A multi-objective optimization method and system for plastic part injection molding parameters

Abstract

The invention provides a method and system for multi-objective optimization of injection molding parameters of plastic parts, and relates to the technical field of optimization of injection molding processing parameters. The invention collects acoustic signal data and heat flow change data in an injection mold cavity in real time, combines fast Fourier transform and differential scanning calorimetry, and quantitatively analyzes the flow characteristics and crystallinity of a melt material. On this basis, an improved non-dominated sorting genetic algorithm is used to optimize the injection speed, injection pressure and cooling time in the injection molding process parameters according to the correlation between an acoustic evaluation coefficient, a heat flow change coefficient and molding quality. The method can not only accurately predict and adjust various process parameters so that the molding quality of the plastic parts reaches the best state, but also ensures that the optimization process can dynamically adapt to changes in actual production through an iterative feedback mechanism.

Description

Technical Field

[0001] The invention relates to the technical field of injection molding process parameter optimization, and in particular to a multi-objective optimization method and system for plastic part injection molding parameters. Background Art

[0002] Injection molding has gradually become an indispensable processing method in the manufacturing industry. However, traditional injection molding process parameter optimization methods mostly rely on experience and experiments, which are difficult to cope with the increasingly complex demands of plastic materials and products. In recent years, researchers have introduced a variety of advanced optimization technologies in the injection molding process, such as designed experimental methods based on statistical analysis, neural network prediction models, and genetic algorithm optimization, in order to improve product quality while reducing production costs.

[0003] In the prior art, the publication number is CN110377948A, and the name is a multi-objective optimization method for injection molding parameters. It is a method for processing multi-objective optimization design based on the improved Kriging agent model Gkriging, the non-dominated sorting genetic algorithm NSGA-Ⅱ and the fuzzy decision Gkriging-NSGA-vague strategy of the vague set. The injection mold runner cross-sectional size parameters and the injection molding process parameters are selected as the decision variables to be optimized. The maximum volume shrinkage rate of the product, the total volume of the runner and the molding cycle are respectively used as evaluation indicators of product quality, production cost and production efficiency. A model is established and the optimal value of the multi-objective quality is obtained, and the multi-objective optimization of the comprehensive quality of the product is achieved through the multi-objective decision method.

[0004] The existing injection molding parameter optimization technology still has many shortcomings: First, the existing technology has limitations in collecting and analyzing real-time data in the mold cavity. Most methods only focus on single parameters such as temperature and pressure, while ignoring more complex multidimensional data such as acoustic signals and heat flow changes. This results in the optimization results often being unable to fully reflect the actual flow characteristics and crystallinity changes of the melt material;

[0005] Secondly, when dealing with multi-objective problems, traditional optimization methods usually cannot effectively take into account the mutual influence between various objectives, resulting in excellent optimization results in some indicators but poor performance in other indicators. In addition, the existing methods lack a dynamic feedback mechanism when dealing with complex models and cannot adjust the optimization strategy according to real-time production conditions, resulting in the inability to achieve the best molding quality in practical applications.

[0006] Therefore, a new injection molding parameter optimization method is needed that can comprehensively consider multiple factors such as acoustic signals and heat flow changes, perform multi-objective optimization, and ensure that the molding quality of plastic parts reaches the best state in all indicators.

[0007] The above information disclosed in the above Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0008] The object of the present invention is to provide a method and system for multi-objective optimization of injection molding parameters of plastic parts to solve the problems raised in the above background technology.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] A multi-objective optimization method for injection molding parameters of plastic parts, the specific steps include:

[0011] Step S1: real-time acquisition of acoustic signal data in the injection mold cavity, the acoustic signal data including pressure fluctuations and melt flow noise in the mold cavity; and acquisition of heat flow change data of the melt material in the injection mold cavity during the melting and cooling process;

[0012] Step S2: Set the time period T immediately before the current moment as the extraction segment. In the extraction segment T, the acoustic signal data and the heat flow change data collected in step S1 are synchronously extracted at the same time interval, and the acoustic signal data and the heat flow change data correspond one to one. According to the extraction segment T and the time interval, the extraction times set is determined to be {1, 2, ..., n}, where i∈{1, 2, ..., n} represents the index of the i-th data extraction, and n represents the total number of data extractions in the extraction segment T;

[0013] Step S3: acquiring historical data of the molding quality of plastic parts under different acoustic signal data and heat flow change data, wherein the molding quality data of the plastic parts includes a volume shrinkage value, a warpage deformation value, and a surface cavitation value, and determining a qualified threshold value of each parameter value in the molding quality data according to the historical data;

[0014] Step S4: after receiving the acoustic signal data extracted n times in the extraction section T, the acoustic frequency domain features of the acoustic signal data are extracted using a fast Fourier transform, and the extracted acoustic frequency domain features are analyzed and processed to generate an acoustic evaluation coefficient, which is used to determine the flow characteristics of the melt material and set a flow characteristic threshold;

[0015] Step S5: After receiving the heat flow change data of n extractions in the extraction section T, the heat flow change coefficient is generated by analyzing and processing. The heat flow change coefficient is used to quantitatively evaluate the crystallinity change trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed using differential scanning calorimetry DSC to obtain the crystallinity value of n extractions;

[0016] Step S6: performing a correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, and selecting a model with the highest fitting degree from the linear model and the nonlinear model according to the correlation analysis result to obtain a prediction model, the prediction model is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and according to the prediction model, setting a target crystallinity value and a corresponding target heat flow variation coefficient;

[0017] Step S7: The acoustic evaluation coefficient approaches the flow characteristic threshold, the heat flow variation coefficient approaches the target heat flow variation coefficient, the crystallinity value approaches the target crystallinity value, and all molding quality parameters are below the corresponding qualified threshold as optimization goals, and the improved non-dominated sorting genetic algorithm is selected as the multi-objective optimization algorithm. The injection molding process parameters are adjusted through the iterative feedback mechanism, and the current optimal injection molding process parameters are calculated. The injection molding process parameters include injection speed, injection pressure and cooling time.

