Parameter transplantation method and terminal

Abstract

The present application discloses a parameter transplantation method and terminal, which relates to the field of parameter transplantation technology. The method obtains the cavity pressure curves of the process before transplantation, the process after the initial transplantation, and the process after the second transplantation, and inputs them into the feature learning model respectively, thereby obtaining the first eigenvector, the second eigenvector, and the third eigenvector. The method calculates the first characteristic distance between the first eigenvector and the second eigenvector, the second characteristic distance between the first eigenvector and the third eigenvector, and the increment between the first characteristic distance and the second characteristic distance. When the increment is not greater than a threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion. The above steps are repeated until it is determined that the iterative termination condition is met, and the operation ends and returns to deduce the process after the second transplantation. The technical solution of the present application can solve the problem that the transplantation efficiency of the existing injection molding process is low and is heavily dependent on the process experience of the process personnel.

Description

Technical Field

[0001] The present application relates to the technical field of parameter transplantation, and in particular to a parameter transplantation method and terminal. Background Art

[0002] During the injection molding process of plastic products, it's often necessary to migrate process parameters from one injection molding machine to another. This can occur, for example, when a mold is delivered to a manufacturer after trial production at the mold manufacturer, when the manufacturer is replacing an old machine with a new one, or when a change in order requires expansion of production. The same process parameters can produce inconsistent product quality across different injection molding machine models. Ideally, if all equipment parameters for both the original and the machine being migrated were available, conversion formulas could be used to determine the migrated process. However, most domestic injection molding machine manufacturers have a relatively low level of information technology, and data access for foreign injection molding machines is even more limited, making it difficult to establish a comprehensive machine database. Even within the same injection molding machine model, significant performance variations can occur due to factors such as screw wear, leading to inconsistent product quality under the same process parameters. Currently, the injection molding process migration relies on a "trial injection"-"correction" model, requiring technicians to repeatedly rely on their own process experience and error. This approach, however, requires a high level of experience and is inefficient, often completely ignoring the use of previously qualified processes. Among them, the cavity pressure curve is the "fingerprint" of the injection molding process of the injection molded product, which can directly reflect the actual change state of the plastic melt in the cavity. Therefore, ensuring the consistency of the cavity pressure curve before and after process transplantation is an important criterion for ensuring the success of process parameter transplantation. However, there is currently no effective method to use the cavity pressure curve for injection molding process parameter transplantation. Summary of the Invention

[0003] The main purpose of this application is to propose a parameter transplantation method and terminal, which aims to solve the problem that the existing injection molding process transplantation is low in efficiency and heavily relies on the process experience of process personnel, effectively utilizes the information of qualified processes before transplantation, and improves the efficiency and accuracy of injection molding process transplantation.

[0004] To achieve the above objectives, the parameter transplantation method proposed in this application includes:

[0005] Obtaining a cavity pressure curve of a process before transplantation, a cavity pressure curve of a process after initial transplantation, and a cavity pressure curve of a process after second transplantation, and inputting them into a feature learning model respectively, to obtain a first eigenvector, a second eigenvector, and a third eigenvector accordingly;

[0006] Calculate the first characteristic distance between the first eigenvector and the second eigenvector, calculate the second characteristic distance between the first eigenvector and the third eigenvector, and calculate the increment between the first characteristic distance and the second characteristic distance. When the increment is not greater than a threshold, accept the new solution. When the increment is greater than the threshold, accept the new solution according to the annealing criterion.

[0007] The steps of loop execution are to obtain the cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after initial transplantation and the cavity pressure curve of the process after second transplantation, and input them into the feature learning model respectively, correspondingly obtaining the first eigenvector, the second eigenvector and the third eigenvector and the steps of calculating the first feature distance between the first eigenvector and the second eigenvector, calculating the second feature distance between the first eigenvector and the third eigenvector, and calculating the increment between the first feature distance and the second feature distance. When the increment is not greater than the threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion until the iteration termination condition is reached. The operation ends and returns to derive the second post-transplantation process.

