Plastic manufacturing process parameter optimization method and system based on neural network

Abstract

The invention provides a plastic manufacturing process parameter optimization method and system based on a neural network, and relates to the technical field of plastic manufacturing process parameter optimizing.The method comprises the following steps that standard plastic is selected, and plastic standard performance parameters of the standard plastic are obtained; based on the plastic reference performance parameters, acquiring plastic reference process parameters through a first optimization model; obtaining plastic target performance parameters; and when the plastic target performance parameter is consistent with the plastic reference performance parameter, directly adopting the plastic reference process parameter as a plastic optimization process parameter, otherwise, based on the plastic reference performance parameter, the plastic reference process parameter and the plastic target performance parameter, obtaining the plastic optimization process parameter through a second optimization model. The method has the advantage of realizing efficient, controllable, accurate and wide-coverage process parameter optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of plastic manufacturing process parameter optimization, and in particular to a plastic manufacturing process parameter optimization method and system based on a neural network. Background Art

[0002] Plastic products are widely used in various industrial and consumer fields such as automobiles, electronics, electrical appliances, packaging, and medical care. With the continuous improvement of manufacturing precision and performance requirements, how to intelligently set manufacturing process parameters according to different plastic performance targets has become one of the key issues in improving plastic quality and production efficiency.

[0003] In existing technologies, the plastics manufacturing process involves the control of a variety of complex process parameters, such as temperature, pressure, cooling time, and injection speed. These parameters have a significant impact on the performance of the final plastic product (such as strength, toughness, hardness, and heat resistance). Traditional process parameter setting relies on empirical rules or repeated trials, which are difficult to adapt to parameter customization under different target performance requirements. They are inefficient, costly, and lack generalization capabilities. With the development of artificial intelligence, the introduction of neural network-based process parameter optimization methods in the field of plastics manufacturing has become a feasible direction. However, in practical applications, the prediction of process parameters for arbitrary target performance conditions still faces challenges. This is mainly manifested in the wide variety of controllable parameters in the manufacturing process and the diverse performance targets. Directly establishing an end-to-end mapping model from "target performance to optimal process" is easily limited by sample coverage and has poor robustness.

[0004] Therefore, there is an urgent need for a process parameter inference method that combines benchmark process knowledge with neural network optimization capabilities, which can adapt to the target performance requirements of various plastic types and achieve efficient, controllable, accurate and wide-coverage process parameter optimization. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for optimizing plastic manufacturing process parameters based on neural networks, which realizes efficient, controllable, accurate and wide-coverage process parameter optimization.

[0006] The present invention is achieved through the following technical solutions: The plastic manufacturing process parameter optimization method based on neural network includes the following steps: Selecting a benchmark plastic and obtaining a plastic benchmark performance parameter of the benchmark plastic, wherein the plastic benchmark performance parameter is used as a benchmark reference value of the performance parameter of the plastic, wherein the performance parameter of the plastic includes D performance parameters for characterizing the performance of the plastic; Based on the plastic benchmark performance parameters, obtaining plastic benchmark process parameters through a first optimization model, where the plastic benchmark process parameters are a set of M types of process parameters that make the finished plastic meet the plastic benchmark performance parameters; Obtaining plastic target performance parameters, where the plastic target performance parameters are target values ​​of performance parameters of the plastic to be manufactured, and are a set of D types of performance parameters; When the plastic target performance parameters are consistent with the plastic benchmark performance parameters, the plastic benchmark process parameters are directly used as the plastic optimization process parameters. Otherwise, based on the plastic benchmark performance parameters, the plastic benchmark process parameters and the plastic target performance parameters, the plastic optimization process parameters are obtained through the second optimization model. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

[0007] Preferably, the method for obtaining the benchmark plastic and the benchmark performance parameters of the plastic is: Obtaining planned plastic production data within a target factory within a time period T, where the planned plastic production data includes the types of plastics produced, the output of each plastic, and target performance parameters of each plastic; Obtaining a baseline evaluation parameter for each plastic based on the planned plastic production data, wherein the baseline evaluation parameter is used to evaluate the suitability of the performance parameter of the plastic as a baseline performance parameter; An evaluation is performed based on the baseline evaluation parameters of all types of plastics, and the plastic with the highest degree of suitability is selected as the baseline plastic, and its target performance parameters are used as the baseline performance parameters.

