Process parameter analysis method, system and storage medium for foam material molding process

Abstract

The invention relates to the technical field of process parameter analysis, and discloses a process parameter analysis method, system and storage medium for a foam material molding process. The method comprises: performing parameter collection and data statistics on an MPPO material under a standard process condition of 160°C to obtain process parameter benchmark data; performing an orthogonal test on a pressure-carrying parameter of the material to obtain a pressure-carrying process parameter database; performing a molding temperature gradient analysis, sampling and testing from 160°C to 190°C at intervals of 5°C to obtain a temperature control curve; performing coupling analysis and pressure dynamic compensation to obtain a molding process parameter combination; performing wet weight, dry weight and size detection on a product produced under the molding process parameter combination, and performing multidimensional parameter calculation to obtain a quality control index; performing correlation verification and parameter correction on the quality control index and the molding process parameter combination, and outputting target process specification data. The invention improves the control accuracy of the process parameters.

Description

Technical Field

[0001] The present invention relates to the field of process parameter analysis, and in particular to a process parameter analysis method, system and storage medium for a foam material molding process. Background Art

[0002] MPPO foam materials are widely used in automotive parts, building insulation and other fields due to their excellent mechanical properties, thermal insulation and lightweight characteristics. However, when traditional MPPO materials are molded at 160°C, there are problems such as long molding cycle, easy mold sticking, and low production efficiency, which seriously restricts product quality and production efficiency.

[0003] At present, the MPPO foam material molding process commonly used in the industry mainly relies on empirical parameter settings and lacks a systematic process parameter optimization method. In the actual production process, due to the fluctuation of process parameters such as temperature and pressure, the product quality is unstable, the yield rate is low, and the production cost is high. In addition, the existing process parameter adjustment method often uses a single parameter optimization and fails to consider the coupling relationship between multiple parameters such as temperature and pressure. Summary of the invention

[0004] The invention provides a process parameter analysis method, system and storage medium for a foam material molding process, and the invention improves the control accuracy of the process parameters.

[0005] In a first aspect, the present invention provides a method for analyzing process parameters of a foam material molding process, the method comprising:

[0006] Under the standard process condition of 160℃, the parameters of MPPO material were collected and data was collected to obtain the process parameter benchmark data.

[0007] Based on the process parameter benchmark data, an orthogonal test of the material pressure parameters is performed to obtain a pressure process parameter database;

[0008] Based on the pressure-carrying process parameter database, a molding temperature gradient analysis is performed, and sampling and testing are performed from 160° C. to 190° C. at intervals of 5° C. to obtain a temperature control curve;

[0009] Perform coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination;

[0010] Conducting wet weight, dry weight and size testing on the products produced under the molding process parameter combination, and performing multi-dimensional parameter calculation to obtain quality control indicators;

[0011] The quality control index and the molding process parameter combination are associated with verification and parameter correction, and target process specification data is output.

[0012] In a second aspect, the present invention provides a process parameter analysis system for a foam material molding process, the process parameter analysis system for a foam material molding process comprising:

[0013] The acquisition module is used to collect parameters and perform data statistics on the MPPO material under the standard process condition of 160°C to obtain the process parameter benchmark data;

[0014] A testing module, used to perform an orthogonal test of the pressure-loading parameters on the material based on the process parameter benchmark data to obtain a pressure-loading process parameter database;

[0015] An analysis module is used to perform a molding temperature gradient analysis based on the pressure-carrying process parameter database, sampling and testing from 160° C. to 190° C. at intervals of 5° C. to obtain a temperature control curve;

[0016] A compensation module, used for performing coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination;

[0017] A calculation module, used to detect the wet weight, dry weight and size of the product produced under the molding process parameter combination, and perform multi-dimensional parameter calculation to obtain quality control indicators;

[0018] The output module is used to perform correlation verification and parameter correction on the combination of the quality control index and the molding process parameter, and output target process specification data.

[0019] The third aspect of the present invention provides a computer device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned process parameter analysis method for the foaming material molding process.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned process parameter analysis method for the foaming material molding process.

[0021] In the technical solution provided by the present invention, the molding temperature is increased from 160°C to the range of 180-190°C through a systematic parameter analysis method, and combined with the precise control of the 7-8KG load pressure, the molding cycle is significantly shortened and the process parameter system is optimized. Based on the orthogonal test of the load pressure process parameters and the temperature gradient analysis, a complete process parameter database is constructed, which realizes the rapid optimization and adjustment of the production process and improves the production cycle. Through multi-dimensional parameter calculation and quality control indicator system, a complete product quality assessment method is established, which effectively reduces the incidence of defects such as product sticking and deformation. By using algorithms such as Gaussian linear regression and bidirectional long short-term memory network, intelligent coupling analysis and dynamic compensation of temperature-pressure parameters are realized, and the control accuracy of process parameters is improved. Through the optimization of process parameters and the establishment of quality control indicators, equipment investment and energy consumption are reduced, while the product yield is improved and production costs are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0023] Figure 1 Schematic diagram of the steps of a method for analyzing process parameters of a foaming material molding process in an embodiment of the present invention;

[0024] Figure 2 It is a structural schematic diagram of a process parameter analysis system for a foam material molding process according to an embodiment of the present invention;

[0025] Figure 3 It is a schematic block diagram of the structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] Embodiments of the present invention provide a method, system and storage medium for analyzing process parameters of a foaming material molding process. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of a method for analyzing process parameters of a foaming material molding process in an embodiment of the present invention includes:

[0028] Step S1, collecting parameters and performing data statistics on the MPPO material under standard process conditions of 160° C. to obtain process parameter benchmark data;

[0029] It is understandable that the execution subject of the present invention may be a process parameter analysis system for the foam material molding process, or may be a terminal or a server, which is not specifically limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0030] Specifically, under the standard process conditions of 160°C, continuous sample production tests were conducted on the MPPO material, and the original process data, including process parameters such as feed rate, silo temperature, mold temperature, and molding time, were collected. The mold sticking characteristics of the sample were analyzed based on the original process data, including the calculation of the mold sticking area, mold sticking degree, and mold sticking position. By acquiring the mold sticking characteristic data, the mold sticking phenomenon that occurred during the molding process was quantified and described. The mold sticking phenomenon is a common problem in the molding of foaming materials. If the mold sticking degree is high, it will directly affect the quality and appearance of the product. By calculating the mold sticking area, mold sticking degree, and mold sticking position, the molding process is effectively optimized and improved to reduce the adverse effects of mold sticking on product quality. According to the original process data, the sample was measured in full size and scanned for surface quality to obtain the product physical property data. The full-size measurement can reflect the dimensional stability and consistency of the product during the molding process, while the surface quality scanning test reveals whether there are defects on the product surface, such as bubbles, cracks, or uneven foaming areas. Through measurement and testing, the obtained product physical property data can fully reflect the quality status of the sample during the molding process. The sticking mold feature data and the physical property data of the product are analyzed in association to obtain the quality quantitative data. Through the association analysis, the sticking mold feature is matched with the physical properties of the final product to find the potential correlation between the two. For example, if the sticking mold phenomenon is more serious under certain specific process parameter combinations and the size or surface quality of the product is affected, it is inferred that these process parameters have an adverse effect on the molding quality. The quality quantitative data is systematically processed, and normal distribution test and outlier screening are performed on them to obtain parameter distribution data. The purpose of the normal distribution test is to determine whether the data conforms to the normal distribution, thereby verifying the stability of the process. If the data presents normal distribution characteristics, it means that the process has high repeatability and consistency, while the outlier screening eliminates data points that deviate from the normal range due to test errors or other accidental factors. Based on the parameter distribution data, the least squares method is used for fitting to obtain parameter fitting data. The least squares method is a data fitting method that obtains an optimal fitting curve by minimizing the error between the data point and the fitting curve, and finds the change trend and interaction law between different parameters. The parameter fitting data is analyzed in time series to obtain the benchmark data of the process parameters. Time series analysis is a data analysis method based on time sequence. By analyzing the changing patterns of data over time, the fluctuation trends and periodic characteristics of the process can be found, thereby judging the stability of the process and the problems that may arise.