[0008] In one embodiment, the steps of establishing the feature learning model specifically include:

[0009] Acquire a cavity pressure curve set, perform data alignment processing on all cavity pressure curves in the cavity pressure curve set, and obtain a cavity pressure training data set;

[0010] A feature learning model is constructed based on the cavity pressure training dataset.

[0011] In one embodiment, the cavity pressure curve set is defined as follows:

[0012] ;

[0013] ;

[0014] Among them, P is the set of cavity pressure curves, is the nth cavity pressure curve, is the pressure value corresponding to the tth sampling point in the nth cavity pressure curve.

[0015] In one embodiment, each of the cavity pressure curves includes a first sampling point, the pressure value corresponding to the first sampling point is the initial pressure value of the cavity, and the first sampling point is the data alignment starting point.

[0016] In one embodiment, the step of obtaining a set of cavity pressure curves includes:

[0017] The data of each cavity pressure curve in the cavity pressure curve set is subjected to outlier filling correction and noise filtering processing.

[0018] In one embodiment, after the step of performing data alignment processing on all cavity pressure curves in the cavity pressure curve set, the method further comprises:

[0019] Normalization processing is performed on the data of each cavity pressure curve in the cavity pressure curve set to obtain a cavity pressure sample data set.

[0020] In one embodiment, the normalized calculation formula is:

[0021] ;

[0022] ;

[0023] Among them, X is the cavity pressure sample data set, is the cavity pressure data corresponding to the i-th cavity pressure curve in the cavity pressure sample data set, is the cavity pressure value corresponding to the kth sampling point.

[0024] In one embodiment, each of the cavity pressure curves includes a second sampling point, a third sampling point, and a fourth sampling point;

[0025] The pressure value corresponding to the second sampling point is the pressure value when the cavity is filled with plastic melt, the pressure value corresponding to the third sampling point is the maximum pressure value of the cavity, and the pressure value corresponding to the fourth sampling point is the pressure value when the gate solidifies;

[0026] The form of any one of the first eigenvector, the second eigenvector, and the third eigenvector is:

[0027] ;

[0028] in, is the pressure value extracted from any one of the cavity pressure curves of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation, 、 、 These are the pressure values ​​corresponding to the second sampling point, the third sampling point, and the fourth sampling point of any one of the cavity pressure curves of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation.

[0029] In one embodiment, the calculation formula of the first characteristic distance or the second characteristic distance is:

[0030] ;

[0031] in, is the pressure value corresponding to the second eigenvector or the third eigenvector, is the pressure value of the first eigenvector.

[0032] This application also proposes a terminal, comprising:

[0033] Memory; and

[0034] A processor, a parameter transplantation program stored in the memory and executed by the processor, wherein the parameter transplantation program implements the parameter transplantation method described above when executed by the processor.

[0035] The technical solution of the present application proposes a parameter transplantation method, which includes: obtaining the cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation, and inputting them into the feature learning model respectively, and correspondingly obtaining the first eigenvector, the second eigenvector, and the third eigenvector; calculating the first characteristic distance between the first eigenvector and the second eigenvector, calculating the second characteristic distance between the first eigenvector and the third eigenvector, and calculating the increment between the first characteristic distance and the second characteristic distance, accepting the new solution when the increment is not greater than a threshold, and accepting the new solution according to the annealing criterion when the increment is greater than the threshold; looping through the steps. The cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after initial transplantation and the cavity pressure curve of the process after second transplantation are obtained, and are input into the feature learning model respectively, and the first eigenvector, the second eigenvector and the third eigenvector are obtained accordingly, and the first characteristic distance between the first eigenvector and the second eigenvector is calculated, the second characteristic distance between the first eigenvector and the third eigenvector is calculated, and the increment between the first characteristic distance and the second characteristic distance is calculated. When the increment is not greater than the threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion. When the iterative termination condition is reached, the operation ends and returns to deduce the process after the second transplantation. The technical solution of the present application can solve the problem that the transplantation efficiency of the existing injection molding process is low and it is heavily dependent on the process experience of the process personnel. It effectively utilizes the information of the qualified process before transplantation and improves the efficiency and accuracy of the injection molding process transplantation. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0038] Figure 1 This is a flow chart of an embodiment of the parameter transplantation method of the present application;

