An injection molding process optimization method and system based on swarm intelligence algorithm

Abstract

The application provides an injection molding process optimization method and system based on a swarm intelligence algorithm, and relates to the technical field of injection molding process optimization.The method comprises the following steps: obtaining various injection molding process parameters in an injection molding process; performing orthogonal tests on the various injection molding process parameters to construct a test data set; performing signal-to-noise ratio analysis on the test data set to construct a recommended injection molding process parameter set; performing variance analysis on the recommended injection molding process parameter set to construct a key injection molding process parameter set; respectively constructing response surface prediction models between various parameters in the key injection molding process parameter set and warping amount, volume shrinkage rate and residual stress; predicting the warping amount, volume shrinkage rate and residual stress through the various response surface prediction models; constructing an injection molding process optimization model based on a swarm intelligence algorithm; and determining an optimal injection molding process parameter set through the injection molding process optimization model with the minimum warping amount, volume shrinkage rate and residual stress as the target.

Description

Technical Field

[0001] This invention relates to the field of injection molding process optimization technology, and in particular to an injection molding process optimization method and system based on swarm intelligence algorithms. Background Technology

[0002] With the widespread application of high-performance polymer-based composite materials in the automotive, electronics, and aerospace industries, injection molding has become an important molding method for plastic parts and fiber-reinforced thermoplastic composite products due to its advantages such as high efficiency, high degree of automation, and suitability for mass production of complex structures. Especially against the backdrop of continuously increasing demands for lightweight, high-precision, and high-consistency manufacturing, the impact of injection molding process parameters on the dimensional accuracy, molding stability, and overall quality of products is becoming increasingly significant.

[0003] In existing technologies, fiber-reinforced composite materials are not only affected by process conditions such as melt temperature, injection pressure, and cooling time during injection molding, but also by the coupled effects of factors such as fiber content, fiber aspect ratio, and orientation distribution. This can easily lead to significant nonlinear changes in warpage, volume shrinkage, and residual stress in the product. Although traditional injection molding process optimization methods have introduced experimental design, simulation analysis, or mathematical modeling, they often lack the ability to systematically evaluate and coordinate different quality responses when faced with multiple quality responses that coexist and are mutually restrictive. It is difficult to accurately identify key process parameters that have a substantial impact on multiple quality indicators, which can easily lead to insufficient parameter selection, one-sided evaluation results, and unstable foundation for subsequent modeling.

[0004] Furthermore, traditional injection molding process optimization methods typically rely on engineering experience, single-factor experiments, repeated trial molding under fixed parameter ranges, or empirical adjustments to local process conditions to achieve better molding results. These methods often focus on optimizing a single quality index or only conduct local analysis on a small number of process parameters. They are insufficient to fully reveal the interrelationships between multiple parameters during the injection molding of fiber-reinforced materials and to take into account the comprehensive requirements of multiple quality indicators. In real-world scenarios where the product structure is complex, the material system exhibits fiber orientation effects, and there are a large number of process parameters, many experiments are often required, the optimization process is time-consuming, and the adaptability and stability of the obtained parameter combinations under different working conditions are limited. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide an injection molding process optimization method based on swarm intelligence algorithms. This method can solve the problems of traditional injection molding process optimization methods, which often lack the ability to systematically evaluate and coordinate different quality responses when faced with multiple coexisting and mutually restrictive quality responses. These methods struggle to accurately identify key process parameters that have a substantial impact on multiple quality indicators, leading to insufficient parameter selection, biased evaluation results, and unstable foundations for subsequent modeling. Furthermore, they typically rely on engineering experience, single-factor experiments, repeated trial molding under fixed parameter ranges, or empirical adjustments to local process conditions to obtain better molding results. They often focus on optimizing a single quality indicator or only conduct local analysis on a small number of process parameters, making it difficult to fully reveal the interrelationships between multiple parameters during the injection molding of fiber-reinforced materials and to consider the comprehensive requirements of multiple quality indicators. In practical scenarios with complex product structures, fiber orientation effects in the material system, and a large number of process parameters, many experiments are often required, the optimization process is time-consuming, and the resulting parameter combinations have limited adaptability and stability under different working conditions.

[0006] A first aspect of this invention proposes a method for optimizing injection molding processes based on swarm intelligence algorithms, comprising:

[0007] S1: Obtain various injection molding process parameters during the injection molding process;

[0008] S2: Conduct orthogonal experiments on various injection molding process parameters and construct an experimental dataset;

[0009] S3: Perform signal-to-noise ratio analysis on various quality response data in the experimental dataset, determine the recommended injection molding process parameter combinations corresponding to various quality response data, and perform comprehensive compromise processing on the recommended injection molding process parameter combinations based on the synergistic relationship between various quality response data to construct a set of recommended injection molding process parameters for multi-quality response synergistic optimization.

[0010] S4: Perform variance analysis on each injection molding process parameter in the recommended injection molding process parameter set to determine the significance probability value of each injection molding process parameter for various quality response data, and screen the injection molding process parameters based on the significance probability value to construct a set of key injection molding process parameters for dimensionality reduction optimization.

[0011] S5: Using a quadratic polynomial general model, response surface prediction models are constructed for the relationship between various parameters in the key injection molding process parameter set and warpage, volume shrinkage rate and residual stress.

[0012] S6: Using various response surface prediction models, warpage, volume shrinkage rate, and residual stress are predicted;

[0013] S7: Construct an injection molding process optimization model based on swarm intelligence algorithms;

[0014] S8: With the goal of minimizing warpage, volume shrinkage, and residual stress, the optimal set of injection molding process parameters is determined through an injection molding process optimization model.

[0015] A second aspect of this invention provides an injection molding process optimization system based on swarm intelligence algorithms, comprising: a processor and a memory;

[0016] The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the injection molding process optimization method based on swarm intelligence algorithms as described in the first aspect.

[0017] A third aspect of the present invention provides a readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the injection molding process optimization method based on swarm intelligence algorithm as described in the first aspect.