[0018] Furthermore, the qualified thresholds of various parameter values ​​in the molding quality data are determined based on the historical data, including:

[0019] The volume shrinkage value, warpage deformation value and surface cavitation value are marked as V s , W d and S b ; Define V s , W d , S b The qualified thresholds are

[0020] Furthermore, the extracted acoustic frequency domain features are analyzed and processed to generate acoustic evaluation coefficients, which are used to determine the flow characteristics of the melt material, including:

[0021] For each extracted acoustic time domain signal value P i Perform fast Fourier transform to extract its frequency domain features; the expression of fast Fourier transform is:

[0022]

[0023] Among them, Y(f) is the frequency domain signal, P(t) is the acoustic time domain signal value, N is the signal length, and f is the sound wave frequency;

[0024] For each extracted acoustic frequency domain feature, the corresponding acoustic time domain signal value P i , use wavelet transform for time-frequency analysis, the formula of wavelet transform is:

[0025]

[0026] Among them, W(a,b) is the wavelet coefficient, a is the scale factor, b is the translation factor; ψ(t) is the selected mother wavelet;

[0027] The acoustic evaluation coefficient C is calculated by the wavelet coefficient W(a,b) obtained from the wavelet transform. The calculation steps are as follows:

[0028] Select key wavelet coefficients from W(a,b) and calculate their mean and standard deviation:

[0029]

[0030] Among them, M is the number of selected key wavelet coefficients, It represents the j′th wavelet coefficient obtained by wavelet transform after the i-th data extraction;

[0031] The calculation formula of acoustic evaluation coefficient C is as follows:

[0032]

[0033] Where k is the adjustment coefficient, 0.82≤k≤2.5; the flow characteristic threshold of the acoustic evaluation coefficient C is set to C th ;

[0034] As C approaches C th When the flow characteristics of the melt material are better;

[0035] As C moves away from C th The flow characteristics of the melt material are worse.

[0036] Further, after receiving the heat flow change data extracted n times in the extraction section T, the heat flow change coefficient is generated by analyzing and processing. The heat flow change coefficient is used to determine the crystallinity change trend of the melt material, specifically including:

[0037] According to the construction data extraction set {(P 1 ,Q 1 ),(P 2 ,Q 2 ),…,(P n ,Q n )}, determine the extracted heat flow change value set as {Q 1 ,Q 2 ,…,Q n}, define the heat flow variation coefficient as C Q , the calculation content is as follows:

[0038] Calculate the mean value μ of the heat flow change Q :

[0039]

[0040] Calculate the standard deviation of the heat flow change σ Q :

[0041]

[0042] The heat flow variation coefficient C Q Defined as the ratio of the standard deviation of the heat flow variation to the mean:

[0043]

[0044] Among them, C Q Expressed as a percentage, it indicates the relative degree of fluctuation of heat flow;

[0045] Differential scanning calorimetry (DSC) was used to perform thermal analysis on the melt material sample. The specific operation was as follows:

[0046] 1) Heat and cool the sample using a DSC device, record the relationship between the heat flow change value Q(t) and the temperature Tw of the sample during the whole process, and draw a heat flow change curve;

[0047] 2) On the heat flow curve, identify the melting temperature Tm and crystallization temperature Tc:

[0048] 3) Define the crystallinity X of the melt material c , the calculation formula of crystallinity is as follows:

[0049]

[0050] The melting heat ΔH is obtained by the melting peak area of ​​the heat flow curve. m The cold crystallization heat ΔH is obtained by the crystallization peak area of ​​the heat flow curve. c ; is the standard melting heat of completely crystalline materials, and the corresponding crystallinity X is calculated according to different time points in the extraction section T. c The values ​​are plotted to form a trend graph of crystallinity changing over time.

[0051] Furthermore, the heat flow variation coefficient C Q The calculation parameter Q i and crystallinity X c A correlation analysis is performed between them, and the correlation coefficient is defined as r. The calculation formula is as follows:

[0052]

[0053] Among them, X ci is the crystallinity of the ith data extraction, is the mean value of crystallinity;

[0054] If r is closer to 1, it means that there is a positive correlation between the heat flow variation coefficient and the crystallinity, and the greater the fluctuation of the heat flow variation data, the higher the crystallinity;

[0055] If r is closer to -1, it means that there is a negative correlation between the heat flow variation coefficient and the crystallinity. The greater the fluctuation of the heat flow variation data, the lower the crystallinity.

[0056] If r is closer to 0, it means that the linear relationship between the heat flow variation coefficient and the crystallinity is less obvious;

[0057] Select the model with the highest fitting degree from the linear model and the nonlinear model for fitting; then obtain the prediction model. The fitted prediction model can calculate the C Q Predict the corresponding X c ;

[0058] Based on the results of the prediction model, the injection molding process parameters including injection speed, injection pressure and cooling time are adjusted to achieve the target crystallinity value:

[0059] Set the target crystallinity value to The target heat flux variation coefficient required by the model is:

[0060] Furthermore, the improved non-dominated sorting genetic algorithm is selected as the multi-objective optimization algorithm to calculate the current optimal injection molding process parameters, which include injection speed, injection pressure and cooling time, specifically including:

[0061] 1) Initial conditions and goal setting;

[0062] For optimization goals:

[0063] The acoustic evaluation coefficient C approaches the flow characteristic threshold C th ;

[0064] Heat flow variation coefficient C Q Approaching the target heat flux variation coefficient

[0065] Crystallinity value X c Approaching the target crystallinity value

[0066] Constraints: The values ​​of various parameters of molding quality must be within the qualified threshold:

[0067] Volume shrinkage Warpage value Surface cavitation value

[0068] 2) Generate an initial population;

[0069] Generate an initial population of N′ individuals, each of which represents a set of injection molding process parameter combinations G i′ =(V sd ,P yl ,t cool ), where V sd , P yl ,t cool Represents injection speed, injection pressure and cooling time respectively:

[0070] 3) Definition and calculation of fitness function:

[0071] Define the fitness function F(G i′ ) is used to evaluate each individual G i′ The formula is as follows:

[0072]

[0073] Where, C is the acoustic evaluation coefficient; C Q is the coefficient of variation of heat flow, X c is the crystallinity value, w 1 、w 2 、w 3 is the weight value of the corresponding parameter, and Indicates that the influence of each optimization objective is equal;

[0074] 4) Genetic operations: Generate the next generation of population through selection, crossover and mutation;

[0075] 5) Constraint processing and penalty function:

