Injection molding process parameter optimization method, device and system

Abstract

The invention discloses an injection molding process parameter optimization method, device and system, and relates to the technical field of injection molding machining.The method comprises the steps that current injection molding production parameters of an injection molding part are collected; performing defect prediction based on the current injection molding production parameters through the trained quality detection model, and outputting a quality detection result; and when the quality detection result is analyzed to determine that the quality of the injection-molded part is abnormal, outputting optimized injection-molded production parameters through a emperor penguin breeding optimization algorithm based on the parameter set of the current injection-molded production parameters. Based on the scheme, efficient defect detection is carried out through the quality detection model, efficient optimization of the injection molding production parameters is completed based on the powerful search capability in combination with the emperor penguin breeding optimization algorithm, and the optimization reliability of the injection molding process parameters is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of injection molding processing, and particularly relates to a method, device and system for optimizing injection molding process parameters. Background Art

[0002] Industrial plastic products are widely used due to their advantages such as light weight. Plastic injection molding based on an injection molding machine is one of the main molding methods for industrial plastic products, and the injection molding process parameters have a key impact on the quality of the finished product.

[0003] In the traditional injection molding process, the selection and configuration of injection molding process parameters mainly rely on manual experience, with low efficiency and difficulty in meeting the requirements of large-scale production. For this reason, in recent years, methods for setting or optimizing injection molding process parameters based on Design of Experiments (DOE) have been proposed. The Design of Experiments method (DOE) aims to improve product quality and optimize the process flow, mainly including steps such as identifying and screening key factors that have a significant impact on the results, analyzing and finding the best combination of process conditions, and verifying the repeatability of the combination. For example, the orthogonal experiment method can use the orthogonal experiment table design to reduce the number of experiments and accelerate the optimization process. However, the actual production of the injection molding machine is interfered by various external factors. Due to changes in external factors such as environmental conditions, data fluctuations will occur. At the same time, the experimental design creates an artificial experimental environment, which is to a certain extent far from the actual production process, and the experimental results may not match the actual situation, making it difficult to cope with the complex and changeable actual production environment. In addition, mainly relying on the experience of process personnel to conduct experiments one mold at a time, the efficiency is extremely low and it is easy to cause waste of materials, resulting in low reliability of the optimization of injection molding process parameters. Summary of the Invention

[0004] The present invention provides a method, device and system for optimizing injection molding process parameters, which solves the technical problem that the existing method for optimizing injection molding process parameters relies on experimental design and is prone to deviation from the actual production situation, resulting in low reliability of the optimization of injection molding process parameters.

[0005] A method for optimizing injection molding process parameters provided by the first aspect of the present invention includes:

[0006] Collecting the current injection molding production parameters of the injection molded part;

[0007] Performing defect prediction based on the current injection molding production parameters through a trained quality detection model, and outputting a quality detection result;

[0008] When it is determined that the injection molded part has a quality abnormality by analyzing the quality detection result, optimizing the injection molding production parameters is output through the emperor penguin breeding optimization algorithm based on the parameter set of the current injection molding production parameters.

[0009] Further, the parameter set based on the current injection molding production parameters outputs optimized injection molding production parameters through the emperor penguin breeding optimization algorithm, including:

[0010] The parameter set based on the current injection molding production parameters is used as emperor penguin members, the population size and the maximum number of iterations are initialized, and the current iteration number, the positions of emperor penguin members, the number of emperor penguin members, and the number of adult penguins in the emperor penguin population are determined;

[0011] The optimized quality inspection results of the positions of emperor penguin members in the emperor penguin population are output through the trained quality inspection model, and the fitness is determined correspondingly based on the optimized quality inspection results;

[0012] The fitness values are sorted in ascending order to form a sorted fitness matrix, and the emperor penguin members in the emperor penguin population are sorted according to the fitness matrix to form a sorted population matrix;

[0013] The emperor penguin members in the sorted population matrix are divided into an adult penguin group and a juvenile penguin group according to the order of the number of adult penguins, and it is judged whether the number of emperor penguin members in the emperor penguin population is less than the corresponding number of adult penguins;

[0014] If the number of emperor penguin members in the penguin population is less than the corresponding number of adult penguins, the position update strategy for adult penguin exploration is executed to update the positions of the emperor penguin members in the adult penguin group, and it is judged whether the number of emperor penguin members in the emperor penguin population satisfies being less than the population size;

[0015] If it is satisfied, the number of emperor penguin members is updated, and the process jumps to execute the step of outputting the optimized quality inspection results of the positions of emperor penguin members in the emperor penguin population through the trained quality inspection model until the number of emperor penguin members in the emperor penguin population is not less than the corresponding number of adult penguins;

[0016] If the number of emperor penguin members in the penguin population is not less than the corresponding number of adult penguins, the position update strategy for juvenile penguin exploration is executed to update the positions of the emperor penguin members in the juvenile penguin group, and the number of emperor penguin members is updated when the number of emperor penguin members in the emperor penguin population is less than the population size until the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size;

[0017] When the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size, the current optimal solution is updated based on the comparison between the position of the emperor penguin member with the minimum fitness in the emperor penguin population and the current optimal solution, and it is judged whether the current iteration number is less than the maximum number of iterations;

[0018] If the current iteration number is not less than the maximum number of iterations, the current optimal solution is output as the optimized injection molding production parameters.

[0019] Further, updating the positions of the emperor penguin members in the adult penguin group by implementing the adult penguin exploration position update strategy includes:

[0020] Determine the emperor penguin member with the smallest fitness in the penguin population as the penguin leader of the penguin population;

[0021] After updating the positions of the emperor penguin members in the adult penguin group based on the penguin leader implementing the first-phase penguin leader guidance strategy, implement the second-phase penguin solo exploration strategy to update the positions of the emperor penguin members in the adult penguin group;

[0022] Implementing the first-phase penguin leader guidance strategy to update the positions of the emperor penguin members in the adult penguin group includes:

[0023] ;

[0024] ;

[0025] Wherein, , , , ;

[0026] In the formula, is the emperor penguin member index, is the number of adult penguins, is the index of the emperor penguin member position coordinate, is the number of emperor penguin member position coordinates, is the first-phase penguin leader guidance strategy, is the th position coordinate of the th emperor penguin member, is the th position coordinate of the th emperor penguin member after being updated by the first-phase penguin leader guidance strategy, is the non-linear parameter, is the penguin leader, is the th position coordinate of the penguin leader, is a random number between [1, 2], is the emperor penguin member position of the th emperor penguin member, is the emperor penguin member position of the th emperor penguin member after being updated by the first-phase penguin leader guidance strategy, is the fitness of the th emperor penguin member, is the fitness of the th emperor penguin member after being updated by the first-phase penguin leader guidance strategy, is the current iteration number, is the exploration constant, is the maximum number of iterations;

[0027] The execution of the second-stage penguin solo exploration strategy updates the positions of the emperor penguin members in the adult penguin group, including:

[0028] ;

[0029] ;

[0030] Among them, ;

[0031] In the formula, is the second-stage penguin solo exploration strategy, is the search step size, is a random number between [0, 1], is the th position coordinate of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the position of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the fitness of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the th lower bound of the position coordinate, is the th upper bound of the position coordinate.