[0018] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0019] In this embodiment of the invention, signal-to-noise ratio analysis is performed on various quality response data in the experimental dataset to determine the recommended injection molding process parameter combinations corresponding to each quality response data. Based on the synergistic relationship between various quality response data, a comprehensive compromise is made on the recommended injection molding process parameter combinations to construct a set of recommended injection molding process parameters for multi-quality response synergistic optimization. Variance analysis is performed on each injection molding process parameter in the recommended injection molding process parameter set to determine the significance probability value of each injection molding process parameter for various quality response data. Based on the significance probability value, the injection molding process parameters are screened to construct a set of key injection molding process parameters for dimensionality reduction optimization. When faced with multiple coexisting and mutually restrictive quality responses, this invention possesses the ability to systematically evaluate and coordinate different quality responses, accurately identify key process parameters that have a substantial impact on multiple quality indicators, and thus is less prone to parameter misalignment. Insufficient data screening, biased evaluation results, and unstable foundations for subsequent modeling all hinder this approach. By constructing an injection molding process optimization model based on a swarm intelligence algorithm, and aiming to minimize warpage, volume shrinkage, and residual stress, the optimal set of injection molding process parameters is determined. This eliminates reliance on engineering experience, single-factor experiments, repeated trial molding under fixed parameter ranges, or empirical adjustments to local process conditions to achieve better molding results. Instead of focusing on a single quality indicator or conducting only localized analysis of a few process parameters, this model fully reveals the interrelationships between multiple parameters during fiber-reinforced material injection molding and considers the comprehensive requirements of multiple quality indicators. In real-world scenarios with complex product structures, fiber orientation effects in the material system, and a large number of process parameters, it requires fewer experiments, shortens the optimization process time, and improves the adaptability and stability of the obtained parameter combinations under different operating conditions. Attached Figure Description

[0020] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0021] Figure 1 This is a schematic flowchart of an injection molding process optimization method based on swarm intelligence algorithm provided in an embodiment of the present invention.

[0022] Figure 2 This is a schematic diagram of the structure of an injection molding process optimization system based on swarm intelligence algorithm provided in an embodiment of the present invention. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] The injection molding process optimization method based on swarm intelligence algorithm provided by the present invention will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0025] Reference manual attached Figure 1 The diagram shows a flowchart of an injection molding process optimization method based on swarm intelligence algorithm provided by an embodiment of the present invention.

[0026] This invention provides a method for optimizing injection molding processes based on swarm intelligence algorithms, which may include the following steps:

[0027] S1: Obtain various injection molding process parameters during the injection molding process.

[0028] Optionally, the injection molding process parameters may include: fiber content, fiber aspect ratio, melt temperature, injection pressure, and cooling time.

[0029] It should be noted that fiber content refers to the mass percentage or volume fraction of reinforcing fibers in injection-molded materials, used to characterize the proportion of fibers added to the matrix material. Its value directly affects the mechanical properties, flow characteristics, and molding shrinkage behavior of the product. Fiber aspect ratio refers to the ratio between the length of a single fiber and its diameter, used to reflect the geometric characteristics of the fiber. A larger fiber aspect ratio is generally beneficial for improving the reinforcement effect of the material, but it also increases melt flow resistance and affects fiber orientation distribution. Melt temperature refers to the melting temperature of the polymer material before injection during the injection molding process. Its value determines the viscosity level and flowability of the melt, and is an important process parameter affecting mold filling behavior, fiber dispersion, and molding defects. Injection pressure refers to the driving force applied to the melt by the injection molding machine during the mold filling stage, used to propel the melt to overcome flow resistance and fill the mold cavity. Its magnitude directly affects the integrity of mold cavity filling, internal stress distribution, and the density of the product. Cooling time refers to the length of time that the melt remains cooled and solidified in the mold cavity after it has been filled into the mold. It is used to ensure that the product achieves sufficient structural stability and dimensional accuracy. Its setting has an important impact on the shrinkage deformation, residual stress release and final molding quality of the product.

[0030] In this embodiment of the invention, by selecting fiber content, fiber aspect ratio, melt temperature, injection pressure and cooling time as injection molding process parameters, the injection molding process is characterized from multiple aspects such as material composition, fiber morphology, thermal conditions and molding process control. This can comprehensively reflect the influence of process conditions on product warpage, volume shrinkage rate and residual stress, and provide an engineering-feasible parameter basis for subsequent statistical analysis, response surface modeling and process parameter optimization.

[0031] In one possible implementation, after S1 and before S2, step S1A is further included:

[0032] S1A: Normalizes injection molding process parameters.

[0033] In this embodiment of the invention, the injection molding process parameters are normalized after step S1 and before step S2. This can eliminate the differences in dimensions and numerical ranges of different process parameters, avoid unreasonable influence of individual parameters on subsequent statistical analysis and optimization calculations, improve the numerical stability and computational efficiency of orthogonal experimental analysis, response surface modeling and swarm intelligence optimization processes, thereby enhancing the reliability and versatility of the injection molding process parameter optimization results.

[0034] S2: Conduct orthogonal experiments on various injection molding process parameters to construct an experimental dataset.

[0035] In one possible implementation, S2 specifically includes sub-steps S201 to S205:

[0036] S201: Construct a factor level table based on the experimental levels of various injection molding process parameters.

[0037] Specifically, based on the injection molding process parameters (such as fiber content and melt temperature) and their physically feasible ranges identified in step S1, and considering material properties, equipment capabilities, and engineering experience, several representative discrete values ​​are scientifically selected as experimental levels for each parameter. For example, for the fiber content parameter, if its range is between 10% and 30%, three levels of 15%, 20%, and 25% can be selected evenly. All parameters and their corresponding level values ​​are organized into a structured table, namely the "factor level table." This table clarifies the adjustable state of each variable in the experimental design and serves as the basis for subsequently selecting an orthogonal experimental table.

[0038] S202: Select the target orthogonal array from the standard orthogonal array library based on the factor level table.

