain multiple film quality data for the photovoltaic encapsulation films. A regression analysis module uses the film quality data for regression analysis to construct a response surface model, which is then optimized to obtain a first set of hot-pressing process parameters. A model adjustment module monitors the execution of the first set of hot-pressing process parameters and adjusts the response surface model based on the monitoring results to generate an optimized response surface model. An intelligent optimization control module performs intelligent optimization control based on the optimized response surface model, outputting a second set of hot-pressing process parameters for the photovoltaic encapsulation films. In other words, by establishing a process parameter impact information database and a response surface model, intelligent optimization control of process parameters is achieved, improving the product quality and production efficiency of photovoltaic encapsulation films.

[0055] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily 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 this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0056] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. An intelligent optimization and control system for photovoltaic encapsulation film hot pressing process parameters, characterized in that, include: The process node analysis module is used to extract multiple process nodes based on the hot pressing process of photovoltaic encapsulation film, and determine multiple process parameters based on the multiple process nodes. The impact information database construction module is used to obtain a process parameter impact information database, which is obtained by performing impact analysis on multiple process flow nodes based on multiple process parameters. The hot pressing molding simulation module is used to perform multiple hot pressing molding simulations based on the process parameter influence information database to obtain multiple film quality data of photovoltaic encapsulation film; The regression analysis module is used to perform regression analysis using the multiple film quality data, construct a response surface model, and perform optimization through the response surface model to obtain the first hot pressing process parameter set. The model adjustment module is used to monitor the first hot pressing process parameter group, adjust the response surface model according to the monitoring results, and generate an optimized response surface model. The intelligent optimization control module is used to perform intelligent optimization control based on the response surface optimization model, and outputs a set of second hot-pressing process parameters for the photovoltaic encapsulation film, including: Intelligent optimization control is achieved using a response surface optimization model to predict the film quality under different combinations of process parameters. The process parameters are adjusted based on the prediction results. Through multiple iterations and adjustments, the combination of process parameters that maximizes product performance or minimizes cost is found. The second set of hot pressing process parameters for photovoltaic encapsulation film is output, including combinations of temperature, pressure, and time process parameters.

2. The intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters as described in claim 1, characterized in that, Regression analysis was performed using the multiple film quality data to construct a response surface model, including: The response surface model is used to preprocess the multiple film quality data to generate multiple film quality standard data. A regression function is introduced to perform regression analysis on the multiple film quality standard data to obtain multiple regression fitting coefficients; Parameter response correlation analysis was performed using the multiple regression fitting coefficients, and a response surface plot was drawn. The response surface model is validated using the response surface graph, and the response surface model is output based on the validation token.

3. The intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters as described in claim 2, characterized in that, By introducing a regression function, regression analysis is performed on the multiple film quality standard data to obtain multiple regression fit coefficients, including: Based on the multiple film quality standard data, a random selection is made, and the regression function is constructed according to the data selection results: ； in, For multiple regression fitting coefficients corresponding to multiple film quality standard data, Let i be the i-th independent variable in multiple film quality standard data. Let j be the j-th independent variable in multiple film quality standard data. This is the intercept term, representing the expected value of the response variable when all independent variables are 0. and The corresponding linear coefficients represent right The linear effect, It is a quadratic term. Is with The corresponding coefficients represent right The impact, Let i be an interactive term, where i ≠ j. Is with The corresponding coefficients represent and The interaction between them The impact, The number of independent variables, i.e., the total number of multiple film quality standard data; The multiple film quality standard data are synchronized to the regression function for regression analysis, and the multiple regression fit coefficients are output.

4. The intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters as described in claim 2, characterized in that, Using the multiple regression fitting coefficients, parameter response correlation analysis is performed, and response surface plots are drawn, including: Based on the multiple regression fitting coefficients and the multiple process parameters, response variable analysis is performed to determine multiple response vectors; Multiple response matrices are constructed by performing association mapping based on the multiple response vectors; Based on the multiple response matrices, determine whether the multiple columns of the multiple response matrices are linearly independent; If the columns of the plurality of response matrices are linearly independent, then the response surface graph is constructed based on the plurality of response matrices.

5. The intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters as described in claim 2, characterized in that, The response surface model is validated using the response surface graph, and the response surface model is output based on the validation token, including: The response surface plot is used in conjunction with the multiple film quality standard data to predict the multiple process parameters, thereby obtaining the predicted values ​​of the process parameter response variables; The predicted values ​​of the process parameter response variables are used as indexes to traverse the response surface plot for retrieval, and a response change trend is constructed based on the retrieval results. Based on the response change trend, extreme points are identified to determine the peak values ​​and valley values ​​of multiple parameters. The response surface model is cross-validated based on the peak values ​​and valley values ​​of the multiple parameters. When the validation passes, the validation token is generated and the response surface model is output.

6. The intelligent optimization control system for photovoltaic encapsulation film hot pressing process parameters as described in claim 5, characterized in that, The first set of hot pressing process parameters is obtained by optimization using the response surface model, including: The optimization objective is set by combining the peak values ​​and valley values ​​of the multiple parameters using the response surface model. The combined gradient of the multiple process parameters is calculated using the gradient descent method combined with the response surface model, and multiple partial derivatives are determined. These multiple partial derivatives have a corresponding relationship with the multiple process parameters. The gradient direction is determined based on the plurality of partial derivatives and the optimization objective, and the plurality of partial derivatives are compared with a preset gradient threshold sequentially according to the gradient direction. Process parameters whose partial derivatives are less than the preset gradient threshold are extracted and added to the candidate parameter group. Process parameters whose partial derivatives are greater than or equal to the preset gradient threshold are extracted and added to the elimination parameter group. This process is iterated until the traversal of the multiple process parameters is completed. The elimination parameter group is deleted, and the candidate parameter group is combined according to the hot pressing molding process flow to output the first hot pressing molding process parameter group.

7. The intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters as described in claim 1, characterized in that, The monitoring of the first hot pressing process parameter group includes: Multiple monitoring points are set up based on the hot pressing process flow, and data monitoring is performed on the first hot pressing process parameter group according to the multiple monitoring points to obtain the first response variable value. The first response variable value has a corresponding relationship with the first hot pressing process parameter group. Based on the value of the first response variable, trend analysis is performed to identify abnormal fluctuation data; Add the abnormal fluctuation data to the monitoring results.

8. The intelligent optimization control system for photovoltaic encapsulation film hot pressing molding process parameters as described in claim 7, characterized in that, The response surface model is adjusted based on the monitoring results to generate an optimized response surface model, including: Based on the correlation analysis between the abnormal fluctuation data and the first hot pressing process parameter group, multiple correlation coefficients are generated. The response surface model is used to predict the first hot pressing process parameter set to generate predicted response values. Based on the multiple correlation coefficients, error analysis is performed between the predicted response value and the abnormal fluctuation data to generate a response error value; Based on the response error value, the response surface model is diagnosed to identify and determine multiple problem areas; The response surface model is refitted according to the multiple problem regions, the fitting results are evaluated, and the optimized response surface model is output.

Images