[0039] Figure 2 This is a flow chart of another embodiment of the parameter transplantation method of the present application;

[0040] Figure 3 This is a flow chart of another embodiment of the parameter transplantation method of the present application;

[0041] Figure 4 This is a flow chart of another embodiment of the parameter transplantation method of the present application;

[0042] Figure 5 This is a flow chart of another embodiment of the parameter transplantation method of the present application;

[0043] Figure 6 This is a schematic diagram of an embodiment of the parameter transplantation method of the present application;

[0044] Figure 7 This is the cavity pressure curve diagram of this application;

[0045] Figure 8 This is a schematic diagram of the circuit function modules of an embodiment of the terminal of the present application.

[0046] Description of Figure Numbers:

[0047] 100. Terminal; 10. Memory; 20. Processor.

[0048] The realization of the objectives, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0050] During the injection molding process of plastic products, it's often necessary to migrate process parameters from one injection molding machine to another. This can occur, for example, when a mold is delivered to a manufacturer after trial production at the mold manufacturer, when the manufacturer is replacing an old machine with a new one, or when a change in order requires expansion of production. The same process parameters can produce inconsistent product quality across different injection molding machine models. Ideally, if all equipment parameters for both the original and the machine being migrated were available, conversion formulas could be used to determine the migrated process. However, most domestic injection molding machine manufacturers have a relatively low level of information technology, and data access for foreign injection molding machines is even more limited, making it difficult to establish a comprehensive machine database. Even within the same injection molding machine model, significant performance variations can occur due to factors such as screw wear, leading to inconsistent product quality under the same process parameters. Currently, the injection molding process migration relies on a "trial injection"-"correction" model, requiring technicians to repeatedly rely on their own process experience and error. This approach, however, requires a high level of experience and is inefficient, often completely ignoring the use of previously qualified processes. Among them, the cavity pressure curve is the "fingerprint" of the injection molding process of the injection molded product, which can directly reflect the actual change state of the plastic melt in the cavity. Therefore, ensuring the consistency of the cavity pressure curve before and after process transplantation is an important criterion for ensuring the success of process parameter transplantation. However, there is currently no effective method to use the cavity pressure curve for injection molding process parameter transplantation.

[0051] Based on this, this application proposes a parameter transplantation method, which aims to solve the problem that the existing injection molding process transplantation is low in efficiency and heavily relies on the process experience of process personnel. It effectively utilizes the information of qualified processes before transplantation and improves the efficiency and accuracy of injection molding process transplantation.

[0052] The parameter transplantation method of the present application can be realized based on terminal 100, that is, the parameter transplantation method is mainly executed on terminal 100, and terminal 100 is used to establish communication connection with the injection molding machine before transplantation and the injection molding machine after transplantation respectively through communication components. Optionally, terminal 100 can refer to a host computer, a mobile phone or a tablet computer, etc., which is not specifically limited here. The communication component can be a wired communication component or a wireless communication component. For example, when the communication component is a wireless communication component, the wireless communication component can specifically be a WiFi component, a Bluetooth component, or a 3G / 4G / 5G component, which is not specifically limited here.

[0053] It is understandable that, in order to improve the efficiency and accuracy of process transplantation, before the process transplantation, the terminal 100 often needs to establish a feature learning model for the transplantation of the injection molding process.

[0054] In one embodiment of the present application, see Figure 1 、 Figure 5 and Figure 7 , Figure 1This is a flow chart of an embodiment of the parameter transplantation method of this application. Figure 5 This is a schematic diagram of an embodiment of the parameter transplantation method of this application. Figure 7 For the cavity pressure curve diagram of this application, the steps of establishing the feature learning model specifically include:

[0055] S101, obtaining a cavity pressure curve set, performing data alignment processing on all cavity pressure curves in the cavity pressure curve set, and obtaining a cavity pressure training data set;

[0056] It can be understood that each injection molding process of the injection molding machine is a molding process of the plastic raw material from solid-liquid-solid. Since the plastic raw material will undergo volume expansion during the transformation from solid to liquid, and volume contraction during the transformation from liquid to solid, after the plastic melt fills the mold cavity, an additional volume of melt is required to fill the mold cavity.