[0032] Furthermore, the execution of the juvenile penguin exploration position update strategy updates the positions of the emperor penguin members in the juvenile penguin group, including:

[0033] After randomly selecting an emperor penguin member from the adult penguin group of the penguin population to execute the first-stage adult penguin random care strategy to update the positions of the emperor penguin members in the juvenile penguin group, execute the second-stage juvenile penguin random skill practice strategy to update the positions of the emperor penguin members in the juvenile penguin group;

[0034] The execution of the first-stage adult penguin random care strategy to update the positions of the emperor penguin members in the juvenile penguin group includes:

[0035] ;

[0036] ;

[0037] Among them, , ;

[0038] In the formula, is the index of Emperor Penguin members, is the number of adult penguins, is the population size, is the index of the position coordinates of Emperor Penguin members, is the number of position coordinates of Emperor Penguin members, is the random care strategy of adult penguins in the first stage, is the th th position coordinate of the th Emperor Penguin member, th th position coordinate of the th Emperor Penguin member after being updated by the random care strategy of adult penguins in the first stage, is a randomly selected number from the set of indices of Emperor Penguin members in the adult penguin group is the th th position coordinate of the th th th position coordinate of the th th position of the Emperor Penguin member of the th Emperor Penguin member after being updated by the random care strategy of adult penguins in the first stage, th position of the Emperor Penguin member of the is the th fitness of the th Emperor Penguin member, is the fitness of the

[0039] The implementation of the second-stage random skill practice strategy for juvenile penguins updates the positions of Emperor Penguin members in the juvenile penguin group, including:

[0040] ;

[0041] ;

[0042] Among them, ;

[0043] In the formula, is the random skill practice strategy for juvenile penguins in the second stage, is the th th position coordinate of the is the current iteration number, is a random number between [0, 1], is the lower bound of the th position coordinate, is the th upper bound of the position coordinate, is a random array, is randomly selected without repetition from the set the numbers, is the number of skills in the random array, is the position of the emperor penguin member of the th emperor penguin member after the update of the random skill practice strategy of the juvenile penguins in the second stage, is the th fitness of the emperor penguin member after the update of the random skill practice strategy of the juvenile penguins in the second stage, is the rounding function.

[0044] Furthermore, it also includes:

[0045] If the current iteration number is less than the maximum iteration number, then reset the number of emperor penguin members in the emperor penguin population and update the current iteration number, the number of adult penguins, and the positions of the emperor penguin members, and then jump to execute the step of outputting the optimized quality inspection results of the positions of each emperor penguin member in the emperor penguin population through the trained quality inspection model until the current iteration number is not less than the maximum iteration number.

[0046] An injection molding process parameter optimization device provided by the second aspect of the present invention includes:

[0047] A parameter acquisition module for acquiring the current injection molding production parameters of an injection molded part;

[0048] A quality inspection module for predicting defects based on the current injection molding production parameters through a trained quality inspection model and outputting a quality inspection result;

[0049] A parameter optimization module for, when analyzing the quality inspection result to determine that the injection molded part has a quality abnormality, outputting optimized injection molding production parameters based on the parameter set of the current injection molding production parameters through the emperor penguin breeding optimization algorithm.

[0050] An injection molding process parameter optimization system provided by the third aspect of the present invention includes:

[0051] A cloud platform for acquiring the current injection molding production parameters of an injection molded part and sending them to a database;

[0052] A database for storing the current injection molding production parameters;

[0053] A backend, which is used to call the current injection molding production parameters from a database and send them to a parameter adjustment system, receive the quality inspection results sent by the parameter adjustment system or / and optimize the injection molding production parameters and send them to a user terminal;

[0054] A parameter adjustment system, which is used to perform defect prediction based on the current injection molding production parameters through a trained quality inspection model, output quality inspection results, and when it is determined that there are quality abnormalities in the injection molded parts by analyzing the quality inspection results, output optimized injection molding production parameters through the emperor penguin breeding optimization algorithm based on the parameter set of the current injection molding production parameters;

[0055] A user terminal, which is used to visually display the quality inspection results or / and the optimized injection molding production parameters.

[0056] A computer device provided in the fourth aspect of the present invention includes a memory and a processor. A computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of the injection molding process parameter optimization method described in any one of the above.

[0057] A computer-readable storage medium provided in the fifth aspect of the present invention has a computer program stored thereon. When the computer program is executed, it implements the injection molding process parameter optimization method described in any one of the above.

[0058] A computer program product provided in the sixth aspect of the present invention includes a computer program / instructions. When the computer program / instructions are executed by a processor, the injection molding process parameter optimization method described in any one of the above is implemented.

[0059] From the above technical solutions, it can be seen that the present invention has the following advantages:

[0060] The above solution of the present invention provides an injection molding process parameter optimization method, including: collecting the current injection molding production parameters of an injection molded part; performing defect prediction based on the current injection molding production parameters through a trained quality inspection model, and outputting quality inspection results; when it is determined that there are quality abnormalities in the injection molded part by analyzing the quality inspection results, outputting optimized injection molding production parameters through the emperor penguin breeding optimization algorithm based on the parameter set of the current injection molding production parameters. Based on the above solution, efficient defect detection is carried out through a quality inspection model, and then combined with the emperor penguin breeding optimization algorithm, the efficient optimization of injection molding production parameters is completed based on its powerful search ability, improving the reliability of injection molding process parameter optimization. Description of the Drawings

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0062] Figure 1 It is a flowchart of the steps of an injection molding process parameter optimization method provided in Embodiment 1 of the present invention;

[0063] Figure 2 It is a schematic diagram of the network architecture of the ANFIS model provided in Embodiment 1 of the present invention;

[0064] Figure 3 It is a schematic diagram of the process of the emperor penguin breeding optimization algorithm provided in Embodiment 1 of the present invention;

[0065] Figure 4 It is a schematic diagram of data processing of the injection molding process parameter optimization method provided in Embodiment 1 of the present invention;

[0066] Figure 5 It is a structural block diagram of an injection molding process parameter optimization device provided in Embodiment 2 of the present invention;

[0067] Figure 6 It is a structural block diagram of an injection molding process parameter optimization system provided in Embodiment 3 of the present invention;

[0068] Figure 7 It is a working flowchart of the injection molding process parameter optimization system provided in Embodiment 3 of the present invention. Detailed implementation manners

[0069] The embodiments of the present invention provide an injection molding process parameter optimization method, device and system, which are used to solve the technical problem that the existing injection molding process parameter optimization method depends on manual experimental design, is prone to deviation from the actual production situation and has low efficiency, resulting in low reliability of injection molding process parameter optimization.

[0070] To make the invention purpose, features and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0071] Please refer to Figure 1 , Figure 1It is a flowchart of the steps of an optimization method for injection molding process parameters provided in Embodiment 1 of the present invention.

[0072] An optimization method for injection molding process parameters provided in this embodiment includes:

[0073] Step 101: Collect the current injection molding production parameters of the injection molded part.

[0074] An injection molded part refers to a plastic product produced by an injection molding process.

[0075] Injection molding production parameters refer to various variable parameters that affect the quality and production efficiency of injection molding during the injection molding production process, such as temperature parameters, time parameters, pressure parameters, speed parameters, and position parameters, etc.; and the current injection molding production parameters refer to the injection molding production parameters currently being used for injection molding production.