[0076] In each iteration, the molding quality parameter V s , W d , S b Evaluate; if any of the individuals does not meet The corresponding qualified threshold is: then the fitness function F(G i ) to impose punishment;

[0077] 6) Non-dominated sorting and Pareto frontier updating;

[0078] At the end of each generation, all individuals in the population are non-dominated and sorted, and the Pareto frontier is updated to obtain the optimal solution set;

[0079] 7) Iteration and convergence;

[0080] The maximum number of iterations is set to L1 generation, and the optimal solution is continuously approached through multiple rounds of genetic operations and non-dominated sorting;

[0081] 8) Optimal solution extraction and practical application;

[0082] After the iteration, the optimal solution G*, that is, the optimal injection molding process parameter combination, is extracted from the final Pareto frontier and obtained

[0083] A multi-objective optimization system for plastic part injection molding parameters, the system is used to execute the multi-objective optimization method for plastic part injection molding parameters, comprising:

[0084] Data acquisition module: used to collect acoustic signal data in the injection mold cavity in real time, including pressure fluctuations and melt flow noise in the mold cavity; at the same time, it collects heat flow change data of the melt material in the injection mold cavity during the melting and cooling process;

[0085] Data extraction module: used to set the time period T immediately before the current moment as the extraction segment, and synchronously extract the acoustic signal data and heat flow change data collected in step S1 at the same time interval within the extraction segment T, and the acoustic signal data and the heat flow change data correspond one to one, and determine the extraction times set as {1,2,…,n} according to the extraction segment T and the time interval, where i∈{1,2,…,n} represents the index of the i-th data extraction, and n represents the total number of data extractions within the extraction segment T;

[0086] Qualified threshold determination module: used to obtain the historical data of plastic parts molding quality under different acoustic signal data and heat flow change data, the molding quality data of plastic parts includes volume shrinkage value, warpage deformation value and surface cavitation value, and determine the qualified threshold of each parameter value in the molding quality data according to the historical data;

[0087] Evaluation coefficient construction module: used for receiving acoustic signal data extracted n times in extraction section T, extracting acoustic frequency domain features of acoustic signal data using fast Fourier transform, analyzing and processing the extracted acoustic frequency domain features, and generating acoustic evaluation coefficients. The acoustic evaluation coefficients are used to determine the flow characteristics of the melt material and set the flow characteristic threshold;

[0088] Variation coefficient building module: used to receive the heat flow variation data of n extractions in the extraction section T, analyze and process it, and generate the heat flow variation coefficient. The heat flow variation coefficient is used to quantitatively evaluate the crystallinity variation trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed using differential scanning calorimetry DSC to obtain the crystallinity value of n extractions;

[0089] Optimization target value determination module: used to perform correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, select the model with the highest fitting degree from the linear model and the nonlinear model for fitting, so as to obtain a prediction model, which is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and set the target crystallinity value and the corresponding target heat flow variation coefficient according to the prediction model;

[0090] Multi-objective optimization module: It is used to take the acoustic evaluation coefficient approaching the flow characteristic threshold, the heat flow variation coefficient approaching the target heat flow variation coefficient, the crystallinity value approaching the target crystallinity value, and all molding quality parameters being below the corresponding qualified threshold as the optimization goals, select the improved non-dominated sorting genetic algorithm as the multi-objective optimization algorithm, adjust the injection molding process parameters through the iterative feedback mechanism, and calculate the current optimal injection molding process parameters. The injection molding process parameters include injection speed, injection pressure and cooling time.

[0091] Compared with the prior art, the beneficial effects of the present invention are as follows: by real-time acquisition of acoustic signal data and heat flow change data in the injection mold cavity, combined with fast Fourier transform and differential scanning calorimetry, the flow characteristics and crystallinity of the melt material are quantitatively analyzed; on this basis, an improved non-dominated sorting genetic algorithm is used to optimize the injection speed, injection pressure and cooling time in the injection molding process parameters according to the correlation between the acoustic evaluation coefficient, the heat flow change coefficient and the molding quality. This method can not only accurately predict and adjust various process parameters so that the molding quality of the plastic parts reaches the best state, but also ensure that the optimization process can dynamically adapt to changes in actual production through an iterative feedback mechanism, overcome the shortcomings of the prior art, and achieve all-round optimization of the injection molding process. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0093] Figure 2 It is a system module block diagram of the present invention. DETAILED DESCRIPTION

[0094] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0095] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0096] Embodiment 1:

[0097] See also Figure 1 , the present invention provides a technical solution:

[0098] A multi-objective optimization method for injection molding parameters of a plastic part is applied to the injection molding cavity of the plastic part, and the specific steps include:

[0099] Step S1: real-time acquisition of acoustic signal data in the injection mold cavity, the acoustic signal data including pressure fluctuations and melt flow noise in the mold cavity; and acquisition of heat flow change data of the melt material in the injection mold cavity during the melting and cooling process;

[0100] Step S2: Set the time period T immediately before the current moment as the extraction segment. In the extraction segment T, the acoustic signal data and the heat flow change data collected in step S1 are synchronously extracted at the same time interval, and the acoustic signal data and the heat flow change data correspond one to one. According to the extraction segment T and the time interval, the extraction times set is determined to be {1, 2, ..., n}, where i∈{1, 2, ..., n} represents the index of the i-th data extraction, and n represents the total number of data extractions in the extraction segment T;

[0101] Step S3: acquiring historical data of the molding quality of plastic parts under different acoustic signal data and heat flow change data, wherein the molding quality data of the plastic parts includes a volume shrinkage value, a warpage deformation value, and a surface cavitation value, and determining a qualified threshold value of each parameter value in the molding quality data according to the historical data;

[0102] Step S4: after receiving the acoustic signal data extracted n times in the extraction section T, the acoustic frequency domain features of the acoustic signal data are extracted using a fast Fourier transform (FFT), and the extracted acoustic frequency domain features are analyzed and processed to generate an acoustic evaluation coefficient, which is used to determine the flow characteristics of the melt material and set a flow characteristic threshold;

[0103] Step S5: After receiving the heat flow change data of n extractions in the extraction section T, the heat flow change coefficient is generated by analyzing and processing. The heat flow change coefficient is used to quantitatively evaluate the crystallinity change trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed using differential scanning calorimetry DSC to obtain the crystallinity value of n extractions;

[0104] Step S6: performing a correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, selecting a model with the highest fitting degree from the linear model and the nonlinear model for fitting to obtain a prediction model, the prediction model is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and setting a target crystallinity value and a corresponding target heat flow variation coefficient according to the prediction model;

[0105] Step S7: The acoustic evaluation coefficient approaches the flow characteristic threshold, the heat flow variation coefficient approaches the target heat flow variation coefficient, the crystallinity value approaches the target crystallinity value, and all molding quality parameters are below the corresponding qualified threshold as optimization goals, and the improved non-dominated sorting genetic algorithm is selected as the multi-objective optimization algorithm. The injection molding process parameters are adjusted through an iterative feedback mechanism, and the current optimal injection molding process parameters are calculated. The injection molding process parameters include injection speed, injection pressure and cooling time to achieve the best molding quality.