[0057] In order to understand the changes in the pressure in the mold cavity during each injection molding process, the injection molding machine before transplantation and the injection molding machine after transplantation respectively include a pressure detection device and a control device. The pressure detection device can be implemented by a pressure sensor, and the control device can be implemented by a controller. The controllers of the injection molding machine before transplantation and the injection molding machine after transplantation respectively establish communication connections with the terminal 100. The pressure sensor is used to detect the pressure generated in the mold cavity and transmit it to the corresponding controller in the form of a signal. The controller processes the received signal and transmits it to the terminal 100. The terminal 100 displays the pressure changes in the mold cavity of the injection molding machine before transplantation or the injection molding machine after transplantation in the form of a curve, such as Figure 7 shown. Figure 7 It can represent the cavity pressure curve of the process before transplantation, or it can represent the cavity pressure curve of the process after transplantation. Figure 7In the figure, before 1 o'clock, it means that the plastic melt is filling the cavity, and the melt has not yet touched the pressure sensor; 1 o'clock means that the plastic melt touches the pressure sensor, and the cavity pressure detected by the pressure sensor; the process from 1 o'clock to 2 o'clock means that the plastic melt continues to fill until the volume in the cavity is completely filled. The pressure rise in this process is the resistance that the plastic melt needs to overcome during the flow process. The slope is related to the filling speed. The faster the speed, the higher the slope, and vice versa. 2 o'clock means that the plastic melt completely fills the cavity in terms of volume. The ideal pressure holding switching point is the pressure holding switching point. When the cavity is about to be filled, the screw of the injection molding machine before transplantation or the injection molding machine after transplantation switches from flow rate control to pressure control. This conversion point is the pressure holding switching point; the process from 2 o'clock to 3 o'clock is that the extra volume of plastic melt is filled in the pressure holding stage, the melt density increases, and the pressure rises; 3 o'clock represents that the cavity pressure reaches the maximum value; the process from 3 o'clock to 4 o'clock is that the plastic melt shrinks significantly, while the pressure holding is still continuing, and the cavity pressure drops relatively slowly; 4 o'clock represents the end of pressure holding, and the gate of the cavity begins to solidify; after 4 o'clock, the plastic melt shrinks rapidly until the cavity pressure returns to atmospheric pressure.

[0058] Thus, before the process parameters are transplanted, after the pre-transplant injection molding machine undergoes multiple injection molding processes, multiple cavity pressure curves of the pre-transplant process can be obtained, and these are composed into a cavity pressure curve set and transmitted to the terminal 100. The terminal 100 aligns the pressure data of all cavity pressure curves in the cavity pressure curve set to obtain a cavity pressure training data set.

[0059] In one embodiment of the present application, the cavity pressure training dataset can be defined as follows:

[0060] ;

[0061] ;

[0062] Among them, P is the set of cavity pressure curves, is the nth cavity pressure curve, is the pressure value corresponding to the t-th sampling point in the n-th cavity pressure curve. Specifically, t can refer to but is not limited to any point from 1 to 4 in the above embodiment, and can also refer to any point between 1 and 4. For example, when t is point 1 in the above embodiment, is the pressure value corresponding to 1 point.

[0063] In one embodiment of the present application, each cavity pressure curve includes a first sampling point, and the pressure value corresponding to the first sampling point is set as the starting point for data alignment. That is, the data of multiple cavity pressure curves are aligned starting from the pressure value corresponding to the first sampling point. In combination with the above embodiment, the first sampling point can be defined as point 1. This is because point 1 indicates that the plastic melt contacts the pressure sensor, and the pressure sensor begins to detect the cavity pressure. In other words, before point 1, the pressure sensor has not detected the cavity pressure, and after point 1, the pressure sensor detects the cavity pressure, and the cavity pressure is always changing. It can be seen that point 1 is the starting point for the pressure sensor to detect the cavity pressure. Taking point 1 as the first sampling point, the data obtained can fully reflect the pressure changes in the cavity, and then reflect the effect of the set process parameters on the injection molding of the plastic.