[0076] It should be noted that in this embodiment, the current injection molding production parameters of the injection molding machine are collected through sensors, cloud platforms, etc., in order to check the defect information of the injection molded parts produced according to these injection molding production parameters.

[0077] In a specific implementation manner of this embodiment, the current injection molding production parameters include melt temperature, injection speed, injection pressure, commutation position, holding pressure, holding time, and cooling time.

[0078] Step 102: Based on the current injection molding production parameters, perform defect prediction through a trained quality inspection model and output a quality inspection result.

[0079] A quality inspection model refers to a model that can predict and output the defect information of injection molded parts based on the input injection molding production parameters through algorithms such as machine learning or deep learning.

[0080] The quality inspection result refers to the inspection result reflecting the defect information of the injection molded part, and the defect information can include forms such as defect type and defect quality score.

[0081] It should be noted that in specific implementation, historical injection molding production parameters of injection molded parts and the corresponding true defect information can be collected first. Based on the historical injection molding production parameters and the true defect information, the quality inspection model to be trained is trained to determine the trained quality inspection model. Among them, the historical injection molding production parameters refer to the injection molding production parameters that have been used to complete the injection molding production. It can be understood that there are various process variable parameters involved in the real-time production process. Therefore, data preprocessing work such as outlier removal and normalization can be carried out on the injection molding production data collected through sensors, cloud platforms, etc. The injection molding production data refers to all process production parameters involved in the injection molding production process. Then, the injection molding production data is screened through expert experience and other methods, and several important influencing parameters are selected as the injection molding production parameters in this embodiment, thereby determining the historical injection molding production parameters. The true defect information is used to calculate the model loss value with the predicted quality inspection result output by the quality inspection model to be trained based on the historical injection molding production parameters, and it can be determined through expert inspection; the current injection molding production parameters are input into the trained quality inspection model to predict the defects that occur, and then the quality inspection result is output and fed back to the emperor penguin breeding optimization algorithm.

[0082] In a specific implementation manner of this embodiment, the quality inspection result includes weight (W), size (S), production time (T), and the degree of shrinkage defect (Q). Q can be divided into 4 levels, 1 represents normal, 2 represents slight shrinkage, 3 represents moderate shrinkage, and 4 represents severe shrinkage.

[0083] In a specific implementation manner of this embodiment, the quality inspection model is an adaptive neuro-fuzzy inference model.

[0084] It should be noted that the adaptive neuro-fuzzy inference model (ANFIS) is a model that combines a fuzzy inference system and a neural network. It can continuously optimize the trainable parameters in the fuzzy inference system through the neural network, making the model have a stable and correct output, and having strong learning ability and less human intervention. It is often used for some complex non-linear problems; ANFIS is a hierarchical architecture, and its network architecture is as Figure 2 shown:

[0085] The first layer (Layer1) is the fuzzification layer. The task of the fuzzification layer is to convert the specific numerical values in the data input to the ANFIS model into fuzzy membership values. The membership value represents the degree to which the input variable belongs to a certain fuzzy set. Its calculation process is shown in Equation (1) (in this embodiment, the membership is defined by the Gaussian membership function):

[0086] (1)

[0087] Among them, , ;

[0088] In the formula, is the index of the input variable, is the index of the fuzzy rule, is the number of input variables, is the number of fuzzy rules, is the th input variable, is the standard deviation, is the mean value, is the membership function, is the output of the first layer; and are the conditional parameters of the ANFIS;

[0089] The second layer (Layer2) is the rule layer, and the nodes are non - adaptive. The output of this layer is the product of all input signals of the nodes. The goal of the rule layer is to calculate the activation degree of the rule according to the fuzzy membership values of the input variables. Use Equation (2) to calculate the membership values obtained from the first layer. In the formula, is the output of the second layer:

[0090] (2)

[0091] The third layer (Layer3) is the normalization layer, where the nodes are non - adaptive. The output of this layer is the ratio of the intensity of a single signal input to the ratio of the intensities of all input signals. Use Equation (3) to calculate the normalized value of the output of the third layer :

[0092] (3)

[0093] The fourth layer (Layer4) is the defuzzification layer. In this layer, the nodes are adaptive. The output of this layer is based on the output of the third layer and the original input. The output result represents the contribution rate of each rule to the ANFIS. Use Equation (4) to calculate the defuzzified value of the output of the fourth layer , is the th correlation coefficient of the

[0094] (4)

[0095] The fifth layer (Layer5) is the summation layer, which calculates the sum of all inputs of the fourth layer; Use Equation (5) to calculate the final output result :

[0096] (5)

[0097] It can be understood that most traditional neural networks or deep learning models usually lack interpretability and require a large amount of data for training. For the optimization of injection molding process parameters, not only accurate prediction of defects is needed, but also an interpretable reasoning process is required to better assist in giving more effective parameter adjustment suggestions. The Adaptive Neuro-Fuzzy Inference System (ANFIS) is a fuzzy Sugeno model that can learn and adapt by learning and adjusting within the framework of an adaptive system. By using ANFIS for fuzzy inference of relevant defect parameters and knowledge extraction of fuzzy rules, the parameter knowledge of corresponding injection molded parts can be extracted through the mapping relationship between input and output. It can not only achieve high model interpretability but also greatly improve the prediction accuracy and stability of the model, and can efficiently and interpretably predict the defects of injection molded parts with a small number of samples. In addition, when the quality inspection results include multiple defect information, such as the quality inspection results include four defect information: weight, size, production time, and shrinkage defect, for any one of the defects, a corresponding trained ANFIS model can be obtained through separate training, that is, four models are trained for prediction.

[0098] Step 103: When it is determined through parsing the quality inspection results that the injection molded part has quality anomalies, optimize the injection molding production parameters through the Emperor Penguin Breeding Optimization Algorithm based on the parameter set of the current injection molding production parameters.

[0099] The Emperor Penguin Breeding Optimization Algorithm is an optimization algorithm inspired by the breeding behavior of emperor penguins.

[0100] The optimized injection molding production parameters refer to the injection molding production parameters obtained after parameter adjustment to meet the quality standards of injection molded parts.

[0101] It should be noted that when it is determined through parsing the quality inspection results that the injection molded parts produced by the injection molding machine have quality anomalies, for example, if the difference from the quality standards of the injection molded parts is greater than or equal to the anomaly threshold representing anomalies, it is considered that there are quality anomalies. To effectively improve the injection molding production parameters of the injection molding machine, this embodiment proposes a new optimization algorithm inspired by the breeding behavior of emperor penguins, namely the Emperor Penguin Breeding Optimization Algorithm (EPBOA). This optimization algorithm has a powerful search ability. It will perform iterative search and stop iterating when the iteration stop condition is reached, and output the best production parameters found, that is, the optimized injection molding production parameters, which can effectively optimize the injection parameters. When it is determined through parsing the quality inspection results that the injection molded parts do not have quality anomalies, the current injection molding production parameters are retained.