[0039] Specifically, based on the number of parameters and the number of levels set for each parameter in the "Factor Level Table" constructed in S201, an orthogonal table that can fully accommodate or efficiently cover all combinations of factors and levels is selected from a standard orthogonal table set (such as an orthogonal table library containing L4, L9, L16, L18, etc.) as the "target orthogonal table". For example, when there are 5 factors, all of which have 3 levels, the L18 (2^1×3^7) table can be selected, and its 7 columns with 3 levels can be used to arrange the experiments of the 5 factors, so as to examine the influence of each parameter evenly in 18 experiments.

[0040] S203: Generate an orthogonal experimental scheme matrix based on the target orthogonal array.

[0041] Specifically, the specific level values ​​of various parameters in the "Factor Level Table" of S201 are filled into the corresponding columns of the "Objective Orthogonal Array" selected in S202 according to the experimental design rules. Each row of the orthogonal array represents a unique combination of process parameters. After filling, a clear and specific "Orthogonal Experimental Scheme Matrix" is formed. This matrix is ​​an n-row, k-column numerical table, where n represents the total number of experiments and k represents the number of process parameters. Each row of the matrix defines all the input conditions for the next experiment (simulation or actual production) to be performed.

[0042] S204: Execute each set of experiments in the orthogonal experimental design matrix and record the various quality response data output by each set of experiments.

[0043] Optionally, the quality response data may specifically include: warpage, volume shrinkage, and residual stress.

[0044] Specifically, following the "orthogonal experimental scheme matrix" generated in S203, each row of parameter combinations is implemented sequentially. Simulations are typically performed using computer-aided engineering (CAE) injection molding simulation software (such as Moldflow), or physical experiments are conducted on a suitable injection molding machine. After each experiment, multiple quality index results corresponding to that set of parameters are precisely measured and recorded, such as warpage (unit: mm), volume shrinkage rate (unit: %), and residual stress (unit: MPa), thereby obtaining multiple sets of "quality response data" equal to the number of rows in the experimental scheme matrix.

[0045] S205: Match various injection molding process parameters and various quality response data to construct an experimental dataset.

[0046] Specifically, the "orthogonal experimental scheme matrix" (containing all input parameters) in S203 and the "quality response data" (containing all output indicators) recorded in S204 are matched one-to-one and horizontally merged according to the experimental sequence. This results in a complete structured data table, the "experimental dataset." Each row of this dataset fully records the input (process parameter values) and output (quality indicator values) of one experiment, providing a direct and standardized sample data foundation for subsequent signal-to-noise ratio analysis, variance analysis, and response surface modeling.

[0047] In this embodiment of the invention, S201 scientifically selects representative experimental levels for each injection molding process parameter and constructs a factor level table, clarifying the adjustable range and discrete state of each parameter, thus providing a foundation for standardized experimental design. S202 selects a target orthogonal array from the standard orthogonal array library based on the factor level table, ensuring that each injection molding process parameter and its experimental level appear in a balanced and orthogonal combination, thereby effectively covering the parameter combination space while reducing the number of experiments. S203 generates an orthogonal experimental scheme matrix based on the target orthogonal array, transforming the experimental design results into specific and executable combinations of process parameters, facilitating subsequent simulation or actual experiment implementation. S204 executes each group of experiments sequentially according to the orthogonal experimental scheme matrix and obtains quality response data such as warpage, volume shrinkage rate, and residual stress, achieving systematic acquisition of molding quality characteristics for different parameter combinations. S205 matches and integrates each group of injection molding process parameters with the corresponding quality response data to form a structured experimental dataset, thus providing a reliable and standardized data foundation for subsequent signal-to-noise ratio analysis, variance analysis, and response surface prediction model construction.

[0048] S3: Perform signal-to-noise ratio analysis on various quality response data in the experimental dataset, determine the recommended injection molding process parameter combinations corresponding to various quality response data, and perform comprehensive compromise processing on the recommended injection molding process parameter combinations based on the synergistic relationship between various quality response data to construct a set of recommended injection molding process parameters for multi-quality response synergistic optimization.

[0049] In one possible implementation, S3 specifically includes sub-steps S301 to S304:

[0050] S301: Calculate the signal-to-noise ratio of various quality response data in the test dataset.

[0051] Specifically, based on the experimental dataset constructed in step S2, the signal-to-noise ratio (SNR) of warpage, volume shrinkage, and residual stress was calculated for the quality response data corresponding to each group of injection molding process parameter combinations. First, according to the optimization target type of each quality response index, the corresponding SNR calculation mode was selected. When the quality response index was of the "smaller is better" type, the small characteristic SNR calculation formula was used. The quality response data of each group of experiments were substituted into the SNR calculation formula to obtain the corresponding SNR value. Second, the SNR values ​​of all experimental groups under each quality response index were calculated and recorded one by one, forming a SNR data set corresponding to the experimental dataset. The SNR value is used to comprehensively characterize the performance level and stability of each injection molding process parameter combination under the corresponding quality response index, providing a unified quantitative evaluation basis for the subsequent optimization and combination of different process parameter levels.

[0052] Optionally, the formula for calculating the signal-to-noise ratio is as follows:

[0053]

[0054] in, The signal-to-noise ratio is represented by log(), the logarithmic function is represented by G, and the number of trials is represented by G. This represents the quality response data of the g-th test.

[0055] S302: Group the injection molding process parameters according to their different test levels, and calculate the average signal-to-noise ratio of all groups of tests under different test levels for each injection molding process parameter based on the signal-to-noise ratio value.

[0056] Specifically, based on the signal-to-noise ratio (SNR) values ​​corresponding to each quality response data obtained in step S301, all test data are grouped according to the type of injection molding process parameter and its corresponding test level. For any injection molding process parameter, all its preset test levels are traversed, and all test groups in the test dataset with the same test level for that injection molding process parameter are grouped into the same group. Within each group, the SNR value of the corresponding test group under the target quality response index is extracted, the SNR values ​​are summed and divided by the number of test groups in that group to obtain the average SNR value of the injection molding process parameter at that test level. The above process is repeated to obtain the set of average SNR values ​​for each injection molding process parameter at each test level, thereby forming an evaluation result used to characterize the degree of influence of different process parameter levels on the quality response.