[0064] In one embodiment of the present application, see Figure 2 , Figure 2 This is a flow chart of another embodiment of the parameter transplantation method of the present application, wherein the step of obtaining the cavity pressure curve set includes:

[0065] S1011 , performing outlier filling correction and noise filtering processing on the data of each cavity pressure curve in the cavity pressure curve set.

[0066] In this embodiment, before aligning the data of multiple cavity pressure curves, outlier correction and noise filtering can be performed on the data of the multiple cavity pressure curves. This is because during the injection molding process, both the injection molding machine before and after transplantation may experience anomalies, resulting in abnormal values ​​and noise. Therefore, by sequentially performing outlier correction and noise filtering on the data of the multiple cavity pressure curves, the outliers and noise values ​​in each cavity pressure curve can be reduced, making the data of each cavity pressure curve more accurate and accurately and comprehensively reflecting the injection molding conditions of each injection molding machine before and after transplantation.

[0067] In one embodiment of the present application, see Figure 3 , Figure 3 This is a flow chart of another embodiment of the parameter transplantation method of the present application, wherein the step of performing data alignment processing on all cavity pressure curves in the cavity pressure curve set includes:

[0068] S1012 , normalizing the data of each cavity pressure curve in the cavity pressure curve set to obtain a cavity pressure sample data set.

[0069] In this embodiment, after the data of the multiple cavity pressure curves are aligned, the data of the multiple cavity pressure curves may be normalized. The normalization process removes the minimum and maximum values ​​of each cavity pressure curve to reduce the error of each cavity pressure curve and improve the data accuracy of each cavity pressure curve.

[0070] In one embodiment of the present application, the normalized calculation formula is:

[0071] ;

[0072] ;

[0073] Among them, X is the cavity pressure sample data set, specifically the matrix X, is the cavity pressure data corresponding to the i-th cavity pressure curve in the cavity pressure sample data set, is the cavity pressure value corresponding to the kth sampling point. It should be noted that k is less than n, which means that the number of entries in matrix X is less than the number of entries in matrix P. In other words, after the multiple cavity pressure curves are normalized, the data are further screened to improve the data accuracy of each cavity pressure curve.

[0074] S102. Construct a feature learning model based on the cavity pressure training data set.

[0075] In this embodiment, after data alignment, terminal 100 constructs a feature learning model based on the cavity pressure training dataset. To improve the accuracy of the feature learning model, in one embodiment of the present application, terminal 100 can construct the feature learning model based on the cavity pressure training dataset that has undergone outlier correction, noise filtering, data alignment, and normalization.

[0076] In one embodiment of the present application, see Figure 4 , Figure 4 This is a flow chart of another embodiment of the parameter transplantation method of the present application. The step of constructing a feature learning model based on the cavity pressure training data set specifically includes:

[0077] S1021. Establish a sparse autoencoder model;

[0078] In this embodiment, the sparse autoencoder model is a model pre-stored in the terminal 100. Specifically, the sparse autoencoder model includes an encoder and a decoder, each of which is implemented using a three-layer fully connected network. The encoder is used to encode the input data, and the decoder is used to decode the data encoded by the encoder and output it. In other words, the encoder serves as the input of the sparse autoencoder model, and the decoder serves as the output of the sparse autoencoder model.

[0079] S1022. Input the cavity pressure sample data set into the sparse autoencoder model for training to obtain a feature learning model.

[0080] In this embodiment, after the cavity pressure training data set is input into the sparse autoencoder training, a feature learning model can be obtained. That is to say, the cavity pressure training data set serves as the input of the sparse autoencoder model, and the feature learning model serves as the output of the sparse autoencoder model. There must be a training error between the input and output of the sparse autoencoder model. In this embodiment, the training error can be implemented using a loss function. The loss function is expressed as the sum of the reconstruction error term and the sparsity penalty term, where the reconstruction error term is calculated using the mean square error MSE, and the sparsity penalty term is calculated using the KL divergence. By using the gradient descent method to continuously reduce the loss and continuously update the parameters of the feature learning model, the training is completed after the loss is lower than the threshold, and the encoder part is saved as the feature learning model.