[0008] Preferably, the method for obtaining the benchmark evaluation parameters of each plastic is: ; ; in, is the benchmark evaluation parameter for the kth plastic, is the output of the kth plastic in the planned plastic production data, is the output of all plastics in the planned plastic production data, e is a natural constant, A is an intermediate parameter, is the dth parameter among the target performance parameters of the kth plastic, It is the average value of the dth parameter among the target performance parameters of all types of plastics.

[0009] Preferably, the first optimization model includes: The input layer is used to receive the plastic benchmark performance parameters and obtain an input vector , ; The first hidden layer consists of 128 neurons using the ReLu activation function, outputting : ; The second hidden layer consists of 64 neurons using the ReLu activation function, outputting : ; Output layer, used to output the vector of the plastic benchmark process parameters : ; in, 、 and is the weight, 、 and represents the bias, Represents the set of real numbers.

[0010] Preferably, the loss function used when training the first optimization model is for: ; ; ; ; in, and are the first loss function and the second loss function respectively, and is the weight, is the dth parameter in the benchmark performance parameter, is the dth parameter among the simulation performance parameters of the plastic obtained by simulation according to the output of the output layer, is the controllable threshold value of the dth parameter in the benchmark performance parameter, is the smoothing term, is a constant, is the variance of the dth parameter among the actual performance parameters of the L benchmark plastics manufactured historically.

[0011] Preferably, the method for obtaining the plastic optimization process parameters through the second optimization model is: Establishing an optimization objective function and constraint conditions through a fitting model based on the plastic benchmark performance parameters, the plastic benchmark process parameters, and the plastic target performance parameters; Obtain multiple candidate solutions based on the optimization objective function and constraint conditions; Establish evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions and select the optimal solution from multiple candidate solutions.

[0012] Preferably, the method for establishing the optimization objective function and the constraint conditions is: Fitting models for the variation of each performance parameter and the variation of the multiple process parameters are established respectively: ; in, is the dependent variable and represents the change in the dth performance parameter, , is the independent variable and represents the change in the mth process parameter, The coefficient of the bth term of the polynomial when calculating the dth change in the performance parameter, where B is the number of terms in the polynomial; The method for establishing the optimization objective function is: ; ; in, As the optimization object, , is the dth parameter in the plastic target performance parameter X, is the dth parameter in the benchmark performance parameter, P is the penalty term, is the penalty item weight, if is the truth function, is the error tolerance value of the dth parameter among the plastic target performance parameters; The method for establishing the constraint condition is to set a threshold range for each of the process parameters.

[0013] Preferably, the method for obtaining multiple candidate solutions based on the optimization objective function and constraint conditions is: Solve the problem based on the optimization objective function and the constraint conditions using the gradient descent method to obtain the first-category candidate process parameters Y, and add the first-category candidate process parameters to the set of candidate solutions; Take the mth process parameter in the first category candidate process parameter Y As the center, , respectively establish normal distribution : ; Based on normal distribution Perform single-point sampling to obtain a Disturbance process value , all The second category candidate process parameters are obtained by combining them, and the operation is repeated to obtain multiple second category candidate process parameters, and all the second category candidate process parameters are added to the set of candidate solutions.

[0014] Preferably, the method of establishing evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions is: Let the i-th candidate solution in the set of candidate solutions be , , is the total number of candidate solutions; The evaluation parameter of the i-th candidate solution in the set of candidate solutions is obtained according to the deviation of the candidate solution from the plastic target performance parameter and the cost consumption. : ; ; ; in, and is the weight, is a natural constant, The fitting model is based on The calculated performance parameters are is the target performance parameter of the plastic, The representative process parameters are The energy cost consumed when stands for finding the Euclidean norm; The candidate solution with the smallest evaluation parameter is selected as the plastic optimization process parameter.