[0031] Step S2, performing an orthogonal test on the pressure-loading parameters of the material based on the process parameter benchmark data to obtain a pressure-loading process parameter database;

[0032] Specifically, the process parameter benchmark data is input into the pressure control system for parameter configuration. By inputting the benchmark data, reasonable initial conditions are set for the pressure system. The pressure control system is composed of a piston structure driven by a servo motor. The piston driven by the servo motor applies pressure to the feeding system to obtain the initialization data of the pressure device. The initialization data of the pressure device includes parameters such as the starting value of the pressure, the driving speed of the servo motor, and the accuracy of the pressure application. During the pressure parameter test, the pressure time is divided to conduct a more detailed study on the influence of the time parameter. The pressure time is divided into five levels of 10 seconds, 20 seconds, 30 seconds, 40 seconds, and 50 seconds, so as to collect data under different pressure durations. The pressure data of different time periods are collected by the pressure sensor to obtain the time classification parameters. The time classification parameters reflect the flow and compression behavior of the material under different pressure time conditions, as well as the changes in the internal stress of the material. Different pressure times will have a significant impact on the foaming effect. By carefully dividing the time periods, the role of the time factor in the pressure process can be better understood. The pressure of the pressure is refined. The load pressure is divided into three levels according to 7KG, 7.5KG, and 8KG to evaluate the behavior characteristics of the material under different pressure conditions. The material change data in different pressure sections are collected by sensors to obtain the pressure classification parameters. The pressure classification parameters are directly related to the compaction degree of the material and the microstructural changes of the material during the molding process. For example, a higher load pressure helps to compact the material more evenly, but too high pressure will cause the destruction of the material structure. The time classification parameters and the pressure classification parameters are orthogonally combined to obtain experimental schemes under different parameter combinations. The orthogonal combination design process generates multiple groups of experimental schemes by systematically cross-pairing the time and pressure parameters to obtain the load pressure test matrix. The load pressure test matrix is ​​a systematic method of designing experiments, which effectively reduces the number of experiments and can fully cover all possible parameter combinations. The material load pressure test is performed on each set of parameters in the load pressure test matrix. Under each set of parameters, the pressure curve and temperature curve of the entire load pressure process are collected by sensors to obtain process monitoring data. The pressure curve reflects how the pressure applied under different parameters affects the flow characteristics and compaction degree of the material, while the temperature curve reveals the temperature changes inside the material during the pressure loading process. Temperature has an important influence on the behavior of foaming materials, because temperature changes will directly affect the fluidity of the material and the degree of foaming reaction. Perform a multi-dimensional analysis of the process monitoring data to calculate the flow state and temperature distribution characteristics of the material. Through a comprehensive analysis of the pressure curve and the temperature curve, the flow behavior of the material under different parameter combinations is obtained, such as whether the material can evenly fill the mold during the pressure loading process, whether there is stagnation or local flow problems, etc. The calculation of the temperature distribution characteristics reveals the temperature uniformity inside the material under different pressure conditions and whether there are local overheating or overcooling areas.Based on the above steps, a parameter database of the pressure-loading process is obtained, which summarizes detailed data such as material flow state, temperature distribution characteristics, pressure curve and temperature curve under different pressure-loading time and pressure-loading conditions.

[0033] Step S3, performing molding temperature gradient analysis based on the pressure-carrying process parameter database, sampling and testing from 160° C. to 190° C. at intervals of 5° C., and obtaining a temperature control curve;

[0034] Specifically, the process data in the pressure-carrying process parameter database is set in temperature segments, and 7 test temperature points are set at intervals of 5°C from 160°C to 190°C to form a complete temperature test sequence. Based on the temperature test sequence, the pressure-carrying parameters are kept unchanged at each set temperature point, and the samples are continuously produced. During the production process, the sample molding process at each temperature point is monitored in real time by an infrared thermometer, and temperature data is collected to obtain temperature sampling data. The infrared thermometer has high-precision measurement capabilities and accurately reflects the distribution of materials at different temperatures without interfering with the molding process. The temperature sampling data can effectively reveal the uniformity of the internal heat distribution of the material during temperature changes. The molding cycle time in the temperature sampling data is recorded and counted to obtain the molding cycle data, which reflects the time required to complete a molding process under different temperature conditions. At the same time, the molding cycle data is correlated with the energy consumption data of the equipment to obtain energy consumption analysis data. Through correlation calculation, the energy consumption characteristics of the equipment under different temperature conditions are clarified, and the specific relationship between molding time and energy consumption is found. The energy consumption analysis data is divided into temperature intervals to obtain temperature efficiency data, which reflects the energy utilization of material molding in different temperature intervals. Through these data, it is found out which temperature interval can achieve the highest energy efficiency. In order to evaluate the impact of different temperatures on product quality, the temperature efficiency data is mapped and analyzed with the product quality test results to obtain quality-related data. The product quality test results include indicators such as material density, surface quality, and foaming uniformity. By mapping with the temperature efficiency data, the relationship between energy efficiency and product quality under different temperature conditions can be seen. A temperature gradient model is established based on quality-related data, and the optimal temperature control strategy is determined by systematic analysis of material properties and production efficiency at different temperature points. The temperature gradient model determines the temperature optimization data through numerical calculation, that is, which temperature points can achieve the best energy efficiency and molding cycle under different process requirements while ensuring product quality. By curve fitting and smoothing the temperature optimization data, the discrete temperature data points are connected in a smooth manner through mathematical methods to form a control curve that can be actually applied to the production process. The smoothing process is to eliminate the fluctuations and noise in the experimental data, so that the temperature control curve is more stable and operable in practical applications, and the final temperature control curve is obtained.

[0035] Step S4, performing coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination;

[0036] Specifically, the temperature control curve and the pressure process parameter database are time-series aligned and data preprocessed to form a standardized parameter matrix. The purpose of time-series alignment is to ensure that the temperature curve and pressure parameters are consistent in the time dimension so as to perform effective data fusion analysis. Data preprocessing includes normalizing the temperature and pressure data to eliminate dimensional differences, ensure the consistency of model input and the effective use of data, and obtain a standardized parameter matrix. The standardized parameter matrix is ​​input into the Gaussian linear regression model, and the coupling coefficient between temperature and pressure is calculated by the maximum likelihood estimation method to obtain parameter coupling data. The Gaussian linear regression model is a probabilistic model that determines the optimal parameters of the model by maximizing the likelihood probability of the data, and the coupling coefficient reflects the interaction between temperature and pressure in the foaming material molding process. The parameter coupling data is input into the bidirectional long short-term memory network (Bi-LSTM), and the pressure compensation sequence is calculated by forward propagation and backward propagation. Bi-LSTM is a deep learning model that can process time series data and has the ability to remember past and future information. Through forward propagation, the model learns the previous state in the time series, and through backward propagation, it predicts the future trend to obtain a more accurate pressure compensation sequence. According to the trend of temperature change, the pressure system is dynamically compensated for pressure, so that the pressure in the molding process can match the temperature change in real time, ensuring the stability and consistency of molding quality. Based on the pressure compensation sequence, the population initialization scheme of the non-dominated sorting genetic algorithm (NSGA-II) is constructed, and the initial solution of pressure compensation is obtained through encoding and decoding. As a multi-objective optimization algorithm, NSGA-II searches for the optimal solution by constructing an initial population. Population initialization is a key step in the algorithm solution, and the quality of the solution is effectively improved through appropriate population settings. The encoding and decoding process converts the continuous pressure compensation data into a form that can be processed by the genetic algorithm so as to operate in the optimization process. The initial solution of pressure compensation is evaluated for fitness and non-dominated sorting to evaluate the advantages and disadvantages of each solution. The purpose of fitness evaluation is to determine the effectiveness of each pressure compensation scheme, that is, whether a specific pressure compensation can effectively improve the quality of the foaming material under given temperature conditions. Through non-dominated sorting, all solutions are hierarchically classified to determine which solutions perform better among multiple objectives (such as minimizing energy consumption and maximizing quality). Through crossover and mutation operations, a multi-objective optimization solution set is obtained. Crossover and mutation are the core operations in genetic algorithms. New solutions are generated through crossover, while mutation increases the diversity of the solution set and improves the algorithm's ability to escape from local optimality. The multi-objective optimization solution set is screened through crowding calculation and elite retention strategy to obtain the final Pareto optimal solution. Crowding calculation is used to evaluate the density of each solution in the solution set, so as to avoid excessive concentration in a specific area and loss of diversity.The elite retention strategy ensures that the best solution can be retained in each generation, thereby continuously promoting the evolution of the solution and obtaining better results. Through this step, a set of Pareto optimal solutions are obtained, which represent the optimal solution to balance multiple objectives such as temperature, pressure, and energy consumption. The decision variables are extracted and the parameters are mapped for the Pareto optimal solution, and the parameter combination of the molding process is generated through reverse decoding. The extraction of decision variables aims to select the most suitable process parameters from the Pareto optimal solution, while the parameter mapping is to map these variables back to the actual physical process parameters. Reverse decoding converts the encoding results of the genetic algorithm into the form of actual process parameters for use in the actual molding process.