[0102] In a specific implementation manner of this embodiment, step 103 includes the following sub-steps:

[0103] S1. Use the parameter set based on the current injection molding production parameters as the emperor penguin members, initialize the population size and the maximum number of iterations, and determine the current iteration number, the positions of the emperor penguin members, the number of emperor penguin members, and the number of adult penguins in the emperor penguin population;

[0104] S2. Output the optimized quality inspection results of the positions of the emperor penguin members in the emperor penguin population through the trained quality inspection model, and determine the fitness corresponding to the optimized quality inspection results;

[0105] S3. Sort the fitness values of each member in ascending order to form a sorted fitness matrix, and sort the emperor penguin members in the emperor penguin population according to the fitness matrix to form a sorted population matrix;

[0106] S4. Divide the emperor penguin members in the sorted population matrix into an adult penguin group and a juvenile penguin group according to the order of the number of adult penguins, and determine whether the number of emperor penguin members in the emperor penguin population is less than the corresponding number of adult penguins;

[0107] S5. If the number of emperor penguin members in the penguin population is less than the corresponding number of adult penguins, execute the adult penguin exploration position update strategy to update the positions of the emperor penguin members in the adult penguin group, and determine whether the number of emperor penguin members in the emperor penguin population satisfies being less than the population size;

[0108] S6. If it is satisfied, update the number of emperor penguin members, and jump to execute the step of outputting the optimized quality inspection results of the positions of the emperor penguin members in the emperor penguin population through the trained quality inspection model, until the number of emperor penguin members in the emperor penguin population is not less than the corresponding number of adult penguins;

[0109] S7. If the number of emperor penguin members in the penguin population is not less than the corresponding number of adult penguins, execute the juvenile penguin exploration position update strategy to update the positions of the emperor penguin members in the juvenile penguin group, and update the number of emperor penguin members when the number of emperor penguin members in the emperor penguin population is less than the population size, until the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size;

[0110] S8. When the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size, update the current optimal solution based on the comparison between the position of the emperor penguin member with the minimum fitness in the emperor penguin population and the current optimal solution, and determine whether the current iteration number is less than the maximum number of iterations;

[0111] S9. If the current iteration number is not less than the maximum number of iterations, output the current optimal solution as the optimized injection molding production parameters.

[0112] The parameter set refers to the set of variable parameters included in the current injection molding production parameters, such as melt temperature, injection speed, injection pressure, switching position, holding pressure, holding time, and cooling time.

[0113] Population size refers to the total number of emperor penguin members participating in the search for the optimal solution in the algorithm.

[0114] Maximum number of iterations refers to the upper limit of iterations set during the operation of the algorithm.

[0115] Current iteration number refers to the number of iteration rounds that the algorithm has completed from the start of operation to the current moment.

[0116] Emperor penguin member position refers to the coordinates of the emperor penguin member in the search space, representing a candidate solution. The search space is understood as the parameter combination space composed of the parameter dimensions in the parameter set, and the emperor penguin member position is the specific value combination of the parameter dimensions in the parameter set.

[0117] Number of emperor penguin members refers to the number of emperor penguin members existing in the emperor penguin population.

[0118] Number of adult penguins refers to the number of adult penguins; adult penguins can be understood as emperor penguin members with certain experience and abilities to participate in important search decision-making and other behaviors, and the opposite is called juvenile penguins.

[0119] Optimization quality detection result refers to the quality detection result output during the optimization process.

[0120] Fitness refers to the index value for measuring the quality of emperor penguin members.

[0121] Adult penguin group refers to the group composed of adult penguins.

[0122] Juvenile penguin group refers to the group composed of juvenile penguins.

[0123] Current optimal solution refers to the injection molding production parameter combination with the best comprehensive performance that can be found during the iterative optimization process that has been carried out. This comprehensive performance can be quantitatively evaluated through fitness.

[0124] In a more specific implementation manner of this embodiment, implementing the adult penguin exploration position update strategy to update the positions of emperor penguin members in the adult penguin group includes:

[0125] Determine the emperor penguin member with the minimum fitness in the penguin population as the penguin leader of the penguin population;

[0126] After updating the positions of emperor penguin members in the adult penguin group by executing the first-stage penguin leader guidance strategy based on the penguin leader, execute the second-stage penguin independent exploration strategy to update the positions of emperor penguin members in the adult penguin group.

[0127] Penguin leader refers to the emperor penguin member with the most experience among adult penguins.

[0128] In a more specific implementation manner of this embodiment, implementing the juvenile penguin exploration position update strategy to update the positions of the emperor penguin members in the juvenile penguin group includes:

[0129] After randomly selecting emperor penguin members from the adult penguin group in the penguin population to execute the first-stage adult penguin random care strategy to update the positions of the emperor penguin members in the juvenile penguin group, execute the second-stage juvenile penguin random skill practice strategy to update the positions of the emperor penguin members in the juvenile penguin group.

[0130] In a more specific implementation manner of this embodiment, it further includes:

[0131] If the current iteration number is less than the maximum iteration number, then reset the number of emperor penguin members in the emperor penguin population and update the current iteration number, the number of adult penguins, and the positions of the emperor penguin members, and then jump to execute the step of outputting the optimized quality inspection results of the positions of each emperor penguin member in the emperor penguin population through the trained quality inspection model until the current iteration number is not less than the maximum iteration number.

[0132] It should be noted that in this embodiment, the objective function of the optimization algorithm is constructed based on the predicted quality defects to calculate the fitness, and then the optimal parameters for production are found through iterative optimization of the optimization algorithm. For the flow of the emperor penguin breeding optimization algorithm, refer to Figure 3 as shown:

[0133] Based on the parameter set of the current injection molding production parameters as the emperor penguin members, the emperor penguin population in EPBOA is divided into adult emperor penguins and juvenile emperor penguins; initialize the population size and the maximum iteration number , and set the current iteration number of the emperor penguin population and the number of emperor penguin members to 1, that is, the emperor penguin member index , and initialize the members of the emperor penguin population using Equation (6):

[0134] (6)

[0135] In the formula, is the emperor penguin member index, is the index of the position coordinates of the emperor penguin member, is the th position coordinate of the th emperor penguin member, is the th lower bound of the position coordinate, is the th upper bound of the position coordinate, is a random number between [0, 1];

[0136] Considering that during the process of going out to hunt for food, emperor penguins may encounter predators such as sea lions, dynamic member changes are introduced. The calculation process of the adult penguin population, that is, the number of adult penguins, can be referred to Equation (7):

[0137] (7)

[0138] In the formula, is the number of adult penguins, is the exploration constant (by default, it can be set to 0.3), is the population size, is the current iteration number, is the maximum iteration number;

[0139] Input the injection molding production parameters represented by each emperor penguin member in the emperor penguin population into the trained quality inspection model to predict and output the optimized quality inspection results, and determine the corresponding fitness through the objective function according to the optimized quality inspection results; after calculating the number of adult emperor penguins, sort them in ascending order according to the fitness to form a sorted fitness matrix, and sort the members according to the order of fitness in the sorted fitness matrix to form a sorted population matrix. The first members are the first group of adult penguins, and the last members are the group of juvenile penguins, as shown in Equations (8) and (9):

[0140] (8)

[0141] (9)

[0142] In the formula, is the sorted population matrix, is the position of the th emperor penguin member, is the number of position coordinates of the emperor penguin member, is the th position coordinate of the th emperor penguin member, is the sorted fitness matrix, is the fitness of the th emperor penguin member; in the matrix , from to the members represent the adult emperor penguin group, and from to the members represent the juvenile penguin group. The sorted fitness matrix successively includes the fitness corresponding to to ;