[0057] S303: Based on the average signal-to-noise ratio, determine the recommended injection molding process parameter combinations corresponding to various quality response data.

[0058] Specifically, based on the set of average signal-to-noise ratios (SNRs) for each injection molding process parameter at different test levels obtained in step S302, parameter optimization is performed independently for each type of quality response data. For any quality response index, each injection molding process parameter is iterated through, and the test level with the highest average SNR among all test levels for that injection molding process parameter is selected as the optimal level for that injection molding process parameter under that quality response index. After selecting the optimal levels for all injection molding process parameters, the optimal levels corresponding to each injection molding process parameter are combined to form a complete set of injection molding process parameter combinations. The above process is repeated to obtain recommended injection molding process parameter combinations for warpage, volume shrinkage, and residual stress, thereby constructing a set of recommended injection molding process parameter combinations corresponding to each quality response data.

[0059] S304: Based on the collaborative relationship between various quality response data, a comprehensive compromise is made on the recommended injection molding process parameter combinations corresponding to various quality response data to construct a set of recommended injection molding process parameters for multi-quality response collaborative optimization.

[0060] Specifically, based on the multiple recommended injection molding process parameter combinations obtained in step S303 for each quality response data, the synergistic relationship between different quality response data is first analyzed. The importance of warpage, volume shrinkage rate, and residual stress in process optimization is uniformly quantified, and a synergistic evaluation criterion between multiple quality responses is constructed. Secondly, based on each recommended injection molding process parameter combination, the recommended levels of each injection molding process parameter under different quality responses are compared one by one to identify the consistency level and conflicts of each parameter under different quality responses. For injection molding process parameters with consistent values ​​in multiple quality responses, their corresponding experimental levels are directly retained. For injection molding process parameters with different values ​​in different quality responses, each candidate experimental level is comprehensively evaluated according to the synergistic evaluation criterion, and the experimental level that can take into account the performance of multiple quality responses is selected as the final value of the injection molding process parameter. After determining the unified values ​​of all injection molding process parameters, the final values ​​of each injection molding process parameter are combined to form a set of recommended injection molding process parameters that meet the requirements of synergistic optimization of multiple quality responses.

[0061] In this embodiment of the invention, step S301 calculates the signal-to-noise ratio (SNR) of each quality response data, transforming the original quality response data into a unified evaluation index that simultaneously reflects performance level and stability. This effectively reduces the interference of random fluctuations on the evaluation results and improves the reliability of parameter evaluation. Step S302 groups the SNR values ​​according to the injection molding process parameters and their experimental levels and calculates the average value, achieving a systematic quantification of the impact of different parameter levels on the quality response, thus providing a more comprehensive and stable characterization of the role of each parameter level. Step S303 independently determines recommended injection molding process parameter combinations for each quality response data, avoiding evaluation bias caused by mutual interference between different quality responses and achieving independent optimal analysis of each quality index. Furthermore, step S304, based on the synergistic relationship between multiple quality responses, performs consistency analysis and conflict resolution on each recommended injection molding process parameter combination, and makes comprehensive compromise decisions on parameters with differences, constructing a unified set of recommended injection molding process parameters. This coordinates multiple quality responses at the parameter level, achieving conflict resolution and consistency optimization at the parameter level. Therefore, this invention, through a multi-level screening and convergence mechanism of "signal-to-noise ratio evaluation - parameter level statistics - construction of optimal combination of single index - multi-index synergistic compromise", can not only effectively reduce the dimension of parameter space and improve the efficiency of subsequent modeling and optimization, but also improve the stability and reliability of multi-index synergistic optimization while ensuring the performance of each quality response, thereby significantly enhancing the engineering applicability of injection molding process optimization methods under complex working conditions.

[0062] S4: Perform variance analysis on each injection molding process parameter in the recommended injection molding process parameter set to determine the significance probability value of each injection molding process parameter for various quality response data, and screen the injection molding process parameters based on the significance probability value to construct a set of key injection molding process parameters for dimensionality reduction optimization.

[0063] In one possible implementation, S4 specifically includes sub-steps S401 to S404:

[0064] S401: Construct an analysis of variance model using various quality response data as dependent variables and each injection molding process parameter in the recommended injection molding process parameter set as independent variables.

[0065] Specifically, based on the experimental dataset constructed in step S2 and combined with the recommended injection molding process parameter set obtained in step S3, corresponding analysis of variance (ANOVA) models are established for each quality response data. For any quality response index, the quality response data is used as the dependent variable, and each injection molding process parameter in the recommended injection molding process parameter set is used as the independent variable. The experimental data are then grouped according to the different experimental levels of each injection molding process parameter in the experimental dataset. Based on this, the between-group and within-group variances of each injection molding process parameter on the quality response data at different experimental levels are calculated, and a multi-factor ANOVA model is constructed to characterize the influence of each injection molding process parameter on the quality response data. The above process is repeated to establish ANOVA models for warpage, volume shrinkage rate, and residual stress, thereby forming a set of ANOVA models for subsequent significance analysis.

[0066] Optionally, the calculation formula for the analysis of variance model is as follows:

[0067]

[0068] in, This represents quality response data (e.g., warpage, volume shrinkage, and residual stress). M represents the overall average value, M represents the number of injection molding process parameters, and A represents the total average value. m This represents the main effect of the m-th injection molding process parameter (i.e., the quality response data of the m-th injection molding process parameter at its different experimental levels). Contribution bias), The term represents the residual. Those skilled in the art can set the size of the residual according to actual needs, but this invention does not limit it.

[0069] S402: Using an analysis of variance model, calculate the significance probability value of each injection molding process parameter in the recommended injection molding process parameter set for various quality response data.

