Industrial production process parameter adjustment method, adjustment equipment and storage medium

Abstract

The present application provides a method for adjusting process parameters of industrial production, an adjustment device, and a computer-readable storage medium. The method for adjusting process parameters of industrial production includes: obtaining historical production data of industrial products, the historical production data including historical parameter values ​​of process parameters of industrial products and historical index values ​​of production result indicators; determining target process parameters related to production result indicators; generating parameter prediction models for industrial products based on target process parameters and multiple groups of historical production data; obtaining target parameter values ​​corresponding to target process parameters when the index value of production result indicators is maximum based on the parameter prediction model; and adjusting target process parameters based on target parameter values ​​during the production of industrial products. The present application can find process parameters that affect production result indicators and the control range of process parameters, and is used to control process parameters of industrial production, thereby improving the quality and output of industrial production.

Description

Technical Field

[0001] The present application relates to the field of industrial control technology, and in particular to a method for adjusting process parameters of industrial production, an adjustment device, and a computer-readable storage medium. Background Art

[0002] Automation technology is widely used in industrial production production lines to improve production efficiency and reduce production costs. In the production process of industrial products, the process parameters of the production process are usually adjusted based on experience to improve the output and quality of industrial products. However, there are many process parameters involved in industrial production (for example, temperature, speed, pressure and other process parameters). How to adjust the process parameters of industrial production to maximize the output and quality of industrial production has become an urgent problem to be solved. Summary of the invention

[0003] The present application provides a method for adjusting process parameters of industrial production, an adjustment device and a computer-readable storage medium, which can find the process parameters that affect production result indicators and the control range of the process parameters, and are used to control the process parameters of industrial production, thereby improving the quality and output of industrial production.

[0004] In a first aspect, the present application provides a method for adjusting process parameters of industrial production, which is applied to an adjustment device in an industrial production system, wherein the industrial production system includes an operation device, an adjustment device, and a parameter detection device, wherein the parameter detection device includes at least two of a temperature sensor, a pressure sensor, a timer, and a speed sensor, and the method includes:

[0005] Acquire multiple groups of historical production data of industrial products, wherein each group of historical production data includes historical parameter values ​​of each process parameter and historical indicator values ​​of production result indicators, wherein the production result indicators include at least one of a yield indicator and a quality indicator;

[0006] Determine, from each of the process parameters, a process parameter related to the production result indicator as a target process parameter, wherein the target process parameter includes at least two of a temperature parameter, a pressure parameter, a time parameter, and a speed parameter, the temperature parameter is detected by the temperature sensor, the pressure parameter is detected by the pressure sensor, the time parameter is collected by the timer, and the speed parameter is detected by the speed sensor;

[0007] Based on the target process parameters and the multiple groups of historical production data, a parameter prediction model of the industrial product is generated, wherein the parameter prediction model is used to indicate the correlation relationship between the target process parameters and the production result indicators;

[0008] Based on the parameter prediction model, obtaining a target parameter value corresponding to the target process parameter when the index value of the production result index is maximum;

[0009] During the production process of the industrial product, the target process parameter is adjusted based on the target parameter value.

[0010] In some embodiments, the target process parameters include P, and generating a parameter prediction model based on the historical parameter values ​​of the target process parameters and the historical indicator values ​​includes:

[0011] Based on the target process parameters, a preliminary multivariate nonlinear regression model is established, wherein the expression of the multivariate nonlinear regression model is: in, is the predicted value of the production result index, P is the total number of target process parameters, β 0 is the regression constant term to be determined, β i is the partial regression coefficient of the first-order term to be determined, b i is the partial regression coefficient of the determined quadratic term, x i is the i-th target process parameter, α is the random error;

[0012] Based on the expression of the nonlinear regression model, a loss function of the multivariate nonlinear regression model is constructed, wherein the expression of the loss function is: in, is the predicted value of the production result index corresponding to the kth group of historical production data, y k is the true value of the production result indicator corresponding to the kth group of historical production data, and m is the number of groups of historical production data;

[0013] With the goal of minimizing the loss function, the target process parameter x in each group of the historical production data is i Substitute the historical parameter values ​​and the historical index values ​​of the production result index values ​​into the loss function, and use the least squares method to solve the regression constant term, the linear term partial regression coefficient and the quadratic term partial regression coefficient to obtain the solved multivariate nonlinear regression model;

[0014] The solved multivariate nonlinear regression model is used as the parameter prediction model.

[0015] In some embodiments, obtaining, based on the parameter prediction model, a target parameter value corresponding to the target process parameter when the index value of the production result index is maximum, includes:

[0016] Get the target process parameter x i The value constraint range of

[0017] Based on the target process parameter xi The parameter prediction model is solved for the maximum value within the value constraint range to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is maximum.

[0018] In some embodiments, the obtaining of the target process parameter x i The value constraints of include:

[0019] Based on the target process parameters, establish the target process parameters x i A basic Gaussian function, where the target process parameter x i The expression of the basic Gaussian function is: Among them, f(x i ) is the production result index, a is the peak height of the Gaussian function curve, μ is the parameter value corresponding to the peak of the Gaussian function curve, σ is the standard deviation of the Gaussian function, and x i is the i-th target process parameter;

[0020] The target process parameter x in each group of historical production data i The historical parameter values ​​and the historical index values ​​of the production result index values ​​are substituted into the target process parameter x i The target process parameter x is obtained by solving the basic Gaussian function of i The solved Gaussian function of ;

[0021] Based on the target process parameter x i The standard deviation of the Gaussian function after solution is used to obtain the target process parameter x i The value constraint range of .