[0106] To further explain, the collection of acoustic signal data and heat flow change data is as follows:

[0107] For acoustic signal data acquisition:

[0108] A highly sensitive acoustic sensor is installed in the mold cavity of the injection molding machine. In this embodiment, a piezoelectric sensor is used. The sensor can collect pressure fluctuations and melt flow noise in the mold cavity in real time.

[0109] The sampling frequency of acoustic signal data acquisition is set to 10kHz to ensure that the slight changes in the melt flow process are captured; the acoustic signal data is represented by the acoustic time domain signal value P(t). At the same time, the calculation formula of P(t) is as follows:

[0110] P(t)=A·sin(2πft+φ)

[0111] Among them, A is the sound amplitude, f is the sound wave frequency, φ is the phase angle, and t is the time. By analyzing these signals, the flow characteristics of the melt can be obtained;

[0112] For heat flow variation data acquisition:

[0113] In this embodiment, the melt material is used as a sample for data collection, and a differential scanning calorimeter (DSC) sensor is installed in the injection mold cavity to regularly collect heat flow change data of the melt material during the melting and cooling process; the sensor is connected to the temperature control system of the mold cavity to accurately control the temperature change;

[0114] The measurement cycle is set to once per second, and the heat flow change data is expressed as the heat flow change value Q(t). The calculation formula of Q(t) is as follows:

[0115] Q(t)=m·Cb·ΔT

[0116] Among them, m is the mass of the melt, Cb is the specific heat capacity of the melt, and ΔT is the temperature change in the time interval; by analyzing the heat flow change data, the change in the crystallinity of the melt can be directly measured, and then the fluidity and crystallization characteristics of the melt during the molding process can be determined.

[0117] Further explanation: extracting segment T and synchronously extracting acoustic signal data and heat flow change data specifically include:

[0118] Set the current time t cur The immediately preceding time period T is used as the data extraction segment, with a length of T = t cur -t prev , where t prev is the time of the previous moment; the length of the extraction segment T is set to a specific value, which is 10 seconds in this embodiment, to ensure that the selected data can reflect the characteristics of melt flow and crystallization changes; the time interval is set to Δt; the time interval Δt is selected to be 0.1 second to ensure sufficient resolution and facilitate detailed data analysis;

[0119] In the extraction section T, an extraction is performed every time Δt. Specifically, the acoustic time domain signal value P(t) and the heat flow change value Q(t) collected in step S1 are synchronously extracted at the same time interval Δt. The data extracted each time are recorded as P i and Q i ; The total number of extractions n is calculated as:

[0120]

[0121] During the extraction process, ensure that each acoustic time domain signal value P i and heat flux change Q i One-to-one correspondence, that is, P i =P(t prev +i·Δt) and Q i =Q(t prev +i·Δt);

[0122] According to the number of extractions n, a data extraction set {(P 1 ,Q 1 ),(P 2 ,Q 2 ),…,(P n ,Q n )}, each data pair (P i , Qi) contains the acoustic time domain signal value and heat flow change value at the same time point in the extraction section T; this data set will be used for subsequent data analysis and feature extraction to help identify the change law of the melt material during the flow and crystallization process;

[0123] After the data extraction is completed, the data in the collection is corrected and preprocessed to eliminate noise and outliers; the moving average filter or Kalman filter algorithm is used to extract the acoustic time domain signal value P i and heat flux change Q i Smoothing is performed to ensure the accuracy and reliability of the data and the quality of input data for subsequent multi-objective optimization algorithms.

[0124] Further explanation: the qualified thresholds of various parameter values ​​in the molding quality data are determined based on historical data, including:

[0125] The volume shrinkage value, warpage deformation value and surface cavitation value are marked as V s , W d and S b ;

[0126] The collected historical data of molding quality parameters are organized into the following Table 1, in the following format:

[0127] Table 1

[0128]

[0129] Calculate the mean μ and standard deviation σ of each molding quality parameter using the following formula:

[0130]

[0131] Among them, X∈{V s , W d , S b}, n is the total number of data extractions in the extraction segment T, and the total number represents the total number of samples of historical data;

[0132] According to the statistical results of historical data, the qualified threshold of molding quality parameters is determined; according to industry standards or enterprise internal control standards, the qualified threshold is set as the mean plus a certain multiple of the standard deviation, and V s , W d , S b The qualified thresholds are

[0133]

[0134] Among them, k1, k2 and k3 are multiples of the control limits, 0.23≤k1≤1.2, 0.43≤k2≤1.68, 0.11≤k3≤1.58; the specific values ​​of k1, k2 and k3 are determined by the expert group based on experimental data.