[0070] Specifically, the variance analysis models corresponding to each quality response constructed in step S401 are solved. For each quality response data point, each injection molding process parameter in the recommended injection molding process parameter set is iterated. Based on the changes in quality response data caused by the injection molding process parameter at different experimental levels, the corresponding between-group mean square and within-group mean square of error are calculated, and the F-statistic of the injection molding process parameter is obtained by the ratio of the two. On this basis, the significance probability value (P-value) of the injection molding process parameter on the current quality response data is calculated by combining the F-statistic and its corresponding degrees of freedom with the F-distribution function. The P-value is used to characterize the probability of observing the current or more extreme F-statistic under the premise that the null hypothesis "the injection molding process parameter has no significant effect on the current quality response" holds. The smaller the P-value, the more significant the influence of the injection molding process parameter on the quality response. The above calculation process is repeated to obtain the significance probability values ​​of each injection molding process parameter under the variance analysis models corresponding to warpage, volume shrinkage rate, and residual stress, thereby forming a set of significance probability values ​​for different quality responses, which are used for subsequent screening and determination of key injection molding process parameters.

[0071] S403: Compare the significance probability value with the preset significance probability value, and select a variety of key injection molding process parameters from the recommended injection molding process parameter set.

[0072] Specifically, based on the significance probability values ​​of each injection molding process parameter obtained in step S402 under different quality response data, a threshold for significance determination is preset as a preset significance probability value. For any injection molding process parameter, its significance probability value corresponding to each quality response data (including warpage, volume shrinkage rate, and residual stress) is extracted, and the significance probability value is compared with the preset significance probability value one by one. When the significance probability value of the injection molding process parameter under at least one quality response data is less than the preset significance probability value, it is determined that the injection molding process parameter has a significant impact on the corresponding quality response, and the injection molding process parameter is marked as a key injection molding process parameter. If the significance probability value of the injection molding process parameter under all quality response data is not less than the preset significance probability value, it is determined that its impact on each quality response is insignificant and is removed. The above determination process is repeated for all injection molding process parameters in the recommended injection molding process parameter set, and finally a set of key injection molding process parameters that have a significant impact on at least one quality response is obtained. Those skilled in the art can set the size of the preset significance probability value according to actual needs, and this invention does not limit it.

[0073] S404: Combine various key injection molding process parameters to construct a set of key injection molding process parameters for dimensionality reduction optimization.

[0074] Specifically, based on the key injection molding process parameters selected in step S403, subsets of parameters deemed significant under different quality response data are extracted, and these subsets are then summarized and integrated. During integration, duplicate injection molding process parameters are deduplicated, retaining only those parameters that are significant under any given quality response data. Based on this, all deduplicated key injection molding process parameters are uniformly combined according to a preset parameter order to construct a complete set of key injection molding process parameters. This set of key injection molding process parameters contains only those that significantly affect at least one quality response, and is used to replace all original process parameters in subsequent modeling and optimization calculations, thereby achieving effective dimensionality reduction of the parameter space and improving the efficiency and stability of optimization calculations.

[0075] In this embodiment of the invention, step S401 constructs a variance analysis model with each quality response data as the dependent variable and the recommended injection molding process parameters as the independent variables, thereby quantitatively characterizing the influence of each injection molding process parameter and unifying the modeling and expression of the relationship between different process parameters and the quality response. Step S402 calculates the significance probability values ​​of each injection molding process parameter under different quality responses based on the variance analysis model, transforming the parameter influence from a qualitative judgment into a quantifiable statistical indicator, thus improving the objectivity and accuracy of parameter evaluation. Further, step S403 systematically compares the significance probability values ​​with preset thresholds, uniformly screening injection molding process parameters across multiple quality response dimensions, avoiding screening biases caused by relying solely on a single indicator or empirical judgment, and achieving accurate identification of key parameters. Based on this, step S404 summarizes, integrates, and deduplicates the key injection molding process parameters screened under each quality response, constructing a unified set of key injection molding process parameters, thus completing the parameter space convergence process driven by multiple quality responses. Therefore, this invention employs a two-stage screening and dimensionality reduction mechanism of "variance modeling - significance quantification - multi-response screening - parameter integration" to ensure that only key parameters that have a significant impact on at least one quality response are retained for subsequent modeling and optimization. This not only effectively reduces the dimensionality of the parameter space and minimizes the interference of redundant variables on model accuracy, but also significantly improves the stability and efficiency of optimization calculations, thereby enhancing the applicability and reliability of the injection molding process optimization method in complex multi-parameter coupling scenarios.

[0076] S5: Using a quadratic polynomial general model, response surface prediction models are constructed for the relationship between various parameters in the key injection molding process parameter set and warpage, volume shrinkage rate, and residual stress.

[0077] Specifically, based on the set of key injection molding process parameters constructed in step S4, each injection molding process parameter is used as the independent variable for modeling, and warpage, volume shrinkage rate, and residual stress are used as the target response variables. A quadratic response surface methodology is employed to establish predictive function models for the three types of quality indicators. These predictive models are used to fit the nonlinear mapping relationship between injection molding process parameters and quality response, constructing a mathematical expression for the parameter-performance relationship. This allows for rapid estimation of quality results in subsequent optimization processes without the need for re-performing physical simulations or experiments. The model possesses good interpretability and differentiability, making it suitable as a basis for fitness calculations using optimization methods such as swarm intelligence algorithms.

[0078] It should be noted that the specific calculation formula for the general model form of the quadratic polynomial is as follows:

[0079]

[0080] Where Y represents quality response data (e.g., warpage, volume shrinkage, and residual stress), a0 represents the constant term coefficient, Z represents the number of key injection molding process parameters, and a z Indicates the injection molding process parameter x of the z-th bond. z The corresponding coefficient of the linear term, a zz Indicates the injection molding process parameter x of the z-th bond. z The corresponding quadratic coefficient, x represents the injection molding process parameter for the z-th bond. z and the Key Injection Molding Process Parameters The interaction term coefficients can be set by those skilled in the art according to actual needs, and the constant term coefficients can be set with respect to the z-th bond injection molding process parameter x. z The corresponding linear coefficient and the injection molding process parameter x of the z-th bond z The corresponding quadratic coefficients and the injection molding process parameter x of the z-th bond z and the Key Injection Molding Process Parameters The magnitude of the interaction term coefficients is not limited in this invention.