[0037] Step S5, testing the wet weight, dry weight and size of the product produced under the molding process parameter combination, and performing multi-dimensional parameter calculation to obtain quality control indicators;

[0038] Specifically, the molding process parameter combination is applied to the actual production process to generate batch samples. On the production line, the wet weight data of the molded product is collected by an electronic balance to obtain the wet weight parameters. The wet weight parameters reflect the overall quality of the product after molding, including the water content. The wet weight product is subjected to standardized drying treatment to remove the moisture inside it. After the drying treatment is completed, the dry weight parameters of the product are measured again using an electronic balance to obtain the net weight of the product after removing the moisture. The dry weight parameter is one of the key indicators for judging the degree of foaming of the material and the uniformity of the internal structure, because different molding conditions will lead to differences in moisture residue and foaming density. By obtaining wet weight and dry weight data, the water absorption and foaming characteristics of the material are evaluated. Based on the collected wet weight parameters and dry weight parameters, differential calculations are performed to obtain the material density distribution data of the product. The calculation of material density can reveal the uniformity of different regions of the product during the foaming process. The uniformity of density has an important influence on the physical properties of the foaming material, which is directly related to the performance indicators such as strength and elasticity of the product. In order to evaluate the geometric characteristics of the product, the product is scanned by a three-dimensional measuring machine, and the geometric dimension data such as length, width, thickness, etc. of the product are obtained through the data acquisition system. These data can reflect the dimensional stability and consistency of the product during the molding process. The material density distribution data and geometric dimension data are standardized to eliminate the influence of different measurement units and dimensions on data analysis, and obtain standardized characteristic data. Based on the standardized characteristic data, a multi-dimensional evaluation is performed to obtain a quality score matrix. Each element in the quality score matrix represents the quality score in a specific dimension. The evaluation dimensions include material uniformity, dimensional accuracy, foaming quality and other aspects. Principal component analysis is performed based on the quality score matrix to obtain the target influencing factors. The multi-dimensional original data is reduced to a few comprehensive indicators, namely the target influencing factors, through linear transformation. The target influencing factors represent the key factors that have the greatest impact on product quality. By performing principal component analysis on the original multi-dimensional quality data, the most representative quality evaluation criteria can be effectively found to simplify the subsequent analysis process. Based on the target influencing factors, multi-dimensional weight calculations are performed to determine the relative importance of each target influencing factor in the overall quality evaluation. By assigning appropriate weights to each factor, a more accurate quality evaluation model is constructed to obtain the final quality control index.

[0039] Step S6: perform correlation verification and parameter correction on the combination of quality control indicators and molding process parameters, and output target process specification data.

[0040] Specifically, the quality control index is combined with the molding process parameters for data matching, the specific relationship between the quality index and each process parameter is found, and the process parameter matching data is obtained. Quality prediction is performed based on the matching data to obtain the parameter correlation matrix. The parameter correlation matrix reflects the correlation between each process parameter and between them and the product quality. Parameter sensitivity analysis is performed based on the parameter correlation matrix to evaluate the degree of influence of different parameters on the quality of the final product. The weight data of the parameters is obtained through sensitivity analysis, which reflects the importance and influence of each process parameter in the overall process. The parameter weight data is imported into the process control system to generate process verification data. In order to ensure the stability of the process, the process verification data is statistically analyzed to obtain process stability data. The process stability data includes the fluctuation range of each parameter in the production process, the change trend of the parameters in different time periods, and the impact on the quality of the final product. By analyzing the process stability data, it is judged whether the current process can be maintained in a stable state for a long time. If some parameters fluctuate greatly, it will lead to unstable molding quality, and these parameters need to be corrected. Based on the process stability data, the key parameters are compensated as necessary to obtain parameter correction data. By adjusting the set values ​​of specific parameters, the negative impact on the process is reduced to ensure the stability and consistency of production. Apply parameter correction data to the continuous production process to verify whether the corrected process parameter combination can meet production requirements. Record various parameters and quality data in the production process through a real-time monitoring system to obtain process control data. Accurately control every link in the production process to ensure that each process step is executed according to the set parameters, and make timely corrections to any deviation from the target parameters. Through real-time production data collection, effectively monitor every detail of the process to ensure production stability. Perform multiple rounds of production verification and parameter optimization on process control data to ensure that the process parameter combination can also maintain consistency and stability in long-term production. Verify the effectiveness of parameter correction through repeated production experiments and find out the further optimization space. Each round of production verification is adjusted based on the results of the previous round, gradually approaching the optimal production conditions, and finally obtaining optimized target process specification data.

[0041] In the embodiment of the present invention, the molding temperature is increased from 160°C to the range of 180-190°C through a systematic parameter analysis method, and combined with the precise control of the 7-8KG load pressure, the molding cycle is significantly shortened and the process parameter system is optimized. Based on the orthogonal test of the load pressure process parameters and the temperature gradient analysis, a complete process parameter database is constructed, which realizes the rapid optimization and adjustment of the production process and improves the production cycle. Through multi-dimensional parameter calculation and quality control indicator system, a complete product quality assessment method is established, which effectively reduces the incidence of defects such as product sticking and deformation. By using algorithms such as Gaussian linear regression and bidirectional long short-term memory network, intelligent coupling analysis and dynamic compensation of temperature-pressure parameters are realized, and the control accuracy of process parameters is improved. Through the optimization of process parameters and the establishment of quality control indicators, equipment investment and energy consumption are reduced, while the product yield is improved and production costs are reduced.

[0042] In a specific embodiment, the process of executing step S1 may specifically include the following steps:

[0043] Conduct continuous production test on MPPO material samples under standard process conditions of 160°C, and collect original process data corresponding to feed speed, silo temperature, mold temperature and molding time;

[0044] Based on the original process data, the mold sticking area, degree of mold sticking and mold sticking position of the sample are calculated to obtain the mold sticking characteristic data. Based on the original process data, the sample is fully measured and the surface quality is scanned and tested to obtain the product physical property data.

[0045] Perform correlation analysis on the sticking mold feature data and the product physical property data to obtain quality quantitative data, and perform normal distribution test and outlier screening on the quality quantitative data to obtain parameter distribution data;

[0046] The least squares method is used to fit the parameter distribution data to obtain the parameter fitting data, and the parameter fitting data is analyzed by time series to obtain the process parameter benchmark data.