[0081] In one embodiment of the present application, see Figure 5 , Figure 5 This is a flow chart of another embodiment of the parameter transplantation method of the present application, which includes:

[0082] S10, obtaining a cavity pressure curve of a process before transplantation, a cavity pressure curve of a process after initial transplantation, and a cavity pressure curve of a process after second transplantation, and inputting them into a feature learning model respectively, to obtain a first eigenvector, a second eigenvector, and a third eigenvector correspondingly;

[0083] In this embodiment, after the process parameters are transplanted, the transplanted injection molding machine undergoes multiple injection molding processes, and multiple cavity pressure curves of the transplanted process are obtained and transmitted to the terminal 100. To determine whether the cavity pressure curves before and after the process transplantation remain consistent, the terminal 100 inputs the cavity pressure curve of the single pre-transplantation process, the cavity pressure curve of the initial post-transplantation process, and the cavity pressure curve of the second post-transplantation process into the feature learning model, thereby obtaining a first eigenvector, a second eigenvector, and a third eigenvector.

[0084] In one embodiment of the present application, each cavity pressure curve includes, in addition to the first sampling point, a second sampling point, a third sampling point, and a fourth sampling point. In combination with the above embodiments, the second sampling point can be defined as two points, the third sampling point as three points, and the fourth sampling point as four points. The second sampling point is the holding pressure switching point of the injection molding machine before transplantation or the injection molding machine after transplantation, at which point the plastic melt completely fills the cavity; the third sampling point is the point of maximum cavity pressure of the injection molding machine before transplantation or the injection molding machine after transplantation; and the fourth sampling point is the end point of holding pressure of the injection molding machine before transplantation or the injection molding machine after transplantation.

[0085] The form of any one of the first eigenvector, the second eigenvector, and the third eigenvector is:

[0086] ;

[0087] in, is the extracted pressure value, 、 、 These are the pressure values ​​corresponding to the second sampling point, the third sampling point, and the fourth sampling point of the cavity pressure curves of the process before transplantation, the process after the initial transplantation, and the process after the second transplantation, respectively.

[0088] For better understanding, the first eigenvector can be in the form of:

[0089] ;

[0090] in, is the pressure value extracted from the cavity pressure curve of the process before transplantation, is the pressure value corresponding to the second sampling point of the cavity pressure curve before transplantation, is the pressure value corresponding to the third sampling point of the cavity pressure curve before transplantation, is the pressure value corresponding to the fourth sampling point of the cavity pressure curve of the process before transplantation.

[0091] Therefore, the second eigenvector or the form of the second eigenvector can be inferred based on the form of the first eigenvector, which will not be repeated here.

[0092] S20, calculating a first characteristic distance between the first eigenvector and the second eigenvector, calculating a second characteristic distance between the first eigenvector and the third eigenvector, and calculating an increment between the first characteristic distance and the second characteristic distance, and accepting a new solution when the increment is not greater than a threshold; and accepting the new solution according to an annealing criterion when the increment is greater than the threshold;

[0093] In this embodiment, the terminal 100 compares the first characteristic distance with a threshold value, and the threshold value may be 0. If the first characteristic distance is close to 0, it means that the consistency between the cavity pressure curve of the process after the initial transplantation and the cavity pressure curve of the process before transplantation is high; if the first characteristic distance is much greater than 0, it means that the consistency between the cavity pressure curve of the process after the initial transplantation and the cavity pressure curve of the process before transplantation is high. It can be seen that the judgment of the second characteristic distance is similar to the judgment of the first characteristic distance, and no repetition is made here. In order to determine whether there is consistency between the initial transplantation and the secondary transplantation, the terminal 100 calculates the increment between the first characteristic distance and the second characteristic distance, and compares the increment with the threshold value, and the threshold value is specifically 0. When the increment between the first characteristic distance and the second characteristic distance is not greater than 0, it means that there is consistency between the initial transplantation and the secondary transplantation, and the product quality under the same process parameters can be kept consistent. The terminal 100 accepts the new solution, that is, accepts the process after the secondary transplantation, otherwise it accepts the new solution according to the annealing criterion.