[0015] The present invention also provides a plastic manufacturing process parameter optimization system based on a neural network, which is applied to the above-mentioned plastic manufacturing process parameter optimization method based on a neural network, comprising: A benchmark plastic selection module is used to select a benchmark plastic and obtain plastic benchmark performance parameters of the benchmark plastic. The plastic benchmark performance parameters are used as benchmark reference values ​​of the performance parameters of the plastic. The performance parameters of the plastic include D types of performance parameters used to characterize the performance of the plastic. A benchmark process parameter calculation module, configured to obtain plastic benchmark process parameters through a first optimization model based on the plastic benchmark performance parameters, wherein the plastic benchmark process parameters are a set of M types of process parameters that ensure that the finished plastic meets the plastic benchmark performance parameters; A plastic target performance parameter acquisition module is used to acquire plastic target performance parameters, where the plastic target performance parameters are target values ​​of performance parameters of the plastic to be produced, and are a collection of D types of performance parameters; The plastic optimization process parameter calculation module directly uses the plastic baseline process parameters as the plastic optimization process parameters when the plastic target performance parameters are consistent with the plastic baseline performance parameters. Otherwise, the plastic optimization process parameters are obtained through the second optimization model based on the plastic baseline performance parameters, the plastic baseline process parameters and the plastic target performance parameters. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

[0016] The technical solution of the present invention has at least the following advantages and beneficial effects: The present invention obtains benchmark plastics and their benchmark performance parameters, uses them as input and a priori basis for the optimization model, and uses a stable reference value as a benchmark for the process parameters of other plastics. This makes the optimization of process parameters highly transferable and stable, and can quickly generate usable solutions in new target scenarios, thereby improving the practicality and robustness of the overall system. The present invention selects the most representative plastic for future production as the benchmark plastic. When the target performance parameters are consistent with the benchmark performance parameters, the benchmark process parameters can be directly used to avoid unnecessary consumption of computing resources. When they are inconsistent, the optimization model is automatically called to obtain adjustment parameters, thereby improving the operating efficiency and intelligence level of the system. The present invention combines the benchmark process knowledge and target performance parameters through the second optimization model to achieve adaptive adjustment of process parameters without the need for a large number of new tests and verifications, effectively shortening the parameter debugging cycle and reducing development costs; When selecting a benchmark plastic, the present invention is based on the planned production data of the target factory within a set time period, so that the evaluation results can adapt to the current production reality and recent process capabilities of the factory, and comprehensively consider the output proportion of this type of plastic in the production plan and the concentration degree of its target performance parameters relative to the performance distribution of all plastics, thereby more scientifically reflecting the current process capabilities and mainstream product characteristics of the target factory. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of a method for optimizing plastic manufacturing process parameters based on a neural network provided in Example 1 of the present invention; Figure 2 This is a schematic diagram of the principle of the neural network-based plastic manufacturing process parameter optimization system provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0019] Example 1 This embodiment provides a method for optimizing plastic manufacturing process parameters based on a neural network. Figure 1 , including the following steps: Step S1: selecting a reference plastic and obtaining a plastic reference performance parameter of the reference plastic, wherein the plastic reference performance parameter is used as a reference value for the performance parameter of the plastic, and the performance parameter of the plastic includes D performance parameters for characterizing the performance of the plastic; In this embodiment, the method for obtaining the reference plastic and the reference performance parameters of the plastic is as follows: Obtaining planned plastic production data within a target factory within a time period T, where the planned plastic production data includes the types of plastics produced, the output of each plastic, and target performance parameters of each plastic; Obtaining a baseline evaluation parameter for each plastic based on the planned plastic production data, wherein the baseline evaluation parameter is used to evaluate the suitability of the performance parameter of the plastic as a baseline performance parameter; An evaluation is performed based on the baseline evaluation parameters of all types of plastics, and the plastic with the highest degree of suitability is selected as the baseline plastic, and its target performance parameters are used as the baseline performance parameters.

[0020] On this basis, the method for obtaining the benchmark evaluation parameters of each plastic is as follows: ; ; in, is the benchmark evaluation parameter for the kth plastic, is the output of the kth plastic in the planned plastic production data, is the output of all plastics in the planned plastic production data, e is a natural constant, A is an intermediate parameter, is the dth parameter among the target performance parameters of the kth plastic, It is the average value of the dth parameter among the target performance parameters of all types of plastics.