[0081] In this embodiment of the invention, by using a quadratic polynomial general model to construct a response surface prediction model between key injection molding process parameters and warpage, volume shrinkage rate and residual stress, it is possible to effectively characterize the nonlinear and interactive influence relationship of process parameters on multiple quality indicators. While ensuring the stability and interpretability of the model, it can achieve rapid and continuous prediction of quality response, thereby providing a reliable and efficient mathematical basis for subsequent optimization calculations.

[0082] S6: Using various response surface prediction models, warpage, volume shrinkage rate, and residual stress are predicted.

[0083] S7: Construct an injection molding process optimization model based on swarm intelligence algorithms.

[0084] Optionally, the swarm intelligence algorithm is specifically the particle swarm optimization algorithm.

[0085] In this embodiment of the invention, by constructing an injection molding process optimization model based on a swarm intelligence algorithm, an efficient global search of the process parameter space can be performed under injection molding conditions with multiple process parameters and strong nonlinear coupling. Combined with a response surface prediction model, the results of multiple quality indicators can be quickly evaluated, thereby reducing computation and experimental costs while improving the reliability and engineering applicability of obtaining the optimal set of injection molding process parameters.

[0086] S8: With the goal of minimizing warpage, volume shrinkage, and residual stress, the optimal set of injection molding process parameters is determined through an injection molding process optimization model.

[0087] In one possible implementation, S8 specifically includes sub-steps S801 and S802:

[0088] S801: Construct an objective function to minimize warpage, volume shrinkage, and residual stress:

[0089]

[0090] Where F(X) represents the objective function value when the set of injection molding process parameters is not X, and ω1 represents the weighting coefficient of warpage. This represents the warpage when the injection molding process parameter set X is not specified, and ω2 represents the weighting coefficient for volume shrinkage. This represents the volume shrinkage rate when the set of injection molding process parameters X is not specified, and ω3 represents the weighting coefficient for residual stress. The residual stress represents the stress when the set of injection molding process parameters is not X. Those skilled in the art can set the weighting coefficients of warpage, volume shrinkage, and residual stress according to actual needs. This invention does not limit these settings.

[0091] S802: Using the reciprocal of the objective function as the fitness function, the optimal set of injection molding process parameters is determined through the injection molding process optimization model.

[0092] In this embodiment of the invention, S801 constructs a unified objective function by weighting warpage, volume shrinkage rate, and residual stress, thereby achieving synergistic optimization of multiple quality indicators. S802 uses the reciprocal of the objective function as the fitness function, ensuring that the optimization search direction aligns with the quality improvement direction, thus efficiently determining the set of injection molding process parameters with the optimal overall quality performance.

[0093] In one possible implementation, S802 specifically includes sub-steps S8021 to S8025:

[0094] S8021: The particle swarm is initialized through an adaptive piecewise nonlinear chaotic mapping algorithm. The particle swarm consists of multiple particles, each representing a feasible set of injection molding process parameters, including fiber content, fiber aspect ratio, melt temperature, injection pressure, and cooling time.

[0095] Optionally, the calculation formula for the adaptive piecewise nonlinear chaotic mapping algorithm is as follows:

[0096]

[0097] Where, x n+1 The value of the chaotic variable is represented by μ in the (n+1)th iteration, μ represents the adaptive modulation factor, α represents the amplitude control coefficient of the exponential nonlinear mapping, and x represents the value of the chaotic variable. n The value of the chaotic variable is represented in the nth iteration, e represents the base of the natural logarithm, β represents the exponential disturbance intensity adjustment coefficient, and mod represents the modulo operation. Those skilled in the art can set the magnitude of the amplitude control coefficient and the exponential disturbance intensity adjustment coefficient of the exponential nonlinear mapping according to actual needs, but this invention does not limit them.

[0098] It should be noted that traditional particle swarm optimization (PSO) algorithms typically generate initial particle positions randomly during initialization, which can easily lead to uneven particle distribution and local clustering in the parameter space. This reduces the algorithm's global search capability and increases the risk of getting trapped in local optima. To address these issues, a chaotic mapping algorithm is introduced for particle swarm initialization. This leverages the ergodicity and initial value sensitivity of chaotic systems to improve the diversity and coverage of the initial particle distribution. However, traditional chaotic mapping algorithms such as Logistic, Tent, and Sine mappings still suffer from boundary clustering, periodic collapse, and strong dimensionality correlation in high-dimensional parameter spaces, making them unsuitable for optimizing complex process parameters. Therefore, this application employs an adaptive piecewise nonlinear chaotic mapping algorithm. By segmenting the chaotic variables and introducing exponential nonlinear perturbations and state-dependent adaptive modulation mechanisms, the ergodicity and nonperiodicity of the chaotic sequence are effectively enhanced, resulting in a more uniform particle distribution in the search space. This improves the global search performance and optimization stability of the PSO algorithm.

[0099] Furthermore, the specific formula for calculating the adaptive modulation factor is as follows:

[0100]

[0101] Where ε represents the modulation perturbation amplitude coefficient, sin() represents the sine function, and π represents pi. Those skilled in the art can set the magnitude of the modulation perturbation amplitude coefficient according to actual needs, but this invention does not limit it.

[0102] S8022: The fitness value of each particle is calculated by using the reciprocal of the objective function as the fitness function, and the particles are clustered by the K-means clustering algorithm to obtain multiple subpopulations.

[0103] It should be noted that the K-means clustering algorithm is an unsupervised clustering method based on distance metrics. It divides samples into multiple categories by pre-setting the number of clusters and uses the distance from the sample to each category center as the basis for division. By iteratively updating the cluster centers and sample affiliation, it maximizes the similarity of samples within each category and the difference between categories, thereby achieving effective grouping of the sample set.

[0104] S8023: Based on the individual optimal position of each particle and the global optimal position of each sub-swarm, an adaptive inertial weight and an adaptive learning factor are introduced to update the velocity and position of each particle, and update the global optimal position of the particle swarm.