[0143] First, execute the adult penguin exploration position update strategy to update the positions of the emperor penguin members in the adult penguin group (from Update to ), and the adult penguin exploration position update strategy refers to the strategy for updating the positions of adult penguins, including the first-stage penguin leader guidance strategy and the second-stage penguin independent exploration strategy:

[0144] The first-stage penguin leader guidance strategy is that the penguin leader leads the adult penguins to go out for exploration:

[0145] The penguin leader is the most experienced individual among the adult penguins. Therefore, let , and the non-linear parameter can help the algorithm achieve a balance between exploration and exploitation and prevent the algorithm from falling into local optimal solutions. The value of can be represented by Equation (10); it can be understood that in the early stage of the EPBOA algorithm, the parameter has a larger value, and the candidate solutions tend to explore more towards the current optimal region. In the later stage, the candidate solutions pay more attention to exploring around themselves. The non-linear parameter can not only accelerate the convergence speed of the algorithm but also effectively avoid the algorithm from falling into local optimal solutions; at this stage, Equation (11) can be used to update the proxy positions of the adult penguins, and then the positions of each member can be adjusted according to Equation (12):

[0146] (10)

[0147] (11)

[0148] (12)

[0149] Among them, , ;

[0150] In the formula, is the emperor penguin member index, is the number of adult penguins, is the index of the emperor penguin member position coordinates, is the number of emperor penguin member position coordinates, is the first-stage penguin leader guidance strategy, is the th th position coordinate of the th emperor penguin member after being updated by the first-stage penguin leader guidance strategy, th th position coordinate of the is the non-linear parameter, is the penguin leader, is the The position coordinate, is a random number between [1, 2], is the position of the th Emperor Penguin member, is the position of the th Emperor Penguin member after the update of the first-stage penguin leader's guiding strategy, is the th Emperor Penguin member's fitness, is the th Emperor Penguin member's fitness after the update of the first-stage penguin leader's guiding strategy, is the current iteration number, is the exploration constant (which can take the value of 0.3 in this embodiment), is the maximum number of iterations;

[0151] The second-stage penguin's solo exploration strategy means that adult penguins leave the group and can only explore independently:

[0152] This strategy can help candidate solutions find better solutions near their own positions without relying on other search positions in the group. The search step size of the current position is represented by Equation (13), and then Equation (14) is used to update the positions of the adult penguin agents in the second stage. Then, the positions of each member are adjusted according to Equation (15):

[0153] (13)

[0154] (14)

[0155] (15)

[0156] In the formula, is the second-stage penguin's solo exploration strategy, is the search step size, is a random number between [0, 1] (when is satisfied, reverse exploration is performed), is the th position coordinate of the th Emperor Penguin member after the update of the second-stage penguin's solo exploration strategy, is the position of the th Emperor Penguin member after the update of the second-stage penguin's solo exploration strategy, is the th Emperor Penguin member's fitness after the update of the second-stage penguin's solo exploration strategy, is the lower bound of the is the upper bound of the coordinate of the

[0157] By increasing the number of emperor penguin members in the emperor penguin population in a loop until the update of the adult penguin group is completed, that is, when the number of emperor penguin members in the emperor penguin population is not less than the number of adult penguins, execute the juvenile penguin exploration position update strategy to update the positions of the emperor penguin members in the juvenile penguin group (from update to ), and the juvenile penguin exploration position update strategy refers to the strategy for updating the positions of juvenile penguins, including the first-stage adult penguin random care strategy and the second-stage juvenile penguin random skill practice strategy:

[0158] In the first-stage adult penguin random care strategy, this strategy can avoid aggregating towards the optimal position and thus failing to achieve the effect of global search. Use Equation (16) to update the positions of the juvenile penguin agents in the first stage, and then adjust the positions of each member according to Equation (17):

[0159] (16)

[0160] (17)

[0161] Where , ;

[0162] In the formula is the emperor penguin member index, is the number of adult penguins, is the population size, is the index of the emperor penguin member position coordinate, is the number of emperor penguin member position coordinates, is the first-stage adult penguin random care strategy, is the th coordinate of the th emperor penguin member, is the th coordinate of the th emperor penguin member after being updated by the first-stage adult penguin random care strategy, is a randomly selected number from the set of emperor penguin member indices is the th coordinate of the th th emperor penguin member, coordinate of the is the The position of the emperor penguin member among the emperor penguin members, is the th position of the emperor penguin member after the update of the adult penguin random care strategy in the first stage, is the th fitness of the emperor penguin member, is the th fitness of the emperor penguin member after the update of the adult penguin random care strategy in the first stage;

[0163] In the second stage, during the random skill practice strategy of the juvenile penguins, the juvenile penguins will explore and practice various skills. This strategy improves the local exploration ability of EPBOA. Each time, the skills in the random array are explored and practiced, the number of skills in , where is the floor function, is the problem dimension, that is, the parameter dimension in the parameter set. Therefore, , where, is randomly selected without repetition from the set ; The position of the juvenile penguin agent in the second stage is updated using Equation (18), and then the position of each member is adjusted according to Equation (19);

[0164] (18)

[0165] (19)

[0166] In the formula, is the random skill practice strategy of the juvenile penguins in the second stage, is the th th position coordinate of the emperor penguin member after the update of the random skill practice strategy of the juvenile penguins in the second stage, is the current iteration number, is a random number between [0, 1], is the th lower bound of the position coordinate, is the th upper bound of the position coordinate, is the random array, is randomly selected without repetition from the set ; is the number of skills in the random array, is the position of the emperor penguin member of the th emperor penguin member after the update of the random skill practice strategy of the juvenile penguins in the second stage, The fitness of the th emperor penguin member after the update of the random skill practice strategy for juvenile penguins in the second stage;

[0167] Since the production process of the injection molding machine belongs to a non-linear problem and there is no global optimal solution, a meta-heuristic optimization algorithm can be used to find the best or near-optimal solution to the problem through the process of searching and optimization; in this embodiment, the fitness of the optimization algorithm can be constructed based on the output optimization quality detection results. For example, the objective function can be constructed by linearly combining the defect information included in the quality detection results according to the corresponding weights. In this embodiment, Equation (20) can be used as the objective function to calculate the fitness of the optimization algorithm:

[0168] (20)

[0169] In the formula, is the fitness, is the weight of weight, is the weight of size, is the weight of production time, is the weight of the degree of shrinkage defect, is the standard weight of the injection molded part, is the weight in the quality detection result, is the standard size of the injection molded part, is the size in the quality detection result, is the production time in the quality detection result, is the degree of shrinkage defect in the quality detection result; it can be understood that it is most important that the produced product has no defects. Therefore, in this embodiment, the weight of the degree of shrinkage defect can be set to be the largest among the four weights. For example, , , and are respectively set to 0.25, 0.25, 0.1 and 0.4;

[0170] By circularly increasing the number of emperor penguin members in the emperor penguin population until the update of the juvenile penguin group is completed, that is, after the number of emperor penguin members in the emperor penguin population is not less than the population size, the position of the emperor penguin member corresponding to the minimum fitness output in the current iteration is compared with the current optimal solution to update the current optimal solution, the current iteration number is incremented by 1. If the current iteration number has not reached the set maximum iteration number, the number of emperor penguin members in the emperor penguin population is reset to 1 and the position is updated until the set maximum iteration number is reached, and then the optimized parameter result is obtained.