[0021] In this step, by using the planned data of the target factory's current production cycle as the basis for analysis, the selected benchmark plastic can truly reflect the factory's current product structure and manufacturing focus. The benchmark evaluation parameters used in this embodiment comprehensively consider the production volume ratio. and performance concentration While ensuring sufficient sample support for the benchmark plastics, the performance parameters were also ensured to be at the general public level. This means that both representativeness and performance stability were taken into account during the selection process. The intermediate parameter A represents performance deviation. It uses a normalized form to measure the relative differences in multi-dimensional performance indicators and evaluates them using the Sigmoid function. This effectively reduces the impact of individual extreme values ​​or a few outliers on the evaluation results, enhancing the robustness and universality of the results.

[0022] Step S2: Based on the plastic benchmark performance parameters, obtain plastic benchmark process parameters through the first optimization model, where the plastic benchmark process parameters are a set of M types of process parameters that make the finished plastic meet the plastic benchmark performance parameters; As a preferred solution, the first optimization model includes: The input layer is used to receive the plastic benchmark performance parameters and obtain an input vector , ; The first hidden layer consists of 128 neurons using the ReLu activation function, outputting : ; The second hidden layer consists of 64 neurons using the ReLu activation function, outputting : ; Output layer, used to output the vector of the plastic benchmark process parameters : ; in, 、 and is the weight, 、 and represents the bias, Represents the set of real numbers.

[0023] Preferably, the loss function used when training the first optimization model is for: ; ; ; ; in, and are the first loss function and the second loss function respectively, and is the weight, is the dth parameter in the benchmark performance parameter, is the dth parameter among the simulation performance parameters of the plastic obtained by simulation according to the output of the output layer, is the controllable threshold value of the dth parameter in the benchmark performance parameter, For example, the smoothing term can be , is a constant, is the variance of the dth parameter among the actual performance parameters of the L benchmark plastics manufactured historically.

[0024] The first optimization model of this embodiment adopts two layers of ReLU hidden layers, has strong nonlinear modeling capabilities, and has a moderate number of network parameters. It can improve training efficiency and generalization ability while ensuring the expressiveness of the model, and is suitable for modeling complex mapping relationships between various plastic properties and process parameters.

[0025] When training the model, the mean square error between the plastic target performance and the simulated manufacturing performance is integrated And the improved Huber loss Optimize the performance of the network. Among them, the traditional Huber loss Design, which represents a threshold in Huber loss, that is, it determines how large the error is when a greater penalty is triggered. Here, the value is , The larger the value, the more tolerant it is, that is, only large deviations will be considered abnormal. In simple terms, It is the penalty threshold of the dth parameter in the benchmark performance parameter, which enables the model to automatically adapt to indicators of different physical magnitudes and unit differences, and improve the reliability of the loss function. The coefficient is set here. Make it sized Based on controllable, generally speaking Can be set to 1-3.

[0026] Step S3: Obtaining plastic target performance parameters, which are target values ​​of performance parameters of the plastic to be produced, and are a set of D types of performance parameters; Step S4: When the plastic target performance parameters are consistent with the plastic baseline performance parameters, the plastic baseline process parameters are directly used as the plastic optimization process parameters; otherwise, based on the plastic baseline performance parameters, the plastic baseline process parameters and the plastic target performance parameters, the plastic optimization process parameters are obtained through the second optimization model. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

[0027] The method for obtaining the plastic optimization process parameters through the second optimization model is: Establishing an optimization objective function and constraint conditions through a fitting model based on the plastic benchmark performance parameters, the plastic benchmark process parameters, and the plastic target performance parameters; Obtain multiple candidate solutions based on the optimization objective function and constraint conditions; Establish evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions and select the optimal solution from multiple candidate solutions.

[0028] On this basis, the method for establishing the optimization objective function and the constraint conditions is: Fitting models for the variation of each performance parameter and the variation of the multiple process parameters are established respectively: ; in, is the dependent variable and represents the change in the dth performance parameter, , is the independent variable and represents the change in the mth process parameter, The coefficient of the bth term of the polynomial when calculating the dth change in the performance parameter, where B is the number of terms in the polynomial; The method for establishing the optimization objective function is: ; ; in, As the optimization object, , is the dth parameter among the plastic target performance parameters, is the dth parameter in the benchmark performance parameter, P is the penalty term, is the penalty item weight, if is the truth function, is the error tolerance value of the dth parameter in the plastic target performance parameter X, that is, X here is a collection of multiple parameters; The method for establishing the constraint condition is to set a threshold range for each of the process parameters.