[0047] Specifically, the MPPO (polyphenylene ether) material samples were tested for continuous production under the standard process conditions of 160°C, and a number of process raw data including feed rate, silo temperature, mold temperature and molding time were collected. The feed rate (v) represents the rate at which the material enters the molding area, which directly affects the material filling and foaming uniformity; the silo temperature (T_silo) reflects the temperature state of the material during the preheating and plasticizing stages, and the temperature at this stage plays an important role in the material's fluidity and the final foaming effect; the mold temperature (T_mold) is the external temperature condition to which the material is subjected during the molding process, which has a direct impact on the material cooling rate and molding quality; the molding time (t) refers to the time required from the material entering the mold to the final cooling and shaping. Too long or too short a molding time will affect the internal structure and appearance characteristics of the material. After collecting these process raw data, the quality of the molded samples was evaluated. In order to analyze the mold sticking of the material during the molding process, the mold sticking area (A sticking), mold sticking degree (D sticking) and mold sticking position (P sticking) of the sample were calculated. The sticky mold area Asticky is calculated by the following formula:

[0048] ;

[0049] in, Represents the adhesion function of the sample surface in the contact area with the mold, It is a small area element. By integrating the entire contact surface of the sample, the total mold sticking area is obtained. It is expressed as the average value of the sticking force:

[0050] ;

[0051] in, It is the adhesion force per unit area of ​​the mold sticking area. It is described by locating the coordinates of the sticking area on the mold, using These data are used to determine which parts of the mold are more likely to stick, and provide a reference for improving mold design. After completing the calculation of the sticking characteristic data, the sample is measured in full size and the surface quality is scanned and tested to obtain the physical property data of the product. The geometric dimensions of the sample after molding are obtained, including length (L), width (W) and thickness (H). These data are used to determine the dimensional stability of the sample during the molding process. These dimensional values ​​are accurately obtained by a three-coordinate measuring machine:

[0052] ;

[0053] The function Indicates that the geometric dimensions are a function of the process parameters. Surface quality scanning is to detect whether there are defects on the sample surface, such as bubbles, cracks or irregular foaming areas. These data are important indicators of product appearance quality. The sticking feature data and the physical property data of the product are correlated and analyzed to reflect the relationship between the sticking phenomenon and product quality. Find the specific relationship between the influence of different process parameters on the sticking characteristics and the physical properties of the product to obtain quality quantitative data. Correlation analysis is achieved by building a regression model, for example:

[0054] ;

[0055] in, Indicates the quality rating of the product. are the regression coefficients of mold sticking area, mold sticking degree and mold sticking position on quality score, is a random error term. The degree of influence of different sticking mold characteristics on product quality is determined by regression analysis. The quality quantitative data is tested for normal distribution and outliers are screened to obtain parameter distribution data. The normal distribution test is performed by the Shapiro-Wilk test or the Kolmogorov-Smirnov test to verify whether the data conforms to the normal distribution. If the data shows obvious deviations, data transformation or outliers are eliminated. Outliers are screened by the 3σ principle, that is, judging whether the data falls within the mean. Three times the standard deviation If it exceeds the range, it is considered an outlier and needs to be eliminated or further analyzed. Based on the screened parameter distribution data, the least squares method is used for fitting to obtain parameter fitting data. The optimal solution of the model parameters is determined by minimizing the sum of squares of the errors between the actual observations and the model predictions. For a polynomial fitting model, it is expressed as:

[0056] ;

[0057] in, represents the parameter data after fitting, represents a process parameter (such as temperature or pressure), are fitting coefficients. These coefficients are determined by the least squares method so that the model can describe the changing trend of the data as accurately as possible. Time series analysis is performed on the fitting data to obtain the benchmark data of the process parameters. The changing trend and periodic characteristics of the data over time are analyzed to determine whether the process is stable. In time series analysis, the autoregressive model (AR model) is used to describe the time dependence of the data:

[0058] ;

[0059] in, Indicates time Parameter data at the moment, is a constant term, is the autoregression coefficient, is a random error term. The autoregressive model is used to determine the fluctuation of process parameters in different time periods, thereby determining the stability and trend of the parameters. The benchmark data of process parameters are obtained through time series analysis.

[0060] In a specific embodiment, the process of executing step S2 may specifically include the following steps:

[0061] Input the process parameter benchmark data into the pressure control system for parameter configuration, apply the loading pressure to the feeding system through the piston structure driven by the servo motor, and obtain the initialization data of the pressure device;

[0062] The pressure-loading time is divided into five levels: 10 seconds, 20 seconds, 30 seconds, 40 seconds, and 50 seconds. The pressure-loading data of different time periods are collected by the pressure sensor to obtain the time classification parameters.

[0063] The load pressure is divided into three levels: 7KG, 7.5KG, and 8KG. The material change data of different pressure sections are collected through sensors to obtain the pressure classification parameters;

[0064] Orthogonal combination design is performed on time classification parameters and pressure classification parameters, and multiple groups of experimental schemes are generated through parameter pairing to obtain the load and pressure test matrix;

[0065] Perform material pressure test on each set of parameters in the pressure test matrix, collect pressure curve and temperature curve through sensors, and obtain process monitoring data;

[0066] Perform multi-dimensional analysis on process monitoring data, calculate material flow status and temperature distribution characteristics, and obtain a pressure-carrying process parameter database.

[0067] Specifically, the process parameter benchmark data is input into the pressure control system for parameter configuration. The benchmark data includes key process parameters such as temperature, time, and pressure involved in the molding process. The piston structure driven by the servo motor applies the loading pressure to the feeding system to form the initial compression state of the material and obtain the initialization data of the pressure device. The driving characteristics of the servo motor provide control over the pressure accuracy and dynamic response, which can ensure that the pressure is applied stably and accurately. Assume that the loading pressure is , then the force applied by the servo motor It is expressed by the following formula:

[0068] ;

[0069] in, is the cross-sectional area of ​​the piston, in units of , is the initial pressure of the load, in units of . By applying loading pressure to the feeding system, ensure that the material can reach the expected physical state before entering the mold. In the process of pressure parameter testing, the pressure time is carefully divided into five levels of 10 seconds, 20 seconds, 30 seconds, 40 seconds, and 50 seconds, so as to study the influence of different pressure duration on material deformation and foaming quality. The pressure sensor collects the pressure data at each time stage to obtain the time classification parameters. Assume that the pressure is at the time Time , the change of pressure with time is expressed by the following formula:

[0070] ;

[0071] in, is the initial pressure, is the rate coefficient of pressure change, in units of . By collecting pressure at different time periods, a series of pressure values ​​are obtained to form time classification parameters. While classifying time, the load pressure is also classified. The load pressure is divided into three levels of 7KG, 7.5KG, and 8KG to clarify the impact of different load pressures on material flow, foaming, and most importantly, molding quality. The sensor collects the change data of the material in each pressure section to obtain the pressure classification parameters. Load pressure As a control parameter, its change will affect the stress distribution inside the material. Assuming that the material is under pressure Volume change under the conditions It is expressed as:

[0072] ;

[0073] in, is the initial volume in , is the bulk modulus of the material, in units of By recording the material changes in different pressure sections, we can understand the effect of pressure on material deformation and flow. Orthogonal combination design is performed on the time classification parameters and pressure classification parameters to generate multiple groups of experimental schemes and obtain the load pressure test matrix. Orthogonal combination design is a statistical experimental design method that covers all possible parameter combinations as comprehensively as possible by reducing the number of experiments. Assume that the load pressure time and pressure are and , then the experimental scheme formed by orthogonal combination is expressed as:

[0074] ;

[0075] By combining, we get a set of A set of experimental matrices with different parameters ensures a comprehensive investigation of the process parameters. Material pressure test is performed on each set of parameters in the pressure test matrix. During the test, the pressure curve and temperature curve are collected by the pressure sensor and temperature sensor respectively to obtain the monitoring data of the whole process. The pressure curve reflects the compression behavior of the material under different conditions, while the temperature curve reveals the thermal distribution of the material during the whole pressure process. Assume that the change of pressure over time is expressed as , and the temperature variation with time is expressed as , then the process monitoring data is represented as a set of functions over time:

[0076] ;

[0077] Perform multi-dimensional analysis on process monitoring data to calculate the material flow state and temperature distribution characteristics. The material flow state is described by the flow index, assuming that the material flow and temperature and load pressure The liquidity index It is expressed as:

[0078] ;

[0079] in, is a constant related to material properties, with units of , is the fluidity coefficient, which is determined by experiment. The temperature distribution characteristics are described by the temperature gradient, assuming that the temperature gradient is , expressed as:

[0080] ;

[0081] in, and are the highest and lowest temperatures in the material, The length of the measurement path. Through the analysis of temperature and pressure, the fluidity of the material during the molding process and its temperature distribution characteristics are obtained. These characteristics help to determine the molding effect of the material under different pressure conditions. Through these multi-dimensional analyses, a parameter database of the pressure process is obtained, including data on material flow characteristics, temperature distribution, pressure curves, etc. under different time and pressure conditions.