[0171] For better illustration, refer to Figure 4, taking the current injection molding production parameters as an example of the quality inspection model, the overall framework diagram of the injection molding process parameter optimization method of this embodiment is shown: input the current injection molding production parameters, and the injection molding production parameters may include parameters such as melt temperature, injection speed, injection pressure, and holding time. The number of input parameters can be increased or decreased according to needs. These parameters are the parameters to be optimized, that is, a set of parameters that need to obtain the most suitable production; input the current injection molding production parameters into each trained ANFIS model (ANFIS1, ANFIS2, ANFIS3, and ANFIS4). ANFIS will perform inference and prediction on the weight, size, production time, and shrinkage defect degree of the injection molded parts, input the predicted results into the optimization algorithm, and find the best production parameters, that is, optimize the injection molding production parameters, through iterative optimization of the optimization algorithm. Comparing the optimized injection molding production parameters with the current injection molding production parameters can find the parameters that need to be specifically adjusted. It should be noted that the specific implementation process of each step can be understood by referring to the relevant content in the foregoing embodiments, and will not be elaborated here.

[0172] In the embodiment of the present invention, the optimization of injection molding production parameters is divided into two parts: defect judgment and parameter adjustment. High-efficiency defect detection is performed through a quality inspection model, and then combined with the emperor penguin breeding optimization algorithm to complete the efficient optimization of injection molding production parameters based on its strong search ability, improving the reliability of injection molding process parameter optimization. <A

[0173] Please refer to Figure 5 , Figure 5 which is the structural block diagram of an injection molding process parameter optimization device provided in the second embodiment of the present invention.

[0174] An injection molding process parameter optimization device provided in this embodiment includes:

[0175] A parameter acquisition module 501, configured to acquire the current injection molding production parameters of the injection molded parts;

[0176] A quality inspection module 502, configured to perform defect prediction based on the current injection molding production parameters through a trained quality inspection model, and output a quality inspection result;

[0177] A parameter optimization module 503, configured to, when it is determined that the injection molded parts have quality abnormalities by analyzing the quality inspection result, output optimized injection molding production parameters based on the parameter set of the current injection molding production parameters through the emperor penguin breeding optimization algorithm.

[0178] Further, outputting optimized injection molding production parameters based on the parameter set of the current injection molding production parameters through the emperor penguin breeding optimization algorithm includes:

[0179] Taking the parameter set based on the current injection molding production parameters as a member of the emperor penguin, initialize the population size and the maximum number of iterations, and determine the current iteration number, the positions of the emperor penguin members, the number of emperor penguin members, and the number of adult penguins in the emperor penguin population;

[0180] Output the optimized quality inspection results of the positions of the emperor penguin members in the emperor penguin population through the trained quality inspection model, and determine the fitness based on the optimized quality inspection results;

[0181] Sort the fitness values in ascending order to form a sorted fitness matrix, and sort the emperor penguin members in the emperor penguin population according to the fitness matrix to form a sorted population matrix;

[0182] Divide the emperor penguin members in the sorted population matrix into an adult penguin group and a juvenile penguin group according to the order of the number of adult penguins, and judge whether the number of emperor penguin members in the emperor penguin population is less than the corresponding number of adult penguins;

[0183] If the number of emperor penguin members in the penguin population is less than the corresponding number of adult penguins, execute the adult penguin exploration position update strategy to update the positions of the emperor penguin members in the adult penguin group, and judge whether the number of emperor penguin members in the emperor penguin population satisfies being less than the population size;

[0184] If it is satisfied, update the number of emperor penguin members, and jump to execute the step of outputting the optimized quality inspection results of the positions of the emperor penguin members in the emperor penguin population through the trained quality inspection model until the number of emperor penguin members in the emperor penguin population is not less than the corresponding number of adult penguins;

[0185] If the number of emperor penguin members in the penguin population is not less than the corresponding number of adult penguins, execute the juvenile penguin exploration position update strategy to update the positions of the emperor penguin members in the juvenile penguin group, and update the number of emperor penguin members when the number of emperor penguin members in the emperor penguin population is less than the population size until the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size;

[0186] When the number of emperor penguin members in the emperor penguin population does not satisfy being less than the population size, update the current optimal solution based on the comparison between the position of the emperor penguin member with the minimum fitness in the emperor penguin population and the current optimal solution, and judge whether the current iteration number is less than the maximum number of iterations;

[0187] If the current iteration number is not less than the maximum number of iterations, output the current optimal solution as the optimized injection molding production parameters.

[0188] Furthermore, executing the adult penguin exploration position update strategy to update the positions of the emperor penguin members in the adult penguin group includes:

[0189] Determine the emperor penguin with the minimum fitness in the penguin population as the penguin leader of the penguin population;

[0190] After the penguin leader executes the first phase of the penguin leader guidance strategy to update the positions of the emperor penguin members in the adult penguin group, the penguin leader executes the second phase of the penguin independent exploration strategy to update the positions of the emperor penguin members in the adult penguin group;

[0191] Execute the first phase of the leader's guidance strategy to update the positions of emperor penguin members in the adult penguin group, including:

[0192] ;

[0193] ;

[0194] in, , , , ;

[0195] Where, Index for Emperor Penguin members, is the number of adult penguins, is the index of the position coordinates of the emperor penguin member, is the number of position coordinates of emperor penguin members, Provide strategy guidance to the penguin leader in the first phase, For the Emperor penguin member Position coordinates, After the first phase of the penguin leader's guidance strategy update Emperor penguin member Position coordinates, is a nonlinear parameter, For the penguin leader, For the penguin leader Position coordinates, is a random number between [1,2], For the The positions of emperor penguin members, After the first phase of the penguin leader's guidance strategy update The positions of emperor penguin members, For the The fitness of each emperor penguin member, After the first phase of the penguin leader's guidance strategy update The fitness of each emperor penguin member, is the current iteration number, To explore the constant, is the maximum number of iterations;

[0196] Execute the second - stage penguin solo exploration strategy to update the positions of the emperor penguin members in the adult penguin group, including:

[0197] ;

[0198] ;

[0199] Among them, ;

[0200] In the formula, is the second - stage penguin solo exploration strategy, is the search step size, is a random number between [0, 1], is the th position coordinate of the th emperor penguin member after being updated by the second - stage penguin solo exploration strategy, is the position of the th emperor penguin member after being updated by the second - stage penguin solo exploration strategy, is the fitness of the th emperor penguin member after being updated by the second - stage penguin solo exploration strategy, is the th lower bound of the position coordinate, is the th upper bound of the position coordinate.