[0029] Based on the above scheme, the method for obtaining multiple candidate solutions based on optimizing the objective function and solving the constraints is: Solve the problem based on the optimization objective function and the constraint conditions using the gradient descent method to obtain the first-category candidate process parameters Y, and add the first-category candidate process parameters to the set of candidate solutions; Take the mth process parameter in the first category candidate process parameter Y As the center, , respectively establish normal distribution : ; Based on normal distribution Perform single-point sampling to obtain a Disturbance process value , all The second category candidate process parameters are obtained by combining them, and the operation is repeated to obtain multiple second category candidate process parameters, and all the second category candidate process parameters are added to the set of candidate solutions.

[0030] Finally, the method of establishing evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions is as follows: Let the i-th candidate solution in the set of candidate solutions be , , is the total number of candidate solutions; The evaluation parameter of the i-th candidate solution in the set of candidate solutions is obtained according to the deviation of the candidate solution from the plastic target performance parameter and the cost consumption. : ; ; ; in, and is the weight, is a natural constant, The fitting model is based on The calculated performance parameters are is the target performance parameter of the plastic, The representative process parameters are The energy cost consumed when stands for finding the Euclidean norm; The candidate solution with the smallest evaluation parameter is selected as the plastic optimization process parameter.

[0031] In this embodiment, a relationship function between performance and process is constructed based on a polynomial fitting model. Specifically, the second optimization model is used to obtain optimized plastic process parameters that meet performance targets and offer cost advantages through fitting modeling and targeted optimization, based on known plastic baseline performance parameters, plastic baseline process parameters, and plastic target performance parameters. During the fitting optimization process, the difference between baseline and target performance is calculated and the process offset is used as the optimization variable. This provides a clear optimization target and a clear path, facilitating rapid convergence to a reasonable process solution.

[0032] In the establishment of the optimization objective function, in addition to introducing the error minimization term , and also introduce the product of penalty weight and penalty term As a penalty, unstable components with excessive deviations are given a larger penalty to improve optimization accuracy. After obtaining an optimal option, namely the first-category candidate process parameter Y, Gaussian perturbations are performed to obtain multiple second-category candidate process parameters, enriching the diversity of the solution space. This approach avoids falling into local optimality and guides the model to explore more potential globally optimal solutions within the feasible domain.

[0033] On this basis, for the set of candidate solutions consisting of the first category candidate process parameters and the second category candidate process parameters, the evaluation function comprehensively considers performance deviation and energy cost, and adopts an exponential weight mechanism to automatically adjust the priority according to their respective values, which has strong adaptability and practical value.

[0034] In summary, in this embodiment, a representative benchmark plastic is first selected and its performance and process parameters are collected. Based on this, a first optimization model is established using a neural network to predict the benchmark process. If the target performance parameters are inconsistent with the benchmark, a second optimization model is used to construct a mapping relationship between performance changes and process changes based on a fitting method. The optimization objective function and constraints are established, and the gradient method is used to obtain an initial solution. Then, multiple sets of candidate solutions are generated through Gaussian perturbation. Finally, a multidimensional evaluation index is constructed by combining performance error and cost. The candidate solutions are comprehensively evaluated and the optimal solution is selected as the final process parameter output. This enables migration analysis based on a highly accurate reference benchmark, avoiding the need to call complex neural networks for extensive training and optimal output for each production run, thereby improving the efficiency, stability, and adaptability of plastic performance optimization to different plastic types.