[0094] In one embodiment of the present application, the calculation formula for the first characteristic distance or the second characteristic distance is:

[0095] ;

[0096] Where g is either the first characteristic distance or the second characteristic distance, is the pressure value corresponding to the second eigenvector or the third eigenvector, is the pressure value of the first eigenvector.

[0097] For ease of understanding, the calculation formula for the first characteristic distance can be:

[0098] ;

[0099] in, is the first characteristic distance, is the pressure value corresponding to the second eigenvector, is the pressure value of the first eigenvector.

[0100] S30, loop execution steps to obtain the cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after initial transplantation and the cavity pressure curve of the process after second transplantation, and input them into the feature learning model respectively, correspondingly obtaining the first eigenvector, the second eigenvector and the third eigenvector and the step of calculating the first characteristic distance between the first eigenvector and the second eigenvector, calculating the second characteristic distance between the first eigenvector and the third eigenvector, and calculating the increment between the first characteristic distance and the second characteristic distance. When the increment is not greater than the threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion until the iteration termination condition is reached. The operation ends and returns to derive the second post-transplantation process.

[0101] In this embodiment, the terminal 100 executes S10 and S20 in a loop and determines whether the iteration termination condition is reached. If the iteration number is reached, it means that the iteration termination condition has been reached, and the operation ends and returns to derive the second post-transplantation process. If the iteration number is reached, it means that the iteration termination condition has not been reached, and then returns to execute S10 and S20 until the iteration termination condition is reached and the output is output.

[0102] The technical solution of the present application proposes a parameter transplantation method, which includes: obtaining the cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation, and inputting them into the feature learning model respectively, and correspondingly obtaining the first eigenvector, the second eigenvector, and the third eigenvector; calculating the first characteristic distance between the first eigenvector and the second eigenvector, calculating the second characteristic distance between the first eigenvector and the third eigenvector, and calculating the increment between the first characteristic distance and the second characteristic distance, accepting the new solution when the increment is not greater than a threshold, and accepting the new solution according to the annealing criterion when the increment is greater than the threshold; looping through the steps. The cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after initial transplantation and the cavity pressure curve of the process after second transplantation are obtained, and are input into the feature learning model respectively, and the first eigenvector, the second eigenvector and the third eigenvector are obtained accordingly, and the first characteristic distance between the first eigenvector and the second eigenvector is calculated, the second characteristic distance between the first eigenvector and the third eigenvector is calculated, and the increment between the first characteristic distance and the second characteristic distance is calculated. When the increment is not greater than the threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion. When the iterative termination condition is reached, the operation ends and returns to deduce the process after the second transplantation. The technical solution of the present application can solve the problem that the transplantation efficiency of the existing injection molding process is low and it is heavily dependent on the process experience of the process personnel. It effectively utilizes the information of the qualified process before transplantation and improves the efficiency and accuracy of the injection molding process transplantation.

[0103] This application also proposes a terminal 100, see Figure 8 The terminal 100 includes a memory 10, a processor 20, and a parameter transplantation program stored in the memory 10 and executable on the processor 20. The parameter transplantation program is configured to implement the steps of the parameter transplantation method.

[0104] The terminal 100 provided in this application utilizes the parameter transplantation method described in the aforementioned embodiment to address the low transplant efficiency of existing injection molding processes and their heavy reliance on the process personnel's experience. This effectively utilizes information about pre-transplant qualified processes, improving the efficiency and accuracy of injection molding process transplantation. Compared to the prior art, the beneficial effects of the terminal 100 provided in this application are the same as those of the parameter transplantation method described in the aforementioned embodiment, and the other technical features of the terminal 100 are the same as those disclosed in the aforementioned embodiment, and are not further detailed here.