[0082] In a specific embodiment, the process of executing step S3 may specifically include the following steps:

[0083] The process data in the pressure-carrying process parameter database is set in temperature segments, and 7 test temperature points are set from 160°C to 190°C with an interval of 5°C to obtain a temperature test sequence;

[0084] Based on the temperature test sequence, the load pressure parameters are kept unchanged at each temperature point to continuously produce samples, and the temperature data is collected by an infrared thermometer to obtain temperature sampling data;

[0085] The molding cycle time in the temperature sampling data is recorded and counted to obtain the molding cycle data, and the molding cycle data is correlated with the equipment energy consumption data to obtain the energy consumption analysis data;

[0086] Divide the energy consumption analysis data into temperature intervals to obtain temperature efficiency data, and map the temperature efficiency data with product quality test results to obtain quality-related data;

[0087] A temperature gradient model is established based on the mass correlation data, the temperature optimization data is determined through numerical calculation, and the temperature optimization data is subjected to curve fitting and smoothing to obtain the temperature control curve.

[0088] Specifically, the process data in the pressure-carrying process parameter database is set in temperature segments. Seven test temperature points are set from 160°C to 190°C at intervals of 5°C to obtain a temperature test sequence. The temperature test sequence is expressed as:

[0089] ;

[0090] in, Indicates temperature test points in degrees Celsius. By setting the temperature in increments, the effect of temperature changes on the molding properties of the foaming material is systematically analyzed. Each temperature point represents a specific molding process condition, ensuring complete coverage of material behavior in different temperature ranges. Based on the temperature test sequence, the load pressure parameters are kept unchanged at each temperature point, and the samples are continuously produced. The effect of temperature as a single variable on the foaming molding of the material is analyzed, thereby eliminating the interference of the load pressure parameters. In the process of sample production at each temperature point, an infrared thermometer is used to collect real-time temperature data of the sample during the molding process to obtain temperature sampling data. The infrared thermometer has high-precision measurement capabilities and can record the temperature changes on the surface of the material in real time without interfering with the molding process. Assuming that the infrared thermometer is at time The collection temperature is , then the temperature sampling data is represented as a function sequence:

[0091] ;

[0092] These data are used to analyze the thermal distribution of materials at different temperature points and their impact on molding quality. After obtaining the temperature sampling data, the molding cycle time of the sample is recorded and counted to obtain the molding cycle data. It refers to the time from when the material enters the mold to when it is completely cooled and demolded. The shorter the molding cycle, the higher the production efficiency, but it will affect the quality of the material. The calculation of the molding cycle is expressed by the following formula:

[0093] ;

[0094] in, It is an indicator function of whether the material has reached a certain molding condition. Its value is 1, which means the molding condition is met, and 0, which means it is not met. The molding cycle data is correlated with the energy consumption data of the equipment to obtain energy consumption analysis data. Energy consumption data Represents the total energy consumed by the equipment to complete a molding cycle under specific temperature and pressure conditions, and is calculated using the following formula:

[0095] ;

[0096] in, is the power of the device in kilowatts, is the molding cycle in hours. Through the relationship between the molding cycle and the energy consumption of the equipment, we can intuitively understand the impact of different temperature points on the energy consumption of the production process. The energy consumption analysis data is divided into temperature intervals to obtain temperature efficiency data. Temperature efficiency It is used to measure the relationship between energy consumption and production efficiency within a certain temperature range, expressed by the following formula:

[0097] ;

[0098] in, is the quality score of the product, expressed in units of mass units, is the energy consumption in kilowatt-hours. The calculation of temperature efficiency helps analyze which range can achieve the highest molding quality and production efficiency with the least energy input under different temperature conditions. Through analysis, the optimal temperature point is determined to achieve a balance between the molding quality of the material and the production energy efficiency. The temperature efficiency data is mapped and analyzed with the quality inspection results of the product to obtain quality-related data. The quality inspection of the product includes multiple indicators such as dimensional accuracy, density uniformity, surface quality, etc. Each indicator reflects the performance of the product during the molding process. Assume that the quality inspection result at a certain temperature point is , the corresponding temperature efficiency is , the quality-related data is obtained through the mapping function:

[0099] ;

[0100] Among them, the mapping function Represents the influence of different temperature efficiencies on product quality. Through mapping, we can understand the specific role of temperature efficiency on the quality of the final product and determine which temperature points achieve the best product performance. Based on the quality-related data, a temperature gradient model is established to describe the overall impact of temperature changes on product quality. The temperature gradient model is established through a fitting function, assuming that the temperature and quality score The relationship between is expressed as a polynomial model:

[0101] ;

[0102] in, The coefficients are fitted by the least square method so that the polynomial can best fit the mass correlation data obtained by measurement. The fitted data is smoothed. The noise in the data is eliminated by moving average or Gaussian filtering, making the fitting curve smoother and forming a temperature control curve.

[0103] In a specific embodiment, the process of executing step S4 may specifically include the following steps:

[0104] Perform time-series alignment and data preprocessing on the temperature control curve and the pressure process parameter database to obtain a standardized parameter matrix;

[0105] The standardized parameter matrix is ​​input into the Gaussian linear regression model, and the temperature-pressure coupling coefficient is calculated by maximum likelihood estimation to obtain parameter coupling data;

[0106] Input the parameter coupling data into the bidirectional long short-term memory network, and calculate the pressure compensation sequence through forward propagation and backward propagation;

[0107] A population initialization scheme of a non-dominated sorting genetic algorithm is constructed based on the pressure compensation sequence, and the initial solution of pressure compensation is obtained through encoding and decoding.

[0108] The initial pressure compensation solutions are evaluated for fitness and non-dominated sorting, and a multi-objective optimization solution set is obtained through crossover and mutation operations.

[0109] The multi-objective optimization solution set is screened through crowding calculation and elite retention strategy to obtain the Pareto optimal solution. The decision variables and parameter mapping of the Pareto optimal solution are then performed, and the molding process parameter combination is obtained through reverse decoding.

[0110] Specifically, the temperature control curve and pressure data are aligned according to the timestamps so that the temperature and pressure data at each time point correspond. Assume that the temperature data in the temperature control curve is represented as , and the pressure data in the pressure-loading process is , then after the timing is aligned, the following standardized parameter matrix is ​​obtained:

[0111] ;

[0112] in, For in time The standardized parameter matrix when Represents temperature data, Represents the corresponding pressure data. The data preprocessing steps include denoising, smoothing and standardization, so that the units of all data points are consistent and can better match the model input requirements. For standardization, it is assumed that the mean values ​​of temperature and pressure are and , and the standard deviations are and , the standardized temperature and pressure data are:

[0113] ;

[0114] After data preprocessing, each column of the standardized parameter matrix becomes data of uniform scale. The standardized parameter matrix is ​​input into the Gaussian linear regression model, and the temperature-pressure coupling coefficient is calculated by maximum likelihood estimation to obtain parameter coupling data. The Gaussian linear regression model predicts the target variable through training data, that is, given temperature data How to estimate stress through regression when The Gaussian linear regression model is expressed as:

[0115] ;

[0116] in, is the regression weight vector, is the bias term, is the standardized input matrix. The optimal regression coefficient is estimated by maximum likelihood estimation. and , so that the pressure output by the model is As close as possible to the actual observed value. The goal of maximum likelihood estimation is to maximize the following likelihood function:

[0117] ;

[0118] in, represents a normal distribution, is the noise variance, is the likelihood function, which indicates the degree of match between the model prediction and the actual observed data. By maximizing this likelihood function, the optimal temperature-pressure coupling coefficient is obtained. and , thus obtaining the coupling data between temperature and pressure. The parameter coupling data is input into the bidirectional long short-term memory (Bi-LSTM) network, and the pressure compensation sequence is calculated through forward propagation and backward propagation. The bidirectional LSTM network is a recurrent neural network that can capture the time dependency of the input sequence in two directions (forward and reverse) at the same time. After inputting the coupling data of temperature and pressure, the LSTM network can learn the complex patterns in the time series. The basic formula of the bidirectional LSTM is:

[0119] ;

[0120] ;

[0121] in, is the input data, and are the hidden states of the forward and reverse LSTMs, respectively, and are the cell states of the forward and reverse LSTM, respectively. Through the information in these two directions, the LSTM network can output a pressure compensation sequence. Assume that the pressure compensation sequence is , then the final pressure compensation sequence is:

[0122] ;

[0123] This sequence represents how to compensate for pressure under given temperature and pressure data. Based on the pressure compensation sequence, a population initialization scheme for the non-dominated sorting genetic algorithm (NSGA-II) is constructed, and the initial solution for pressure compensation is obtained through encoding and decoding. Genetic algorithm is an optimization method that simulates natural selection and finds the optimal solution to the problem through operations such as selection, crossover, and mutation. In this problem, the optimization goal is to determine the optimal process parameter configuration based on the pressure compensation sequence, thereby improving the quality of the molded product. Initialize a population in which each individual represents a candidate solution. The solution is encoded in binary encoding or real-valued encoding. Assume that the parameters of each candidate solution are , represents the corresponding process parameters. Through crossover and mutation operations, new solutions are generated and compared with the current solutions to select the individuals with the best fitness. Fitness evaluation is performed by calculating the objective function value of each solution. The objective function is a weighted sum based on multiple evaluation indicators (for example, product quality, energy consumption, production efficiency, etc.). In each generation of the genetic algorithm, individuals in the population are screened by non-dominated sorting and crowding calculation to obtain a multi-objective optimization solution set. In multi-objective optimization, non-dominated sorting divides the solution set into different levels, each of which contains a non-dominated solution, indicating that it is better than other solutions in a certain goal. Through operations such as crossover and mutation, new solutions are generated and fitness evaluation is performed. The final solution set is the Pareto optimal solution set, which means that the optimal solution is achieved in multiple goals. The final molding process parameter combination is obtained by extracting decision variables and mapping parameters of the Pareto optimal solution. Assume that the Pareto optimal solution is , then the final molding process parameter combination is:

[0124] ;

[0125] in, It is to convert the optimal solution code into actual process parameters.

[0126] In this embodiment, the standardized parameter matrix is ​​input into the Gaussian linear regression model, and the temperature-pressure coupling coefficient is calculated by maximum likelihood estimation to obtain parameter coupling data, including: performing eigendecomposition and singular value analysis on the standardized parameter matrix, and obtaining a eigenvector sequence by matrix transformation; grouping the eigenvector sequence according to temperature and pressure variables, and obtaining a variable correlation matrix by covariance calculation; constructing a Gaussian kernel function for the variable correlation matrix, and obtaining a kernel function matrix by radial basis function calculation; inputting the kernel function matrix into the linear regression layer for optimization calculation, and obtaining regression coefficients by gradient descent method; performing maximum likelihood estimation on the regression coefficients, and obtaining likelihood function values ​​by probability density function calculation; constructing a target optimization function based on the likelihood function value, and obtaining an optimal solution by Newton iteration method; substituting the optimal solution into the Gaussian linear regression model, and obtaining a temperature-pressure mapping relationship by forward calculation; performing residual analysis and error correction on the temperature-pressure mapping relationship, and obtaining parameter coupling data by reverse updating.

[0127] The parameter coupling data is input into a bidirectional long short-term memory network, and a pressure compensation sequence is calculated through forward propagation and backward propagation, including: performing time window segmentation and data normalization processing on the parameter coupling data to obtain a network input sequence; sending the network input sequence into the input layer, and obtaining an input layer feature map through weight matrix and bias vector calculation; constructing a forget gate based on the input layer feature map, and selectively forgetting the historical state information through the sigmoid activation function to obtain a forget gate output; combining the forget gate output with the new input information, and calculating the candidate memory unit state through the tanh activation function to obtain the candidate state; performing weighted fusion on the candidate state and the forget gate output, and controlling the degree of information update through the input gate to obtain the current memory unit state; calculating the output gate state based on the current memory unit state, and extracting information from the memory unit through the tanh function to obtain the hidden layer output; performing feature dimensionality reduction on the hidden layer output through the fully connected layer, and combining the residual connection to process the temporal dependency to obtain the output layer feature; performing back propagation error calculation and gradient update on the output layer feature to obtain the pressure compensation sequence.

[0128] Among them, the initial solution of pressure compensation is evaluated for fitness and non-dominated sorting, and a multi-objective optimization solution set is obtained through crossover mutation operation, including: constructing a fitness function for the initial solution of pressure compensation, calculating the fitness value of each solution through weighted combination of target parameters, and obtaining fitness score data; sorting and grading the fitness score data, stratifying the solution space through dominance relationship judgment, and obtaining non-dominated hierarchical data; calculating the congestion of each layer of solution set in the non-dominated hierarchical data, and obtaining the congestion distribution data through distribution density analysis of the solution in the target space; and performing individual selection based on the congestion distribution data. The parent individuals are determined through the tournament selection mechanism to obtain mating pair data; a simulated binary crossover operation is performed on the mating pair data, and the parameter exchange process is controlled by the crossover probability to obtain the offspring crossover data; a polynomial mutation operation is performed on the offspring crossover data, and the parameter perturbation intensity is adjusted by the mutation probability to obtain the mutation optimization data; new solutions are evaluated and screened based on the mutation optimization data, and the population is re-stratified through the fast non-dominated sorting algorithm to obtain the hierarchical sorting data; convergence judgment and elite retention are performed on the hierarchical sorting data, and the iterative process is controlled by genetic algebra to obtain a multi-objective optimization solution set.

[0129] In a specific embodiment, the process of executing step S5 may specifically include the following steps:

[0130] Apply the molding process parameter combination to product production, collect the wet weight data of the products off the production line through an electronic balance to obtain the wet weight parameters, and perform standard drying treatment on the products off the production line, and measure the dry weight parameters of the products through an electronic balance;

[0131] Based on the wet weight parameters and dry weight parameters, differential calculation is performed to obtain the material density distribution data of the product, and the product is scanned by a three-dimensional coordinate measuring machine. The length, width, and thickness values ​​of the product are obtained through the data acquisition system to obtain the geometric dimension data;

[0132] Standardize the material density distribution data and geometric dimension data to obtain standardized characteristic data, and perform multi-dimensional evaluation based on the standardized characteristic data to obtain a quality scoring matrix;

[0133] Principal component analysis is performed based on the quality score matrix to obtain the target impact factor, and multi-dimensional weight calculation is performed based on the target impact factor to obtain the quality control index.

[0134] Specifically, the optimized molding process parameter combination is used for production, and the wet weight data of the products off the production line are collected through an electronic balance. The wet weight data represents the weight of the product after absorbing moisture during the production process, which is a key parameter for judging the molding quality of the product. Assume that the wet weight data is , the wet weight of each sample is recorded by an electronic balance. In order to eliminate the influence of moisture on quality, the product is subjected to standard drying treatment. The moisture in the product is removed to obtain the dry weight data. The dry weight data is recorded as The sample is placed in an oven for treatment until its weight stabilizes, and the dry weight of the product is measured using an electronic balance. and dry weight After that, differential calculation is performed to obtain the material density distribution data of the product. Assuming that the volume of the product The material density of the product is obtained by three-dimensional measurement or estimation through geometric formula. Calculated by the following formula:

[0135] ;

[0136] in, is the dry weight, is the volume of the product. By calculating the material density of all products, a density distribution data set is obtained. The product is scanned by a three-coordinate measuring machine to obtain the product's geometric dimension data. The three-coordinate measuring machine measures the length, width and thickness of the product, which are recorded as , and The material density distribution data and geometric dimension data are standardized to eliminate the influence of different dimensions and magnitudes on the analysis results. Assume that the material density distribution data is And the geometrical dimension data is ,in Indicates The data obtained after standardization is:

[0137] ;

[0138] in, and are the mean and standard deviation of the material density, and are the mean and standard deviation of the geometric dimension data. Multi-dimensional evaluation is performed on the standardized feature data. The impact of multiple factors on product quality, such as material density and geometric dimensions, is considered comprehensively. By constructing a quality scoring matrix, each feature of each sample is weighted and evaluated. Assume that the quality scoring matrix is ​​recorded as , where each row represents the evaluation result of a sample and each column represents the score of a feature. The quality score matrix is ​​calculated by the following formula:

[0139] ;

[0140] in, Indicates The samples in The ratings on the features, It is The samples in Data on standardized features, It is The weight of the feature. Indicates the degree of influence of the feature on the overall quality score. The overall quality score of each sample is obtained by weighted summing the scores of all standardized features. Principal component analysis is performed based on the quality score matrix. The high-dimensional quality score matrix is ​​reduced by linear transformation to extract the principal component that best explains the data variation. Assume that the quality score matrix is , find a new coordinate system that maximizes the variance of the data in this coordinate system. Principal component analysis obtains the principal components by solving the following eigenvalue problem:

[0141] ;

[0142] in, It is principal components, is the corresponding eigenvalue. Indicates the size of the variance that the principal component can explain. By selecting the first few principal components with larger eigenvalues, the dimension of the data is reduced while retaining most of the variation information. Based on the results of principal component analysis, the target influencing factor is obtained. Multi-dimensional weight calculation is performed based on the target influencing factor. By evaluating each target influencing factor, the relative importance of each factor to product quality is determined. This is achieved through the following formula:

[0143] ;

[0144] in, is the final weight, It is the eigenvalue obtained through principal component analysis. The final quality control index is obtained by calculating the weight of each target influencing factor.

[0145] In a specific embodiment, the process of executing step S6 may specifically include the following steps:

[0146] The quality control index is combined with the molding process parameters to perform data matching to obtain process parameter matching data, and quality prediction is performed based on the process parameter matching data to obtain a parameter correlation matrix;

[0147] Perform parameter sensitivity analysis based on the parameter correlation matrix to obtain parameter weight data, and import the parameter weight data into the process control system to obtain process verification data;

[0148] Perform statistical analysis on process validation data to obtain process stability data, and perform parameter compensation based on the process stability data to obtain parameter correction data;

[0149] The parameter correction data is applied to the continuous production process. The production parameters and quality data are recorded through a real-time monitoring system to obtain process control data. Multiple rounds of production verification and parameter optimization are performed on the process control data to obtain the target process specification data.

[0150] Specifically, the quality control index is combined with the molding process parameters for data matching to obtain the process parameter matching data. Effectively combine data from different sources, including quality control data from the production process (such as wet weight, dry weight, geometric dimensions, etc.) and input parameters of the molding process (such as temperature, pressure, time, speed, etc.). Assume that the quality control index is , which includes wet weight, dry weight, geometric dimensions, material density and other data, and the process parameter combination is , then the process parameter matching data is expressed as:

[0151] ;

[0152] Quality prediction is performed based on process parameter matching data. The relationship between process parameters and quality control indicators is modeled using regression analysis or other prediction models to achieve quality prediction. Assuming that a linear regression model is used for quality prediction, the prediction function is:

[0153] ;

[0154] in, is a quality control indicator for the prediction, is the intercept, Each process parameter The relevant regression coefficients, is the number of process parameters. By training the model, a parameter correlation matrix is ​​obtained, which represents the correlation between each process parameter and the quality control index. Parameter sensitivity analysis is performed based on the parameter correlation matrix to identify which process parameters have a significant impact on the quality control index. Parameter sensitivity analysis is performed by calculating the partial derivative of the parameter to the quality prediction function. Assume that the prediction function of the quality control index and the process parameter is , then for the Process parameters , the sensitivity analysis is performed using the following formula:

[0155] ;

[0156] in, Indicates The sensitivity of each process parameter reflects the degree of influence of process parameter changes on quality control indicators. Parameters with higher sensitivity mean that they have a greater impact on product quality and require special attention. By analyzing the sensitivity of each process parameter, the weight data of the parameter is obtained. The weight is expressed as the relative importance of the parameter's contribution to the quality control indicator. Assume that the weight data is , and its calculation formula is:

[0157] ;

[0158] in, It is The weight of each process parameter indicates the proportion of the parameter's influence on the overall quality control. The parameter weights are imported into the process control system for process verification. The process control system regulates the process parameters in the production process according to these weights to ensure that the product quality reaches the expected target. In the process verification stage, the process verification data is processed through statistical analysis. The process verification data is the actual production data obtained based on the optimized parameter combination, including quality control indicators and various process parameters in the production process. By performing statistical analysis on these data, process stability data is obtained. Process stability data can help the system evaluate whether there are unstable factors in the production process, such as excessive temperature fluctuations, rapid pressure changes, etc. Assume that the process stability data is , which is obtained by calculating the variance or standard deviation:

[0159] ;

[0160] in, is the number of process parameters, It is The actual value of the process parameters, is the mean of the process parameters. A smaller standard deviation indicates a higher process stability. Parameter compensation is performed based on the process stability data. Assume that the parameter compensation data is , the process parameters are corrected by the following formula:

[0161] ;

[0162] in, is the compensation amount, which is adjusted based on the process stability analysis results. The compensated parameters make the production process more stable, thereby improving the consistency of product quality. The corrected parameters are applied to the continuous production process, and the production parameters and quality data are recorded through the real-time monitoring system. Assume that the real-time monitoring data is , including various process parameters in the production process and their corresponding quality control data. During the production process, the real-time monitoring system records the production process of each sample, and the data is described by the following formula:

[0163] ;

[0164] Through multiple rounds of production verification and parameter optimization of process control data, the optimization effect of the production process is obtained. These verification processes adjust and optimize process parameters multiple times to ensure that each link in the production process can operate in the optimal state. The target process specification data is obtained by summarizing and analyzing multiple rounds of verification data.

[0165] The above describes the process parameter analysis method of the foam material molding process in the embodiment of the present invention. The following describes the process parameter analysis system of the foam material molding process in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a process parameter analysis system for a foam material molding process includes:

[0166] The acquisition module is used to collect parameters and perform data statistics on the MPPO material under the standard process condition of 160°C to obtain the process parameter benchmark data;

[0167] The test module is used to perform an orthogonal test on the material pressure parameters based on the process parameter benchmark data to obtain a pressure process parameter database;

[0168] An analysis module is used to analyze the molding temperature gradient based on the pressure-carrying process parameter database, sampling and testing from 160°C to 190°C at intervals of 5°C to obtain a temperature control curve;

[0169] The compensation module is used to perform coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination;

[0170] The calculation module is used to detect the wet weight, dry weight and size of the products produced under the combination of molding process parameters, and to calculate multi-dimensional parameters to obtain quality control indicators;

[0171] The output module is used to perform correlation verification and parameter correction on the combination of quality control indicators and molding process parameters, and output target process specification data.

[0172] Through the synergy of the above components and the systematic parameter analysis method, the molding temperature is increased from 160℃ to 180-190℃, and combined with the precise control of 7-8KG load pressure, the molding cycle is significantly shortened and the process parameter system is optimized. Based on the orthogonal test of the load pressure process parameters and the temperature gradient analysis, a complete process parameter database is constructed, which realizes the rapid optimization and adjustment of the production process and improves the production cycle. Through multi-dimensional parameter calculation and quality control index system, a complete product quality evaluation method is established, which effectively reduces the incidence of defects such as product sticking and deformation. Using algorithms such as Gaussian linear regression and bidirectional long short-term memory network, intelligent coupling analysis and dynamic compensation of temperature-pressure parameters are realized, and the control accuracy of process parameters is improved. Through the optimization of process parameters and the establishment of quality control indicators, equipment investment and energy consumption are reduced, while the product yield rate is improved and production costs are reduced.