[0105] Optionally, the velocity of each particle can be updated according to the following formula:

[0106]

[0107] Among them, v i (t+1) represents the velocity of the i-th particle in the (t+1)-th iteration, ω(t) represents the adaptive inertia weight in the t-th iteration, and v i (t) represents the velocity of the i-th particle in the t-th iteration, c1 represents the individual adaptive learning factor, r1 represents the first random number in the interval [0,1], and p i (t) represents the optimal position of the i-th particle in the t-th iteration, x i (t) represents the position of the i-th particle in the t-th iteration, min represents taking the minimum value, c2 represents the population adaptive learning factor, r2 represents the second random number in the interval [0,1], and q j (t) represents the globally optimal position of the j-th subpopulation at the t-th iteration.

[0108] Optionally, the positions of each particle can be updated according to the following formula:

[0109]

[0110] Where, x i (t+1) represents the position of the i-th particle in the (t+1)-th iteration.

[0111] It should be noted that the specific formulas for calculating the velocity update and position update in the traditional particle swarm optimization algorithm are as follows:

[0112]

[0113]

[0114] Where C1 represents the individual learning factor, C2 represents the group learning factor, and p g (t) represents the global optimal position at the t-th iteration.

[0115] It should be noted that traditional particle swarm optimization (PSO) algorithms typically employ fixed inertia weights, individual learning factors, and swarm learning factors when updating particle velocity and position. This keeps the particle's dependence on historical velocity, individual optimal position, and global optimal position constant throughout the iteration process. This can easily lead to insufficient global exploration capability in the early stages of the search or slower convergence speed in the later stages, increasing the risk of getting trapped in local optima. To overcome these problems, this application introduces adaptive inertia weights and adaptive learning factors during particle velocity updates. This allows the weights of particles on historical search information, individual experience, and global experience of the subpopulation to be dynamically adjusted during the iteration process. This enhances the global search capability of the particle swarm in the early stages and gradually improves the local convergence capability of particles in the later stages, achieving a balance between global search performance and local optimization accuracy. This improves the solution stability and optimization effect of the PSO algorithm in complex parameter spaces.

[0116] Furthermore, the specific formula for calculating the adaptive inertia weight is as follows:

[0117]

[0118] Where t represents the current iteration number, The inertial weight segmentation control coefficient is represented by T, which represents the maximum number of iterations. Those skilled in the art can set the size of the inertial weight segmentation control coefficient according to actual needs, but this invention does not limit it.

[0119] Furthermore, the specific formula for calculating the adaptive learning factor includes:

[0120]

[0121]

[0122] Among them, c max c represents the maximum value of the learning factor. min This represents the minimum value of the learning factor.

[0123] S8024: Determine if the global optimal position of the particle swarm remains unchanged for a preset number of consecutive iterations. If yes, introduce a particle re-initialization mechanism to re-initialize some particles in each sub-swarm, and return to S8023. Otherwise, proceed to S8025.

[0124] Optionally, the calculation formula for the particle re-initialization mechanism is as follows:

[0125]

[0126] Where, x new q represents the new position of the particle after the re-initialization operation. j Let δ represent the global optimal position of the j-th subpopulation, and let l represent the directional displacement control coefficient. j Let represent the local optimal solution of the j-th subpopulation, γ represent the random perturbation intensity coefficient, and R represent the random vector.

[0127] It should be noted that during the iterative process of the particle swarm optimization algorithm, when the global optimal position of the particle swarm remains unchanged for multiple consecutive iterations, it indicates that the algorithm may be trapped in a local optimum, and the overall search activity of the particle swarm decreases significantly. To address this problem, traditional improved particle swarm algorithms typically intervene with the particles using methods such as Cauchy perturbation, Gaussian perturbation, or completely random reinitialization. However, these methods often have drawbacks such as strong randomness in the perturbation direction, insufficient utilization of existing search information, or disruption of the particle swarm structure stability, which can easily lead to reduced search efficiency. Therefore, this application, when detecting a swarm search stagnation, performs a reinitialization operation only on a portion of the particles in each subpopulation. Based on the global optimal position of the subpopulation, a controlled displacement along the direction of the subpopulation's local optimum is introduced, and a random perturbation term is superimposed to generate new particle positions. This effectively restores the diversity of the particle swarm while preserving existing excellent search information and subpopulation structure, enhances the algorithm's ability to escape local optima, and improves overall search stability and optimization performance.

[0128] S8025: Determine if the current iteration count has reached the maximum iteration count. If yes, determine the set of injection molding process parameters represented by the particles corresponding to the current global optimal solution as the optimal injection molding process parameter set. Otherwise, return to continue iterating.

[0129] In this embodiment of the invention, S8021 introduces an adaptive piecewise nonlinear chaotic mapping algorithm to initialize the particle swarm, making the particles more evenly distributed in the injection molding process parameter search space, enhancing the diversity and global coverage of the initial solutions, thereby reducing the risk of getting trapped in local optima during the optimization process. S8022 uses the reciprocal of the objective function as the fitness function and combines it with a K-means clustering algorithm to cluster the particles, forming multiple subpopulations, enabling parallel search of the particle swarm in the parameter space, balancing search diversity with the preservation of excellent solutions. S8023 introduces adaptive inertia weights and adaptive learning factors during particle velocity and position updates, allowing the algorithm to focus on global exploration in the early stages of the search and on local fine-tuning in the later stages, thereby improving convergence speed and optimization stability. S8024, when particle swarm search stagnation is detected, a particle re-initialization mechanism based on the global optimal position of the subpopulation is introduced, restoring particle swarm diversity while preserving existing effective search information, enhancing the ability to escape local optima. S8025 sets the maximum number of iterations as the termination condition and outputs the set of injection molding process parameters corresponding to the current globally optimal particle as the final result, ensuring that the optimization process is controllable and the result is clear, thereby reliably determining the optimal set of injection molding process parameters that meets the comprehensive requirements of multiple quality indicators.

[0130] Reference manual attached Figure 2 The diagram shows a structural schematic of an injection molding process optimization system based on a swarm intelligence algorithm provided in an embodiment of the present invention.