[0201] Furthermore, execute the juvenile penguin exploration position update strategy to update the positions of the emperor penguin members in the juvenile penguin group, including:

[0202] After randomly selecting emperor penguin members from the adult penguin group of the penguin population to execute the first - stage adult penguin random care - taking strategy to update the positions of the emperor penguin members in the juvenile penguin group, execute the second - stage juvenile penguin random skill practice strategy to update the positions of the emperor penguin members in the juvenile penguin group;

[0203] Execute the first - stage adult penguin random care - taking strategy to update the positions of the emperor penguin members in the juvenile penguin group, including:

[0204] ;

[0205] ;

[0206] Among them, , ;

[0207] In the formula, is the emperor penguin member index, is the number of adult penguins, is the population size, Is the index of the position coordinates of the emperor penguin members, Is the number of position coordinates of the emperor penguin members, Is the random care strategy for adult penguins in the first stage, Is the th position coordinate of the Is the th position coordinate of the Is the randomly selected number from the set of indices of emperor penguin members in the adult penguin group Randomly selected number, Is the th position coordinate of the Is the th position coordinate of the Is the th Is the position of the emperor penguin member of the th Is the th Is the fitness of the th

[0208] Execute the random skill practice strategy for juvenile penguins in the second stage to update the positions of emperor penguin members in the juvenile penguin group, including:

[0209] ;

[0210] ;

[0211] Among them, ;

[0212] In the formula, Is the random skill practice strategy for juvenile penguins in the second stage, Is the th position coordinate of the Is the current iteration number, Is a random number between [0,1], Is the lower bound of the Is the The upper bound of the position coordinates, is a random array, is randomly selected without repetition from the set for non-repeating numbers, is the number of skills in the random array, is the position of the th emperor penguin member after the random skill practice strategy update of the second-stage juvenile penguins, is the th emperor penguin member's fitness after the random skill practice strategy update of the second-stage juvenile penguins, is the rounding function.

[0213] Furthermore, it further includes:

[0214] If the current iteration number is less than the maximum iteration number, then reset the number of emperor penguin members in the emperor penguin population and update the current iteration number, the number of adult penguins, and the positions of the emperor penguin members, and then jump to execute the step of outputting the optimized quality inspection results of the positions of each emperor penguin member in the emperor penguin population through the trained quality inspection model until the current iteration number is not less than the maximum iteration number.

[0215] Please refer to Figure 6 , Figure 6 which is the structural block diagram of an injection molding process parameter optimization system provided in Embodiment 3 of the present invention.

[0216] An injection molding process parameter optimization system provided in this embodiment includes:

[0217] A cloud platform 601, configured to collect the current injection molding production parameters of the injection molded part and send them to the database 602;

[0218] The database 602 is configured to store the current injection molding production parameters;

[0219] The backend 603 is configured to call the current injection molding production parameters from the database 602 and send them to the parameter adjustment system 604, and receive the quality inspection results or / and optimized injection molding production parameters sent by the parameter adjustment system 604 and send them to the user terminal 605;

[0220] The parameter adjustment system 604 is configured to perform defect prediction based on the current injection molding production parameters through the trained quality inspection model, output the quality inspection results, and when it is determined that there is a quality anomaly in the injection molded part by analyzing the quality inspection results, output the optimized injection molding production parameters based on the parameter set of the current injection molding production parameters through the emperor penguin reproduction optimization algorithm;

[0221] The user terminal 605 is configured to visually display the quality inspection results or / and the optimized injection molding production parameters.

[0222] It should be noted that the working process of the device in this embodiment can be referred to Figure 7 as follows: First, the cloud platform 601 collects the injection molding production data of the corresponding injection molding machine during the production of injection molded parts, and performs data preprocessing work such as outlier removal and normalization on the data. After that, several important influencing parameters are selected as the current injection molding production parameters, and are sent to the database 602 (MySQL) for storage for subsequent calls. After the backend 603 calls the current injection molding production parameters from the database 602, it is transmitted to the parameter adjustment system 604 through the API interface. In the parameter adjustment system 604, defect prediction and injection molding production parameter optimization are performed according to the deployed and trained quality inspection model and the emperor penguin breeding optimization algorithm, and the output quality inspection results or / and optimized injection molding production parameters are fed back to the backend and then transmitted to the user terminal 605, so as to be presented to users such as production control engineers and managers in a visual form, so as to make decisions according to the quality inspection results or / and optimized injection molding production parameters, such as adjusting the injection molding production parameters, etc.

[0223] It can be understood that the cloud platform 601 is also used to collect the historical injection molding production parameters of the injection molded parts and the real defect information corresponding to the historical injection molding production parameters and send them to the database 602; the database 602 is also used to store the historical injection molding production parameters and the real defect information; the backend 603 is also used to call the historical injection molding production parameters and the real defect information from the database 602 and send them to the parameter adjustment system; the parameter adjustment system 604 is used to train the quality inspection model to be trained according to the historical injection molding production parameters and the real defect information to determine the trained quality inspection model.

[0224] The embodiment of the present invention also provides a computer device, including a memory and a processor, and a computer program is stored in the memory; when the computer program is executed by the processor, the processor executes the steps of the injection molding process parameter optimization method in any of the above embodiments.

[0225] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program / instruction is stored, and when the computer program / instruction is executed by the processor, the steps of the injection molding process parameter optimization method in any of the above embodiments are realized.

[0226] The embodiment of the present invention also provides a computer program product, including a computer program / instruction, and when the computer program / instruction is executed by the processor, the steps of the injection molding process parameter optimization method in any of the above embodiments are realized.

[0227] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and modules can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0228] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0229] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0230] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0231] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this 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 for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0232] As described above, the above embodiments are only used to illustrate the technical solutions 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 of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A method for optimizing injection molding process parameters, characterized in that: Including: Collecting the current injection molding production parameters of the injection molded part; Performing defect prediction based on the current injection molding production parameters through a trained quality inspection model, and outputting a quality inspection result; When it is determined that the injection molded part has a quality anomaly by analyzing the quality inspection result, outputting optimized injection molding production parameters through the Emperor Penguin Breeding Optimization Algorithm based on the parameter set of the current injection molding production parameters.

2. The optimization method of injection molding process parameters according to claim 1, characterized in that The outputting of the optimized injection molding production parameters through the Emperor Penguin Breeding Optimization Algorithm based on the parameter set of the current injection molding production parameters includes: Regarding the parameter set of the current injection molding production parameters as Emperor Penguin members, initializing the population size and the maximum number of iterations, and determining the current iteration number, the positions of Emperor Penguin members, the number of Emperor Penguin members, and the number of adult penguins in the Emperor Penguin population; Outputting the optimized quality inspection results of the positions of Emperor Penguin members in the Emperor Penguin population through a trained quality inspection model, and correspondingly determining the fitness based on the optimized quality inspection results; Sorting the fitness values in ascending order to form a sorted fitness matrix, and sorting the Emperor Penguin members in the Emperor Penguin population according to the fitness matrix to form a sorted population matrix; Dividing the Emperor Penguin members in the sorted population matrix into an adult penguin group and a juvenile penguin group in the order of the number of adult penguins, and determining whether the number of Emperor Penguin members in the Emperor Penguin population is less than the corresponding number of adult penguins; If the number of Emperor Penguin members in the penguin population is less than the corresponding number of adult penguins, execute the adult penguin exploration position update strategy to update the positions of Emperor Penguin members in the adult penguin group, and determine whether the number of Emperor Penguin members in the Emperor Penguin population satisfies being less than the population size; If it is satisfied, update the number of Emperor Penguin members, and jump to execute the step of outputting the optimized quality inspection results of the positions of Emperor Penguin members in the Emperor Penguin population through a trained quality inspection model until the number of Emperor Penguin members in the Emperor Penguin population is not less than the corresponding number of adult penguins; If the number of Emperor Penguin members in the penguin population is not less than the corresponding number of adult penguins, execute the juvenile penguin exploration position update strategy to update the positions of Emperor Penguin members in the juvenile penguin group, and update the number of Emperor Penguin members when the number of Emperor Penguin members in the Emperor Penguin population is less than the population size until the number of Emperor Penguin members in the Emperor Penguin population does not satisfy being less than the population size; When the number of Emperor Penguin members in the Emperor Penguin population does not satisfy being less than the population size, update the current optimal solution based on the comparison between the position of the Emperor Penguin member with the minimum fitness in the Emperor Penguin population and the current optimal solution, and determine whether the current iteration number is less than the maximum number of iterations; If the current iteration number is not less than the maximum number of iterations, output the current optimal solution as the optimized injection molding production parameters.