[0135] It is further explained that the extracted acoustic frequency domain features are analyzed and processed to generate acoustic evaluation coefficients, which are used to determine the flow characteristics of the melt material, including:

[0136] For each extracted acoustic time domain signal value P i Perform fast Fourier transform to extract its frequency domain features; the expression of fast Fourier transform is:

[0137]

[0138] Among them, Y(f) is the frequency domain signal, P(t) is the acoustic time domain signal value, N is the signal length, j is the imaginary unit, and f is the sound wave frequency; through this transformation, the performance of the acoustic signal in the frequency domain can be obtained;

[0139] For each extracted acoustic time domain signal value P i , use wavelet transform for time-frequency analysis, the formula of wavelet transform is:

[0140]

[0141] Among them, W(a,b) is the wavelet coefficient, a is the scale factor, and b is the translation factor; ψ(t) is the selected mother wavelet; through wavelet transform, the local characteristics of the acoustic signal at different times and scales are obtained, thereby capturing the transient behavior of the melt flow;

[0142] The acoustic evaluation coefficient C is calculated by the wavelet coefficient W(a,b) obtained from the wavelet transform. The calculation steps are as follows:

[0143] Select key wavelet coefficients from W(a,b) and calculate their mean and standard deviation:

[0144]

[0145] Among them, M is the number of selected key wavelet coefficients, It represents the j′th wavelet coefficient obtained by wavelet transform after the i-th data extraction;

[0146] The calculation formula of acoustic evaluation coefficient C is as follows:

[0147]

[0148] Wherein, k is the adjustment coefficient, 0.82≤k≤2.5; k is determined by the expert group through experimental data, aiming to balance the flow characteristics and production efficiency;

[0149] Set the flow characteristic threshold of the acoustic evaluation coefficient C to C th ;

[0150] As C approaches C th When the flow characteristics of the melt material are better;

[0151] As C moves away from C th The flow characteristics of the melt material are worse.

[0152] Further explanation: after receiving the heat flow variation data extracted n times in the extraction section T, the heat flow variation coefficient is generated after analysis and processing. The heat flow variation coefficient is used to determine the crystallinity variation trend of the melt material, specifically including:

[0153] According to the construction data extraction set {(P 1 ,Q 1 ),(P 2 ,Q 2 ),…,(P n ,Q n )}, determine the extracted heat flow change value set as {Q 1 ,Q 2 ,…,Q n}, define the heat flow variation coefficient as C Q , the calculation content is as follows:

[0154] Calculate the mean value μ of the heat flow change Q :

[0155]

[0156] Calculate the standard deviation of the heat flow change σ Q :

[0157]

[0158] The heat flow variation coefficient C Q Defined as the ratio of the standard deviation of the heat flow variation to the mean:

[0159]

[0160] Among them, C Q Expressed as a percentage, it indicates the relative degree of fluctuation of heat flow;

[0161] Differential scanning calorimetry (DSC) was used to perform thermal analysis on the melt material sample. The specific operation was as follows:

[0162] 1) Heat and cool the sample using a DSC device, record the relationship between the heat flow change value Q(t) and the temperature Tw of the sample during the whole process, and draw a heat flow change curve;

[0163] 2) On the heat flow curve, identify the melting temperature Tm and crystallization temperature Tc:

[0164] Melting temperature Tm: During the heating process, the heat flow curve has a peak value, and the corresponding temperature is the melting temperature of the sample;

[0165] Crystallization temperature Tc: During the cooling process, the heat flow curve shows an exothermic peak, and the corresponding temperature is the crystallization temperature of the sample;

[0166] 3) Define the crystallinity X of the melt material c , the calculation formula of crystallinity is as follows:

[0167]

[0168] The melting heat ΔH is obtained by the melting peak area of ​​the heat flow curve. m The cold crystallization heat ΔH is obtained by the crystallization peak area of ​​the heat flow curve. c ; The standard heat of fusion of a completely crystalline material can be found in the literature or determined experimentally based on the properties of the material;

[0169] Calculate the corresponding crystallinity X at different time points in the extraction section T c Values ​​are calculated to form a trend graph of crystallinity changing over time; this trend graph can reflect the crystallization behavior of the material during the injection molding process;

[0170] Further explanation: the heat flow variation coefficient C Q The calculation parameter Q i and crystallinity X c A correlation analysis is performed between them, and the correlation coefficient is defined as r. The calculation formula is as follows:

[0171]

[0172] Among them, X ci is the crystallinity of the ith data extraction, is the mean value of crystallinity;

[0173] If r is closer to 1, it means that there is a positive correlation between the heat flow variation coefficient and the crystallinity, and the greater the fluctuation of the heat flow variation data, the higher the crystallinity;

[0174] If r is closer to -1, it means that there is a negative correlation between the heat flow variation coefficient and the crystallinity. The greater the fluctuation of the heat flow variation data, the lower the crystallinity.

[0175] If r is closer to 0, it means that the linear relationship between the heat flow variation coefficient and the crystallinity is less obvious;

[0176] For the linear model, assuming C Q With X c Linear relationship:

[0177] X c =h1·C Q +h2

[0178] Among them, h1 and h2 are coefficients obtained through regression analysis;

[0179] For nonlinear models, assuming that C Q With X c A nonlinear relationship:

[0180]

[0181] Among them, h3, h4 and h5 are the coefficients obtained by nonlinear regression;

[0182] According to the actual data distribution and the results of correlation analysis, the model with the highest fitting degree is selected from the linear model and the nonlinear model for fitting to obtain the prediction model. The fitted prediction model can be calculated based on the current measured C Q Predict the corresponding X c ;

[0183] Based on the results of the prediction model, the injection molding process parameters including injection speed, injection pressure and cooling time are adjusted to achieve the target crystallinity value:

[0184] Set the target crystallinity value to The target heat flux variation coefficient required by the model is:

[0185] For example, if the target crystallinity is And the prediction model is a linear model X c =h1·C Q +h2, required for reverse thrust

[0186]

[0187] It is further explained that the improved non-dominated sorting genetic algorithm (NSGA-II) is selected as the multi-objective optimization algorithm to calculate the current optimal injection molding process parameters. The injection molding process parameters include injection speed, injection pressure and cooling time, specifically including:

[0188] 1) Initial conditions and goal setting;

[0189] For optimization goals:

[0190] The acoustic evaluation coefficient C approaches the flow characteristic threshold C th In this embodiment, C th =70%;

[0191] Heat flow variation coefficient C Q Approaching the target heat flux variation coefficient In this embodiment

[0192] Crystallinity value X c Approaching the target crystallinity value In this embodiment

[0193] Constraints: The values ​​of various parameters of molding quality must be within the qualified threshold:

[0194] Volume shrinkage Warpage value Surface cavitation value

[0195] In this embodiment, The values ​​are 2%, 0.5%, and 0.2% respectively;

[0196] These thresholds are based on criteria set based on historical data and experimental results;

[0197] 2) Generate an initial population;