[0131] This invention provides an injection molding process optimization system 20 based on swarm intelligence algorithms, comprising: a processor 201 and a memory 202;

[0132] The memory 202 stores programs or instructions that can run on the processor 201. When the program or instructions are executed by the processor 201, they implement the steps of the above-mentioned injection molding process optimization method based on swarm intelligence algorithm and achieve the same technical effect. To avoid repetition, the present invention will not elaborate further.

[0133] It should be understood that the processor 201 in this embodiment of the invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0134] It should also be understood that the memory 202 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0135] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0136] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0137] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0138] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0139] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0140] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0141] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0142] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0143] This invention provides a readable storage medium comprising: storing a program or instructions on the readable storage medium, wherein when the program or instructions are executed by a processor, the program or instructions implement the steps of the above-described injection molding process optimization method based on swarm intelligence algorithm, and can achieve the same technical effect. To avoid repetition, this invention will not elaborate further.

[0144] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing injection molding process based on swarm intelligence algorithm, characterized in that, include: S1: Obtain various injection molding process parameters during the injection molding process; S2: Conduct orthogonal experiments on various injection molding process parameters to construct an experimental dataset; S3: Perform signal-to-noise ratio analysis on various quality response data in the test dataset, determine the recommended injection molding process parameter combinations corresponding to various quality response data, and perform comprehensive compromise processing on the recommended injection molding process parameter combinations based on the synergistic relationship between various quality response data to construct a set of recommended injection molding process parameters for multi-quality response synergistic optimization. S4: Perform variance analysis on each injection molding process parameter in the recommended injection molding process parameter set to determine the significance probability value of each injection molding process parameter to various quality response data, and screen the injection molding process parameters based on the significance probability value to construct a key injection molding process parameter set for dimensionality reduction optimization. S5: Using a quadratic polynomial general model, construct response surface prediction models for various parameters in the set of key injection molding process parameters and warpage, volume shrinkage rate and residual stress respectively. S6: The warpage, volume shrinkage rate, and residual stress are predicted using various response surface prediction models. S7: Construct an injection molding process optimization model based on swarm intelligence algorithms; S8: With the goal of minimizing the warpage, the volume shrinkage rate, and the residual stress, the optimal set of injection molding process parameters is determined using the injection molding process optimization model.

2. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, The injection molding process parameters specifically include: fiber content, fiber aspect ratio, melt temperature, injection pressure, and cooling time.

3. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, After S1 and before S2, it also includes: S1A: Normalize the injection molding process parameters.

4. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, S2 specifically includes: S201: Construct a factor level table based on the experimental levels of various injection molding process parameters; S202: Select a target orthogonal array from the standard orthogonal array library based on the aforementioned factor level table; S203: Generate an orthogonal experimental scheme matrix based on the target orthogonal array; S204: Execute each group of experiments in the orthogonal experimental scheme matrix and record the various quality response data output by each group of experiments; S205: Match the various injection molding process parameters and the various quality response data to construct the test dataset.

5. The injection molding process optimization method based on swarm intelligence algorithm according to claim 4, characterized in that, The quality response data specifically includes: the warpage, the volume shrinkage rate, and the residual stress.

6. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, S3 specifically includes: S301: Calculate the signal-to-noise ratio of various quality response data in the test dataset; S302: Group the various injection molding process parameters according to their different test levels, and calculate the average signal-to-noise ratio of all groups of tests under different test levels for the various injection molding process parameters based on the signal-to-noise ratio value. S303: Based on the average signal-to-noise ratio, determine the recommended injection molding process parameter combinations corresponding to each of the quality response data; S304: Based on the collaborative relationship between various quality response data, a comprehensive compromise is made on the recommended injection molding process parameter combinations corresponding to various quality response data to construct a set of recommended injection molding process parameters for multi-quality response collaborative optimization.

7. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, S4 specifically includes: S401: Using various quality response data as dependent variables and each injection molding process parameter in the recommended injection molding process parameter set as independent variables, construct an analysis of variance model; S402: Using the variance analysis model, calculate the significance probability value of each injection molding process parameter in the recommended injection molding process parameter set for various quality response data; S403: Compare the significance probability value with the preset significance probability value, and select a variety of key injection molding process parameters from the recommended injection molding process parameter set; S404: Combine various key injection molding process parameters to construct a set of key injection molding process parameters for dimensionality reduction optimization.

8. The injection molding process optimization method based on swarm intelligence algorithm according to claim 1, characterized in that, The swarm intelligence algorithm is specifically: particle swarm optimization algorithm; S8 specifically includes: S801: Construct an objective function with the goal of minimizing the warping, the volume shrinkage rate, and the residual stress; S802: Using the reciprocal of the objective function as the fitness function, the optimal set of injection molding process parameters is determined through the injection molding process optimization model.

9. The injection molding process optimization method based on swarm intelligence algorithm according to claim 8, characterized in that, Specifically, S802 includes: S8021: Initialize a particle swarm using an adaptive piecewise nonlinear chaotic mapping algorithm. The particle swarm includes multiple particles, each of which represents a feasible set of injection molding process parameters, including fiber content, fiber aspect ratio, melt temperature, injection pressure, and cooling time. S8022: Using the reciprocal of the objective function as the fitness function, calculate the fitness value of each particle, and perform clustering operation on each particle using the K-means clustering algorithm to obtain multiple subpopulations; S8023: Based on the individual optimal position of each particle and the global optimal position of each sub-population, an adaptive inertial weight and an adaptive learning factor are introduced to update the velocity and position of each particle, and update the global optimal position of the particle swarm. S8024: Determine whether the global optimal position of the particle swarm remains unchanged for a preset number of consecutive times; if yes, introduce a particle re-initialization mechanism to re-initialize some particles in each sub-swarm and return to S8023; otherwise, proceed to S8025. S8025: Determine whether the current iteration count has reached the maximum iteration count; if so, determine the set of injection molding process parameters represented by the particles corresponding to the current global optimal solution as the optimal injection molding process parameter set; otherwise, return to continue iterating.

10. A swarm intelligence algorithm-based injection molding process optimization system, characterized in that, include: Processor and memory; The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the injection molding process optimization method based on swarm intelligence algorithm as described in any one of claims 1 to 9.

Images