3. The method for optimizing injection molding process parameters according to claim 2, wherein The execution of the adult penguin exploration position update strategy to update the positions of Emperor Penguin members in the adult penguin group includes: Determining the Emperor Penguin member with the minimum fitness in the penguin population as the penguin leader of the penguin population; After updating the positions of Emperor Penguin members in the adult penguin group by executing the first-stage penguin leader guidance strategy based on the penguin leader, execute the second-stage penguin solo exploration strategy to update the positions of Emperor Penguin members in the adult penguin group; The execution of the first-stage penguin leader guidance strategy to update the positions of Emperor Penguin members in the adult penguin group includes: ； ； Among them, , , , ; Wherein, is the index of emperor penguin members, is the number of adult penguins, is the index of the position coordinates of emperor penguin members, is the number of the position coordinates of emperor penguin members, is the guiding strategy of the penguin leader in the first stage, is the th th position coordinate of the th emperor penguin member, is the th position coordinate of the th emperor penguin member after being updated by the guiding strategy of the penguin leader in the first stage, is the non - linear parameter, is the penguin leader, is the th position coordinate of the penguin leader, is a random number between [1, 2], is the th emperor penguin member's position, is the th emperor penguin member's position after being updated by the guiding strategy of the penguin leader in the first stage, is the th emperor penguin member's fitness, is the th emperor penguin member's fitness after being updated by the guiding strategy of the penguin leader in the first stage, is the current iteration number, is the exploration constant, is the maximum iteration number; The second stage of executing the penguin's independent exploration strategy to update the positions of the emperor penguin members of the adult penguin group includes: ； ； Among them, ; In the formula, is the second-stage penguin solo exploration strategy, is the search step size, is a random number between [0, 1], is the th position coordinate of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the emperor penguin member position of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the fitness of the th emperor penguin member after being updated by the second-stage penguin solo exploration strategy, is the th lower bound of the position coordinate, is the th upper bound of the position coordinate.

4. The optimization method of injection molding process parameters according to claim 2, wherein, The step of executing the juvenile penguin exploration position update strategy to update the positions of emperor penguin members in the juvenile penguin group includes: Randomly select emperor penguins from the adult penguin group of the penguin population and execute the first stage of the adult penguin random care strategy to update the positions of the emperor penguin members of the juvenile penguin group. Then, execute the second stage of the juvenile penguin random skill practice strategy to update the positions of the emperor penguin members of the juvenile penguin group. The method of executing the first phase of the random care strategy of adult penguins to update the positions of emperor penguin members of the juvenile penguin group includes: ； ； Among them, , ; Where, Index for Emperor Penguin members, is the number of adult penguins, is the population size, is the index of the emperor penguin member's position coordinates, is the number of position coordinates of emperor penguin members, is the random care strategy of adult penguins in the first stage, For the Emperor penguin member Position coordinates, The first stage is the second stage after the adult penguin random care strategy is updated. Emperor penguin member Position coordinates, Index collection for emperor penguin members from the adult penguin group Randomly selected numbers, For the Emperor penguin member Position coordinates, For the Emperor penguin member Position coordinates, For the The positions of emperor penguin members, The first stage is the second stage after the adult penguin random care strategy is updated. The positions of emperor penguin members, For the The fitness of each emperor penguin member, The first stage is the second stage after the adult penguin random care strategy is updated. The fitness of each emperor penguin member; The second stage of executing the random skill training strategy for young penguins to update the positions of emperor penguin members in the young penguin group includes: ； ； Among them, ; In the formula, is the random skill practice strategy for juvenile penguins in the second stage, is the th position coordinate of the th Emperor Penguin member after updating the random skill practice strategy for juvenile penguins in the second stage, is the current iteration number, is a random number between [0, 1], is the lower bound of the th position coordinate, is the upper bound of the th position coordinate, is a random array, is a randomly selected non-repeating number from the set , is the number of skills in the random array, is the Emperor Penguin member position of the th Emperor Penguin member after updating the random skill practice strategy for juvenile penguins in the second stage, is the fitness of the th Emperor Penguin member after updating the random skill practice strategy for juvenile penguins in the second stage, is the rounding function.

5. The method for optimizing injection molding process parameters according to claim 2, wherein Also includes: If the current number of iterations is less than the maximum number of iterations, the number of emperor penguin members in the emperor penguin population is reset and the current number of iterations, the number of adult penguins and the positions of emperor penguin members are updated, and then the process jumps to the step of outputting the optimized quality detection results of the positions of emperor penguin members of each emperor penguin member in the emperor penguin population through the trained quality detection model until the current number of iterations is not less than the maximum number of iterations.

6. An injection molding process parameter optimization device, characterized in that, include: Parameter collection module, used to collect the current injection molding production parameters of the injection molded parts; A quality inspection module, configured to predict defects based on the current injection molding production parameters using a trained quality inspection model and output quality inspection results; A parameter optimization module is used to output optimized injection molding production parameters based on the parameter set of the current injection molding production parameters through the emperor penguin breeding optimization algorithm when analyzing the quality inspection results to determine that the injection molding part has quality abnormalities.

7. An injection molding process parameter optimization system, characterized in that, include: The cloud platform is used to collect the current injection molding production parameters of the injection molded parts and send them to the database; Database, used to store current injection molding production parameters; The backend is used to call the current injection molding production parameters from the database and send them to the parameter adjustment system, receive the quality inspection results and / or optimize the injection molding production parameters from the parameter adjustment system and send them to the user end; The parameter adjustment system is used to predict defects based on the current injection molding production parameters using a trained quality inspection model, output quality inspection results, and, when the analysis of the quality inspection results determines that the injection molded parts have quality anomalies, output optimized injection molding production parameters based on the current injection molding production parameter set using the Emperor Penguin Breeding Optimization Algorithm; The user end is used to visualize quality inspection results and / or optimize injection molding production parameters.

8. A computer device, characterized in that, The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the method for optimizing injection molding process parameters according to any one of claims 1 to 5.

9. A computer-readable storage medium having computer programs / instructions stored thereon, characterized in that, When the computer program / instructions are executed by a processor, the steps of the method for optimizing injection molding process parameters according to any one of claims 1 to 5 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by a processor, the steps of the method for optimizing injection molding process parameters according to any one of claims 1 to 5 are implemented.

Images