[0198] Generate an initial population containing N′ individuals. In this embodiment, N′=100, and each individual represents a set of injection molding process parameter combinations G i′ =(V sd ,P yl ,t cool ), where V sd , P yl ,t cool Represents injection speed, injection pressure and cooling time respectively:

[0199] Specifically, the injection speed V sd Randomly generated between 10-100mm / s;

[0200] Injection pressure P yl Randomly generated between 50-150MPa;

[0201] Cooling time cool Randomly generated between 10-40 seconds;

[0202] Make sure the initial population has enough diversity to cover the parameter space;

[0203] 3) Definition and calculation of fitness function:

[0204] Define the fitness function F(G i′ ) is used to evaluate each individual G i′ The formula is as follows:

[0205]

[0206] Where, C is the acoustic evaluation coefficient; C Q is the coefficient of variation of heat flow, X c is the crystallinity value, w 1 、w 2 、w 3 is the weight value of the corresponding parameter, and Indicates that the influence of each optimization objective is equal;

[0207] 4) Genetic operations: Generate the next generation of population through selection, crossover and mutation;

[0208] In each generation, the following genetic operations are performed:

[0209] Selection: Use the tournament selection method to select the individual with the smallest fitness (i.e., the smallest error) from the current population for crossover operation;

[0210] Crossover: Use single-point crossover to select two parent individuals for gene exchange and generate two new individuals (offspring); the crossover probability is set to P c =80%;

[0211] Mutation: with mutation probability P m =5%, randomly changing the parameter values ​​of offspring individuals to increase population diversity;

[0212] Through these operations, the next generation of population is generated;

[0213] 5) Constraint processing and penalty function:

[0214] In each iteration, the molding quality parameter V s , W d , S b Evaluate; if any of the individuals does not meet The corresponding qualified threshold is:

[0215] Then the fitness function F(G i′ ) is punished, and the specific penalty function is:

[0216] F′(G i′ )=F(G i′ )+Penalty

[0217] Among them, the penalty term Penalty = 10; the penalty function forces the optimization algorithm to tend to the solution that satisfies the constraints;

[0218] 6) Non-dominated sorting and Pareto frontier updating;

[0219] At the end of each generation, all individuals in the population are non-dominated and sorted, and the Pareto frontier is updated to obtain the optimal solution set;

[0220] Non-dominated sorting: For any two individuals G in the population i′ and G i′+1 , if G i′ Not inferior to G in all objectives i′+1 , and is better than G in at least one objective i′+1 , then G i′ Dominate G i′+1 ;

[0221] Pareto frontier: select non-dominated solutions to form the Pareto frontier as the current optimal solution set;

[0222] 7) Iteration and convergence;

[0223] The maximum number of iterations is set to L1 generations. In this embodiment, L1=200 generations is set. Through multiple rounds of genetic operations and non-dominated sorting, the optimal solution is continuously approached.

[0224] 8) Optimal solution extraction and practical application;

[0225] After the iteration, the optimal solution G*, that is, the optimal injection molding process parameter combination, is extracted from the final Pareto frontier and obtained

[0226] These parameters are applied in the actual production process, and the acoustic evaluation coefficient C and heat flow change coefficient C are detected through the real-time monitoring system. Q and crystallinity X c If the actual value deviates from the target value, the feedback mechanism is used to further adjust the injection molding process parameters until all optimization goals are met.

[0227] See also Figure 2 , a plastic part injection molding parameter multi-objective optimization system, the system is used to execute the plastic part injection molding parameter multi-objective optimization method, comprising:

[0228] Data acquisition module: used to collect acoustic signal data in the injection mold cavity in real time, including pressure fluctuations and melt flow noise in the mold cavity; at the same time, it collects heat flow change data of the melt material in the injection mold cavity during the melting and cooling process;

[0229] Data extraction module: used to set the time period T immediately before the current moment as the extraction segment, and synchronously extract the acoustic signal data and heat flow change data collected in step S1 at the same time interval within the extraction segment T, and the acoustic signal data and the heat flow change data correspond one to one, and determine the extraction times set as {1,2,…,n} according to the extraction segment T and the time interval, where i∈{1,2,…,n} represents the index of the i-th data extraction, and n represents the total number of data extractions within the extraction segment T;

[0230] Qualified threshold determination module: used to obtain the historical data of plastic parts molding quality under different acoustic signal data and heat flow change data, the molding quality data of plastic parts includes volume shrinkage value, warpage deformation value and surface cavitation value, and determine the qualified threshold of each parameter value in the molding quality data according to the historical data;

[0231] Evaluation coefficient construction module: used for receiving acoustic signal data extracted n times in extraction section T, extracting acoustic frequency domain features of acoustic signal data using fast Fourier transform, analyzing and processing the extracted acoustic frequency domain features, and generating acoustic evaluation coefficients. The acoustic evaluation coefficients are used to determine the flow characteristics of the melt material and set the flow characteristic threshold;

[0232] Variation coefficient building module: used to receive the heat flow variation data of n extractions in the extraction section T, analyze and process it, and generate the heat flow variation coefficient. The heat flow variation coefficient is used to quantitatively evaluate the crystallinity variation trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed using differential scanning calorimetry DSC to obtain the crystallinity value of n extractions;

[0233] Optimization target value determination module: used to perform correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, select the model with the highest fitting degree from the linear model and the nonlinear model for fitting, so as to obtain a prediction model, which is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and set the target crystallinity value and the corresponding target heat flow variation coefficient according to the prediction model;

[0234] Multi-objective optimization module: It is used to take the acoustic evaluation coefficient approaching the flow characteristic threshold, the heat flow variation coefficient approaching the target heat flow variation coefficient, the crystallinity value approaching the target crystallinity value, and all molding quality parameters being below the corresponding qualified threshold as the optimization goals, select the improved non-dominated sorting genetic algorithm as the multi-objective optimization algorithm, adjust the injection molding process parameters through the iterative feedback mechanism, and calculate the current optimal injection molding process parameters. The injection molding process parameters include injection speed, injection pressure and cooling time.