[0035] Example 2 The plastic manufacturing process parameter optimization system based on neural network in this embodiment is applied to the plastic manufacturing process parameter optimization method based on neural network in the above embodiment, see Figure 2 ,include: A benchmark plastic selection module is used to select a benchmark plastic and obtain plastic benchmark performance parameters of the benchmark plastic. The plastic benchmark performance parameters are used as benchmark reference values ​​of the performance parameters of the plastic. The performance parameters of the plastic include D types of performance parameters used to characterize the performance of the plastic. A benchmark process parameter calculation module, configured to obtain plastic benchmark process parameters through a first optimization model based on the plastic benchmark performance parameters, wherein the plastic benchmark process parameters are a set of M types of process parameters that ensure that the finished plastic meets the plastic benchmark performance parameters; A plastic target performance parameter acquisition module is used to acquire plastic target performance parameters, where the plastic target performance parameters are target values ​​of performance parameters of the plastic to be produced, and are a collection of D types of performance parameters; The plastic optimization process parameter calculation module directly uses the plastic baseline process parameters as the plastic optimization process parameters when the plastic target performance parameters are consistent with the plastic baseline performance parameters. Otherwise, the plastic optimization process parameters are obtained through the second optimization model based on the plastic baseline performance parameters, the plastic baseline process parameters and the plastic target performance parameters. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

[0036] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for optimizing plastic manufacturing process parameters based on neural networks, characterized in that: The following steps are involved: Selecting a benchmark plastic and obtaining a plastic benchmark performance parameter of the benchmark plastic, wherein the plastic benchmark performance parameter is used as a benchmark reference value of the performance parameter of the plastic, wherein the performance parameter of the plastic includes D performance parameters for characterizing the performance of the plastic; Based on the plastic benchmark performance parameters, obtaining plastic benchmark process parameters through a first optimization model, where the plastic benchmark process parameters are a set of M types of process parameters that make the finished plastic meet the plastic benchmark performance parameters; Obtaining plastic target performance parameters, where the plastic target performance parameters are target values ​​of performance parameters of the plastic to be manufactured, and are a set of D types of performance parameters; When the plastic target performance parameters are consistent with the plastic benchmark performance parameters, the plastic benchmark process parameters are directly used as the plastic optimization process parameters. Otherwise, based on the plastic benchmark performance parameters, the plastic benchmark process parameters and the plastic target performance parameters, the plastic optimization process parameters are obtained through the second optimization model. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

2. The method for optimizing plastic manufacturing process parameters based on a neural network according to claim 1, characterized in that: The method for obtaining the benchmark plastic and the benchmark performance parameters of the plastic is as follows: Obtaining planned plastic production data within a target factory within a time period T, where the planned plastic production data includes the types of plastics produced, the output of each plastic, and target performance parameters of each plastic; Obtaining a baseline evaluation parameter for each plastic based on the planned plastic production data, wherein the baseline evaluation parameter is used to evaluate the suitability of the performance parameter of the plastic as a baseline performance parameter; An evaluation is performed based on the baseline evaluation parameters of all types of plastics, and the plastic with the highest degree of suitability is selected as the baseline plastic, and its target performance parameters are used as the baseline performance parameters.

3. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 2, characterized in that: The method for obtaining the benchmark evaluation parameters of each plastic is as follows: ； ； in, is the benchmark evaluation parameter for the kth plastic, is the output of the kth plastic in the planned plastic production data, is the output of all plastics in the planned plastic production data, e is a natural constant, A is an intermediate parameter, is the dth parameter among the target performance parameters of the kth plastic, It is the average value of the dth parameter among the target performance parameters of all types of plastics.

4. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 1, characterized in that: The first optimization model includes: The input layer is used to receive the plastic benchmark performance parameters and obtain an input vector , ; The first hidden layer consists of 128 neurons using the ReLu activation function, outputting : ； The second hidden layer consists of 64 neurons using the ReLu activation function, outputting : ； Output layer, used to output the vector of the plastic benchmark process parameters : ； in, 、 and is the weight, 、 and represents the bias, Represents the set of real numbers.

5. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 4, characterized in that: The loss function used when training the first optimization model for: ； ； ； ； in, and are the first loss function and the second loss function respectively, and is the weight, is the dth parameter in the benchmark performance parameter, is the dth parameter among the simulation performance parameters of the plastic obtained by simulation according to the output of the output layer, is the controllable threshold value of the dth parameter in the benchmark performance parameter, is the smoothing term, is a constant, is the variance of the dth parameter among the actual performance parameters of the L benchmark plastics manufactured historically.