[0173] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0174] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0175] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0176] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0177] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0178] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0179] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing process parameters of a foam material molding process, characterized in that: The method comprises: Under the standard process condition of 160°C, the MPPO material is subjected to parameter collection and data statistics to obtain process parameter benchmark data; specifically, the method comprises: conducting sample continuous production test on the MPPO material under the standard process condition of 160°C, and collecting process original data corresponding to the feed speed, silo temperature, mold temperature and molding time; based on the process original data, calculating the mold sticking area, mold sticking degree and mold sticking position of the sample to obtain mold sticking characteristic data, and performing full-size measurement and surface quality scanning inspection on the sample according to the process original data to obtain product physical property data; performing correlation analysis on the mold sticking characteristic data and the product physical property data to obtain quality quantification data, and performing normal distribution test and outlier screening on the quality quantification data to obtain parameter distribution data; performing least squares fitting based on the parameter distribution data to obtain parameter fitting data, and performing time series analysis on the parameter fitting data to obtain process parameter benchmark data; Based on the process parameter benchmark data, an orthogonal test of the material pressure parameters is performed to obtain a pressure process parameter database; specifically comprising: inputting the process parameter benchmark data into the pressure control system for parameter configuration, applying pressure to the feed system through a piston structure driven by a servo motor, and obtaining pressure device initialization data; dividing the pressure time into five levels of 10 seconds, 20 seconds, 30 seconds, 40 seconds, and 50 seconds, collecting pressure data in different time periods through a pressure sensor, and obtaining time classification parameters; dividing the pressure into three levels of 7KG, 7.5KG, and 8KG, collecting material change data in different pressure sections through a sensor, and obtaining pressure classification parameters; orthogonally combining the time classification parameters and the pressure classification parameters, generating multiple groups of experimental schemes through parameter pairing, and obtaining a pressure test matrix; performing material pressure test on each group of parameters in the pressure test matrix, collecting pressure curves and temperature curves through sensors, and obtaining process monitoring data; performing multi-dimensional analysis on the process monitoring data, calculating material flow status and temperature distribution characteristics, and obtaining a pressure process parameter database; Based on the pressure-carrying process parameter database, a molding temperature gradient analysis is performed, and sampling and testing are performed from 160° C. to 190° C. at intervals of 5° C. to obtain a temperature control curve; Perform coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination; Conducting wet weight, dry weight and size testing on the products produced under the molding process parameter combination, and performing multi-dimensional parameter calculation to obtain quality control indicators; The quality control index and the molding process parameter combination are associated with verification and parameter correction, and target process specification data is output.

2. The process parameter analysis method for foam material molding process according to claim 1, characterized in that: The molding temperature gradient analysis is performed based on the pressure-carrying process parameter database, and sampling and testing are performed from 160° C. to 190° C. at intervals of 5° C. to obtain a temperature control curve, including: The process data in the pressure-carrying process parameter database is subjected to temperature segmentation setting, and 7 test temperature points are set at intervals of 5° C. from 160° C. to 190° C. to obtain a temperature test sequence; Based on the temperature test sequence, the load pressure parameter is kept unchanged at each temperature point to continuously produce samples, and temperature data is collected by an infrared thermometer to obtain temperature sampling data; Recording and counting the molding cycle time in the temperature sampling data to obtain molding cycle data, and correlating and calculating the molding cycle data with the equipment energy consumption data to obtain energy consumption analysis data; Dividing the energy consumption analysis data into temperature intervals to obtain temperature efficiency data, and mapping and analyzing the temperature efficiency data with product quality test results to obtain quality-related data; A temperature gradient model is established based on the mass correlation data, temperature optimization data is determined through numerical calculation, and curve fitting and smoothing processing are performed on the temperature optimization data to obtain a temperature control curve.

3. The process parameter analysis method for foam material molding process according to claim 2, characterized in that: The coupling analysis and pressure dynamic compensation of the temperature control curve and the pressure-carrying process parameter database are performed to obtain a molding process parameter combination, including: Performing time sequence alignment and data preprocessing on the temperature control curve and the pressure-loading process parameter database to obtain a standardized parameter matrix; The standardized parameter matrix is ​​input into a Gaussian linear regression model, and the temperature-pressure coupling coefficient is calculated by maximum likelihood estimation to obtain parameter coupling data; Inputting the parameter coupling data into a bidirectional long short-term memory network, and calculating a pressure compensation sequence through forward propagation and backward propagation; A population initialization scheme of a non-dominated sorting genetic algorithm is constructed based on the pressure compensation sequence, and an initial pressure compensation solution is obtained through encoding and decoding; Performing fitness evaluation and non-dominated sorting on the initial pressure compensation solutions, and obtaining a multi-objective optimization solution set through crossover and mutation operations; The multi-objective optimization solution set is screened through crowding calculation and elite retention strategy to obtain the Pareto optimal solution, and decision variables are extracted and parameter mapped for the Pareto optimal solution, and a molding process parameter combination is obtained through reverse decoding.

4. The process parameter analysis method for foam material molding process according to claim 3, characterized in that: The wet weight, dry weight and size of the product produced under the molding process parameter combination are tested, and multi-dimensional parameter calculation is performed to obtain quality control indicators, including: The molding process parameter combination is applied to product production, wet weight data of products off the production line are collected by an electronic balance to obtain wet weight parameters, and standard drying treatment is performed on the products off the production line, and dry weight parameters of the products are measured by an electronic balance; Perform differential calculation based on the wet weight parameter and the dry weight parameter to obtain material density distribution data of the product, scan the product with a three-dimensional coordinate measuring machine, obtain the length, width, and thickness values ​​of the product through a data acquisition system, and obtain geometric dimension data; Standardizing the material density distribution data and the geometric dimension data to obtain standardized characteristic data, and performing multi-dimensional evaluation based on the standardized characteristic data to obtain a quality scoring matrix; A principal component analysis is performed based on the quality score matrix to obtain a target influencing factor, and a multi-dimensional weight calculation is performed based on the target influencing factor to obtain a quality control indicator.

5. The process parameter analysis method for foam material molding process according to claim 4, characterized in that: The method of performing correlation verification and parameter correction on the combination of the quality control index and the molding process parameter and outputting target process specification data includes: The quality control index is combined with the molding process parameter to perform data matching to obtain process parameter matching data, and quality prediction is performed based on the process parameter matching data to obtain a parameter correlation matrix; Performing parameter sensitivity analysis based on the parameter correlation matrix to obtain parameter weight data, and importing the parameter weight data into a process control system to obtain process verification data; Performing statistical analysis on the process verification data to obtain process stability data, and performing parameter compensation based on the process stability data to obtain parameter correction data; The parameter correction data is applied to the continuous production process, the production parameters and quality data are recorded by a real-time monitoring system to obtain process control data, and the process control data is subjected to multiple rounds of production verification and parameter optimization to obtain target process specification data.

6. A process parameter analysis system for a foam material molding process, characterized in that: A method for analyzing process parameters of a foaming material molding process according to any one of claims 1 to 5, the system comprising: The acquisition module is used to collect parameters and perform data statistics on the MPPO material under the standard process condition of 160°C to obtain the process parameter benchmark data; A testing module, used to perform an orthogonal test of the pressure-loading parameters on the material based on the process parameter benchmark data to obtain a pressure-loading process parameter database; An analysis module is used to perform a molding temperature gradient analysis based on the pressure-carrying process parameter database, sampling and testing from 160° C. to 190° C. at intervals of 5° C. to obtain a temperature control curve; A compensation module, used for performing coupling analysis and pressure dynamic compensation on the temperature control curve and the pressure-carrying process parameter database to obtain a molding process parameter combination; A calculation module, used to detect the wet weight, dry weight and size of the product produced under the molding process parameter combination, and perform multi-dimensional parameter calculation to obtain quality control indicators; The output module is used to perform correlation verification and parameter correction on the combination of the quality control index and the molding process parameter, and output target process specification data.

7. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and is characterized in that when the processor executes the computer program, the process parameter analysis method of the foaming material molding process described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the processor is enabled to execute the process parameter analysis method for a foaming material molding process according to any one of claims 1 to 5.

Images