[0235] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0236] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0237] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0238] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A multi-objective optimization method for injection molding parameters of plastic parts, characterized in that: The specific steps include: Step S1: real-time acquisition of acoustic signal data in the injection mold cavity, the acoustic signal data including pressure fluctuations and melt flow noise in the mold cavity; and acquisition of heat flow change data of the melt material in the injection mold cavity during the melting and cooling process; Step S2: Set the time period T immediately before the current moment as the extraction section. In the extraction section T, the acoustic signal data and the heat flow change data collected in step S1 are synchronously extracted at the same time interval, and the acoustic signal data and the heat flow change data correspond one to one. According to the extraction section T and the time interval, the extraction times set is determined to be {1, 2, ..., n}, where i∈{1, 2, ..., n} represents the index of the i-th data extraction, and n represents the total number of data extractions in the extraction section T; Step S3: acquiring historical data of the molding quality of plastic parts under different acoustic signal data and heat flow change data, wherein the molding quality data of the plastic parts includes a volume shrinkage value, a warpage deformation value, and a surface cavitation value, and determining a qualified threshold value of each parameter value in the molding quality data according to the historical data; Step S4: after receiving the acoustic signal data extracted n times in the extraction section T, the acoustic frequency domain features of the acoustic signal data are extracted using fast Fourier transform, and the extracted acoustic frequency domain features are analyzed and processed to generate an acoustic evaluation coefficient, which is used to determine the flow characteristics of the melt material and set the flow characteristic threshold; specifically including: Perform fast Fourier transform on the acoustic time domain signal value extracted each time to extract its frequency domain features; For each extracted acoustic frequency domain feature, wavelet transform is used to perform time-frequency analysis; The acoustic evaluation coefficient C is calculated by the wavelet coefficients obtained from the wavelet transform. The calculation steps are as follows: Select key wavelet coefficients from the wavelet coefficients, and calculate the mean and standard deviation of the key wavelet coefficients respectively: The calculation formula of acoustic evaluation coefficient C is as follows: Among them, k is the adjustment coefficient, μ W is the mean of the key wavelet coefficients, σ W is the standard deviation of the key wavelet coefficients, 0.82≤k≤2.5; the flow characteristic threshold of the acoustic evaluation coefficient C is set to C th ; As C approaches C th When the flow characteristics of the melt material are better; As C moves away from C th When the flow characteristics of the melt material are poorer; Step S5: After receiving the heat flow change data of n extractions in the extraction section T, the heat flow change coefficient is generated by analyzing and processing. The heat flow change coefficient is used to quantitatively evaluate the crystallinity change trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed by differential scanning calorimetry to obtain the crystallinity value of n extractions; Step S6: performing a correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, and selecting a model with the highest fitting degree from the linear model and the nonlinear model according to the correlation analysis result to obtain a prediction model, the prediction model is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and according to the prediction model, setting a target crystallinity value and a corresponding target heat flow variation coefficient; Step S7: The acoustic evaluation coefficient approaches the flow characteristic threshold, the heat flow variation coefficient approaches the target heat flow variation coefficient, the crystallinity value approaches the target crystallinity value, and all molding quality parameters are below the corresponding qualified threshold as optimization goals, and the improved non-dominated sorting genetic algorithm is selected as the multi-objective optimization algorithm. The injection molding process parameters are adjusted through the iterative feedback mechanism, and the current optimal injection molding process parameters are calculated. The injection molding process parameters include injection speed, injection pressure and cooling time.

2. A multi-objective optimization method for injection molding parameters of plastic parts according to claim 1, characterized in that: The collection of acoustic signal data and heat flow change data is as follows: For acoustic signal data acquisition: The acoustic signal data is expressed as the acoustic time domain signal value P(t). At the same time, the calculation formula of P(t) is as follows: P(t)=A·sin(2πft+φ) Among them, A is the sound amplitude, f is the sound wave frequency, φ is the phase angle, and t is the time; For heat flow variation data acquisition: The heat flow change data is expressed as the heat flow change value Q(t), and the calculation formula of Q(t) is as follows: Q(t)=m·Cb·ΔT Where m is the mass of the melt, Cb is the specific heat capacity of the melt, and ΔT is the temperature change in the time interval; Set the current time t cur The immediately preceding time period T is used as the data extraction segment, with a length of T = t cur -t prev , where t prev is the time of the previous moment; the length of the extraction segment T is set to a specific value, and the time interval is set to Δt; In the extraction section T, an extraction is performed every time Δt. Specifically, the acoustic time domain signal value P(t) and the heat flow change value Q(t) collected in step S1 are synchronously extracted at the same time interval Δt. The data extracted each time are recorded as P i and Q i ; According to the number of extractions n, a data extraction set {(P1,Q1),(P2,Q2),…,(P n ,Q n )}, each data pair (P i ,Qi) contains the acoustic time domain signal value and heat flow change value at the same time point in the extraction segment T.

3. A multi-objective optimization method for plastic part injection molding parameters according to claim 2, characterized in that: Determine the qualified thresholds of various parameter values ​​in the molding quality data based on historical data, including: The volume shrinkage value, warpage deformation value and surface cavitation value are marked as V s , W d and S b ; Calculate the mean and standard deviation of each molding quality parameter using the following formula: Among them, X∈{V s , W d , S b }, n is the total number of data extractions in the extraction segment T, and the total number represents the total number of samples of historical data; V is defined respectively s , W d , S b The qualified thresholds are Among them, k1, k2 and k3 are multiples of the control limits, 0.23≤k1≤1.2, 0.43≤k2≤1.68, 0.11≤k3≤1.58 respectively.

4. A multi-objective optimization method for plastic part injection molding parameters according to claim 3, characterized in that: According to the constructed data extraction set {(P1,Q1),(P2,Q2),…,(P n ,Q n )}, determine the extracted heat flow change value set as {Q1,Q2,…,Q n }, define the heat flow variation coefficient as C Q , the calculation content is as follows: Calculate the mean value μ of the heat flux variation Q : Calculate the standard deviation of the heat flow change σ Q : The heat flow variation coefficient C Q Defined as the ratio of the standard deviation of the heat flow variation to the mean: Among them, C Q Expressed as a percentage, it indicates the relative degree of fluctuation of heat flow; Differential Scanning Calorimetry (DSC) is used to analyze the melt material sample and define the crystallinity X of the melt material. c , the calculation formula of crystallinity is as follows: The melting heat ΔH is obtained by the melting peak area of ​​the heat flow curve. m The cold crystallization heat ΔH is obtained by the crystallization peak area of ​​the heat flow curve. c ; is the standard melting heat of completely crystalline materials, and the corresponding crystallinity X is calculated according to different time points in the extraction section T. c The values ​​are plotted to form a trend graph of crystallinity changing over time.