[0105] The above description is only a preferred embodiment of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application description and drawings under the inventive concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A parameter transplantation method, characterized in that: include: Obtaining a cavity pressure curve of a process before transplantation, a cavity pressure curve of a process after initial transplantation, and a cavity pressure curve of a process after second transplantation, and inputting them into a feature learning model respectively, to obtain a first eigenvector, a second eigenvector, and a third eigenvector accordingly; Calculate the first characteristic distance between the first eigenvector and the second eigenvector, calculate the second characteristic distance between the first eigenvector and the third eigenvector, and calculate the increment between the first characteristic distance and the second characteristic distance. When the increment is not greater than a threshold, accept the new solution. When the increment is greater than the threshold, accept the new solution according to the annealing criterion. The steps of loop execution are to obtain the cavity pressure curve of the process before transplantation, the cavity pressure curve of the process after initial transplantation and the cavity pressure curve of the process after second transplantation, and input them into the feature learning model respectively, correspondingly obtaining the first eigenvector, the second eigenvector and the third eigenvector and the steps of calculating the first feature distance between the first eigenvector and the second eigenvector, calculating the second feature distance between the first eigenvector and the third eigenvector, and calculating the increment between the first feature distance and the second feature distance. When the increment is not greater than the threshold, the new solution is accepted. When the increment is greater than the threshold, the new solution is accepted according to the annealing criterion until the iteration termination condition is reached. The operation ends and returns to derive the second post-transplantation process.

2. The parameter transplantation method according to claim 1, wherein: The steps of establishing the feature learning model specifically include: Acquire a cavity pressure curve set, perform data alignment processing on all cavity pressure curves in the cavity pressure curve set, and obtain a cavity pressure training data set; A feature learning model is constructed based on the cavity pressure training dataset.

3. The parameter transplantation method according to claim 2, wherein: The definition of the cavity pressure curve set is as follows: ； ； Among them, P is the set of cavity pressure curves, is the nth cavity pressure curve, is the pressure value corresponding to the tth sampling point in the nth cavity pressure curve.

4. The parameter transplantation method according to claim 2, wherein: Each of the cavity pressure curves includes a first sampling point. The pressure value corresponding to the first sampling point is the initial pressure value of the cavity. The first sampling point is the data alignment starting point.

5. The parameter transplantation method according to claim 2, wherein: The step of obtaining the cavity pressure curve set includes: The data of each cavity pressure curve in the cavity pressure curve set is subjected to outlier filling correction and noise filtering processing.

6. The parameter transplantation method according to claim 2, wherein: After the step of performing data alignment processing on all cavity pressure curves in the cavity pressure curve set, the following steps are included: Normalization processing is performed on the data of each cavity pressure curve in the cavity pressure curve set to obtain a cavity pressure sample data set.

7. The parameter transplantation method according to claim 6, wherein: The normalized calculation formula is: ； ； Among them, X is the cavity pressure sample data set, is the cavity pressure data corresponding to the i-th cavity pressure curve in the cavity pressure sample data set, is the cavity pressure value corresponding to the kth sampling point.

8. The parameter transplantation method according to claim 2, wherein: Each of the cavity pressure curves includes a second sampling point, a third sampling point, and a fourth sampling point; The pressure value corresponding to the second sampling point is the pressure value when the cavity is filled with plastic melt, the pressure value corresponding to the third sampling point is the maximum pressure value of the cavity, and the pressure value corresponding to the fourth sampling point is the pressure value when the gate solidifies; The form of any one of the first eigenvector, the second eigenvector, and the third eigenvector is: ； in, is the pressure value extracted from any one of the cavity pressure curves of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation, 、 、 These are the pressure values ​​corresponding to the second sampling point, the third sampling point, and the fourth sampling point of any one of the cavity pressure curves of the process before transplantation, the cavity pressure curve of the process after the initial transplantation, and the cavity pressure curve of the process after the second transplantation.

9. The parameter transplantation method according to claim 8, wherein: The calculation formula of the first characteristic distance or the second characteristic distance is: ； in, is the pressure value corresponding to the second eigenvector or the third eigenvector, is the pressure value of the first eigenvector.

10. A terminal, characterized in that: include: Memory; as well as A processor, a parameter transplantation program stored in the memory and executed by the processor, wherein the parameter transplantation program implements the parameter transplantation method according to any one of claims 1 to 9 when executed by the processor.

Images