6. The method for optimizing plastic manufacturing process parameters based on a neural network according to claim 1, characterized in that: The method for obtaining the plastic optimization process parameters through the second optimization model is: Establishing an optimization objective function and constraint conditions through a fitting model based on the plastic benchmark performance parameters, the plastic benchmark process parameters, and the plastic target performance parameters; Obtain multiple candidate solutions based on the optimization objective function and constraint conditions; Establish evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions and select the optimal solution from multiple candidate solutions.

7. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 6, characterized in that: The method for establishing the optimization objective function and the constraint conditions is: Fitting models for the variation of each performance parameter and the variation of the multiple process parameters are established respectively: ； in, is the dependent variable and represents the change in the dth performance parameter, , is the independent variable and represents the change in the mth process parameter, The coefficient of the bth term of the polynomial when calculating the dth change in the performance parameter, where B is the number of terms in the polynomial; The method for establishing the optimization objective function is: ； ； in, As the optimization object, , is the dth parameter in the plastic target performance parameter X, is the dth parameter in the benchmark performance parameter, P is the penalty term, is the penalty item weight, if is the truth function, is the error tolerance value of the dth parameter among the plastic target performance parameters; The method for establishing the constraint condition is to set a threshold range for each of the process parameters.

8. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 7, characterized in that: The method for obtaining multiple candidate solutions based on the optimization objective function and constraint conditions is: Solve the problem based on the optimization objective function and the constraint conditions using the gradient descent method to obtain the first-category candidate process parameters Y, and add the first-category candidate process parameters to the set of candidate solutions; Take the mth process parameter in the first category candidate process parameter Y As the center, , respectively establish normal distribution : ； Based on normal distribution Perform single-point sampling to obtain a Disturbance process value , all The second category candidate process parameters are obtained by combining them, and the operation is repeated to obtain multiple second category candidate process parameters, and all the second category candidate process parameters are added to the set of candidate solutions.

9. The method for optimizing plastic manufacturing process parameters based on neural network according to claim 8, characterized in that: The method of establishing evaluation indicators to comprehensively evaluate multiple candidate solutions from multiple dimensions is as follows: Let the i-th candidate solution in the set of candidate solutions be , , is the total number of candidate solutions; The evaluation parameter of the i-th candidate solution in the set of candidate solutions is obtained according to the deviation of the candidate solution from the plastic target performance parameter and the cost consumption. : ； ； ； in, and is the weight, is a natural constant, The fitting model is based on The calculated performance parameters are is the target performance parameter of the plastic, The representative process parameters are The energy cost consumed when stands for finding the Euclidean norm; The candidate solution with the smallest evaluation parameter is selected as the plastic optimization process parameter.

10. A plastic manufacturing process parameter optimization system based on a neural network, applied to the plastic manufacturing process parameter optimization method based on a neural network according to any one of claims 1 to 9, characterized in that: include: A benchmark plastic selection module is used to select a benchmark plastic and obtain plastic benchmark performance parameters of the benchmark plastic. The plastic benchmark performance parameters are used as benchmark reference values ​​of the performance parameters of the plastic. The performance parameters of the plastic include D types of performance parameters used to characterize the performance of the plastic. A benchmark process parameter calculation module, configured to obtain plastic benchmark process parameters through a first optimization model based on the plastic benchmark performance parameters, wherein the plastic benchmark process parameters are a set of M types of process parameters that ensure that the finished plastic meets the plastic benchmark performance parameters; A plastic target performance parameter acquisition module is used to acquire plastic target performance parameters, where the plastic target performance parameters are target values ​​of performance parameters of the plastic to be produced, and are a collection of D types of performance parameters; The plastic optimization process parameter calculation module directly uses the plastic baseline process parameters as the plastic optimization process parameters when the plastic target performance parameters are consistent with the plastic baseline performance parameters. Otherwise, the plastic optimization process parameters are obtained through the second optimization model based on the plastic baseline performance parameters, the plastic baseline process parameters and the plastic target performance parameters. The plastic optimization process parameters are a set of M types of process parameters that make the finished plastic meet the plastic target performance parameters.

Images