5. A multi-objective optimization method for plastic part injection molding parameters according to claim 4, characterized in that: The heat flow variation coefficient C Q The calculation parameters Qi and crystallinity X c A correlation analysis is performed between them, and the correlation coefficient is defined as r. The calculation formula is as follows: Among them, X ci is the crystallinity of the ith data extraction, is the mean value of crystallinity; If r is closer to 1, it means that there is a positive correlation between the heat flow variation coefficient and the crystallinity, and the greater the fluctuation of the heat flow variation data, the higher the crystallinity; If r is closer to -1, it means that there is a negative correlation between the heat flow variation coefficient and the crystallinity. The greater the fluctuation of the heat flow variation data, the lower the crystallinity. Select the model with the highest fitting degree from the linear model and the nonlinear model for fitting; then obtain the prediction model. The fitted prediction model can calculate the C Q Predict the corresponding X c ; Based on the results of the prediction model, the injection molding process parameters including injection speed, injection pressure and cooling time are adjusted to achieve the target crystallinity value: Set the target crystallinity value to The target heat flux variation coefficient required by the prediction model is:

6. A multi-objective optimization method for plastic part injection molding parameters according to claim 5, characterized in that: The improved non-dominated sorting genetic algorithm is selected as the multi-objective optimization algorithm to calculate the current optimal injection molding process parameters. The injection molding process parameters include injection speed, injection pressure and cooling time, specifically including: 1) Initial conditions and goal setting; For optimization goals: The acoustic evaluation coefficient C approaches the flow characteristic threshold C th ; Heat flow variation coefficient C Q Approaching the target heat flux variation coefficient Crystallinity value X c Approaching the target crystallinity value Constraints: The values ​​of various parameters of molding quality must be within the qualified threshold: Volume shrinkage Warpage value Surface cavitation value 2) Generate an initial population; Generate an initial population of N individuals, each of which represents a set of injection molding process parameter combinations G i′ =(V sd ,P yl ,t cool ), where V sd , P yl ,t cool Represents injection speed, injection pressure and cooling time respectively: 3) Definition and calculation of fitness function: Define the fitness function F(G i′ ) is used to evaluate each individual G i′ The formula is as follows: Where, C is the acoustic evaluation coefficient; C Q is the coefficient of variation of heat flow, X c is the crystallinity value, w1, w2, w3 are the weight values ​​of the corresponding parameters, and Indicates that the influence of each optimization objective is equal; 4) Genetic operations: Generate the next generation of population through selection, crossover and mutation; 5) Constraint processing and penalty function: In each iteration, the molding quality parameter V s , W d , S b Evaluate; if any of the individuals does not meet The corresponding qualified threshold is: then the fitness function F(G i′ ) to impose punishment; 6) Non-dominated sorting and Pareto frontier updating; At the end of each generation, all individuals in the population are non-dominated and sorted, and the Pareto frontier is updated to obtain the optimal solution set; 7) Iteration and convergence; The maximum number of iterations is set to L1 generation, and the optimal solution is continuously approached through multiple rounds of genetic operations and non-dominated sorting; 8) Optimal solution extraction and practical application; After the iteration, the optimal solution G is extracted from the final Pareto frontier * , that is, the optimal combination of injection molding process parameters, is obtained 7. A multi-objective optimization system for plastic injection molding parameters, characterized in that: The system is used to execute the multi-objective optimization method for plastic part injection molding parameters according to any one of claims 1 to 6, comprising: Data acquisition module: used to collect acoustic signal data in the injection mold cavity in real time, including pressure fluctuations and melt flow noise in the mold cavity; at the same time, it collects heat flow change data of the melt material in the injection mold cavity during the melting and cooling process; Data extraction module: used to set the time period T immediately before the current moment as the extraction segment, and synchronously extract the acoustic signal data and heat flow change data collected in step S1 at the same time interval within the extraction segment T, and the acoustic signal data and the heat flow change data correspond one to one, and determine the extraction times set as {1,2,…,n} according to the extraction segment T and the time interval, where i∈{1,2,…,n} represents the index of the i-th data extraction, and n represents the total number of data extractions within the extraction segment T; Qualified threshold determination module: used to obtain the historical data of plastic parts molding quality under different acoustic signal data and heat flow change data, the molding quality data of plastic parts includes volume shrinkage value, warpage deformation value and surface cavitation value, and determine the qualified threshold of each parameter value in the molding quality data according to the historical data; Evaluation coefficient construction module: used for receiving acoustic signal data extracted n times in extraction section T, extracting acoustic frequency domain features of acoustic signal data using fast Fourier transform, analyzing and processing the extracted acoustic frequency domain features, and generating acoustic evaluation coefficients. The acoustic evaluation coefficients are used to determine the flow characteristics of the melt material and set the flow characteristic threshold; Variation coefficient building module: used to receive the heat flow variation data of n extractions in the extraction section T, analyze and process it, and generate the heat flow variation coefficient. The heat flow variation coefficient is used to quantitatively evaluate the crystallinity variation trend of the melt material. In the extraction section T, the melt material sample is thermally analyzed using differential scanning calorimetry DSC to obtain the crystallinity value of n extractions; Optimization target value determination module: used to perform correlation analysis between the calculation parameters of the heat flow variation coefficient and the crystallinity, select the model with the highest fitting degree from the linear model and the nonlinear model for fitting, so as to obtain a prediction model, which is used to predict the corresponding crystallinity value according to the currently calculated heat flow variation coefficient, and set the target crystallinity value and the corresponding target heat flow variation coefficient according to the prediction model; Multi-objective optimization module: It is used to take the acoustic evaluation coefficient approaching the flow characteristic threshold, the heat flow variation coefficient approaching the target heat flow variation coefficient, the crystallinity value approaching the target crystallinity value, and all molding quality parameters being below the corresponding qualified threshold as the optimization goals, select the improved non-dominated sorting genetic algorithm as the multi-objective optimization algorithm, adjust the injection molding process parameters through the iterative feedback mechanism, and calculate the current optimal injection molding process parameters. The injection molding process parameters include injection speed, injection pressure and cooling time.

Images