[0022] In some embodiments, the target process parameter x i The value constraint range of is [μ-3σ,μ+3σ], where σ is the target process parameter x i The standard deviation of the Gaussian function after solution, μ is the target process parameter x i The peaks of the solved Gaussian function curve correspond to the parameter values.

[0023] In some embodiments, the P target process parameters are respectively a temperature parameter, a pressure parameter, a time parameter, and a speed parameter;

[0024] The expression of the multivariate nonlinear regression model is:

[0025]

[0026] Among them, x 1 is the temperature parameter, x 2 is the pressure parameter, x 3 is the time parameter, x 4is the speed parameter, β 1 is the partial regression coefficient of the temperature parameter to be determined, β 2 is the partial regression coefficient of the pressure parameter to be determined, β 3 is the partial regression coefficient of the time parameter to be determined, β 4 is the partial regression coefficient of the velocity parameter to be determined, b 1 is the partial regression coefficient of the temperature parameter to be determined, b 2 is the partial regression coefficient of the quadratic term to be determined for the pressure parameter, b 3 is the partial regression coefficient of the quadratic term to be determined for the time parameter, b 4 is the partial regression coefficient of the quadratic term of the speed parameter to be determined, and α is the random error.

[0027] In some embodiments, determining, from among the process parameters, a process parameter related to the production result indicator as a target process parameter, includes:

[0028] Traversing multiple groups of the historical production data, dividing each group of the historical production data into positive and negative samples, if the historical index value of the production result index of the current traversal data group is greater than the preset index threshold, the current traversal data group is divided into a positive sample; if the historical index value of the production result index of the current traversal data group is less than or equal to the preset index threshold, the current traversal data group is divided into a negative sample, and the label value of each group of the historical production data is obtained;

[0029] Based on the label value of each group of the historical production data and the historical parameter value of each process parameter corresponding to each group of the historical production data, the preset Xgboot model is trained to obtain a trained Xgboot model;

[0030] Calculating the importance score of each of the process parameters using the trained Xgboot model;

[0031] From the process parameters, process parameters whose importance scores are greater than a preset score threshold are extracted as the target process parameters.

[0032] In some embodiments, the target process parameters include temperature parameters, pressure parameters, time parameters, and speed parameters;

[0033] In the production process of the industrial product, adjusting the target process parameter based on the target parameter value includes:

[0034] Detecting a current parameter value of the temperature parameter by the temperature sensor;

[0035] Detecting a current parameter value of the pressure parameter by the pressure sensor;

[0036] Collecting the current parameter value of the time parameter by the timer;

[0037] Detecting a current parameter value of the speed parameter by the speed sensor;

[0038] If the current parameter value of the temperature parameter is not equal to the target parameter value corresponding to the temperature parameter, adjusting the temperature parameter to the target parameter value corresponding to the temperature parameter;

[0039] If the current parameter value of the pressure parameter is not equal to the target parameter value corresponding to the pressure parameter, adjusting the pressure parameter to the target parameter value corresponding to the pressure parameter;

[0040] If the current parameter value of the time parameter is not equal to the target parameter value corresponding to the time parameter, adjusting the time parameter to the target parameter value corresponding to the time parameter;

[0041] If the current parameter value of the speed parameter is not equal to the target parameter value corresponding to the speed parameter, the speed parameter is adjusted to the target parameter value corresponding to the speed parameter.

[0042] In a second aspect, the present application further provides an adjustment device, which includes a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, it executes any industrial production process parameter adjustment method provided in the present application.

[0043] In a third aspect, the present application also provides a computer-readable storage medium on which a computer program is stored, and the computer program is loaded by a processor to execute the process parameter adjustment method for industrial production.

[0044] In the present application, historical production data of industrial products are obtained; target process parameters related to production result indicators are determined; a parameter prediction model for industrial products is generated based on the target process parameters and multiple sets of historical production data; based on the parameter prediction model, the target parameter value corresponding to the target process parameter when the indicator value of the production result indicator is maximum is obtained; in the production process of industrial products, the target process parameters are adjusted based on the target parameter value, thereby finding the target process parameters that affect the production result indicators and the control range of the target process parameters, which are used to control the process parameters of industrial production, thereby improving the quality and output of industrial production. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 It is a schematic block diagram of the structure of an industrial production system provided in an embodiment of the present application;

[0047] Figure 2 is a schematic block diagram of the structure of an adjustment device provided in an embodiment of the present application;

[0048] Figure 3 This is a flow chart of a method for adjusting process parameters of industrial production provided by an embodiment of the present application;

[0049] Figure 4 is a flowchart of an embodiment of step 303 provided in an embodiment of the present application;

[0050] Figure 5 is a flowchart of an embodiment of step 304 provided in an embodiment of the present application;

[0051] Figure 6 It is a schematic diagram illustrating the Gaussian function relationship between target process parameters and production result indicators provided in the embodiments of the present application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0053] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0054] In the description of the embodiments of the present application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0055] In order to enable any person skilled in the art to implement and use the present application, the following description is provided. In the following description, details are listed for the purpose of explanation. It should be understood that those of ordinary skill in the art can recognize that the present application can also be implemented without using these specific details. In other examples, the known process will not be elaborated in detail to avoid unnecessary details that make the description of the present application embodiment obscure. Therefore, the present application is not intended to be limited to the embodiments shown, but is consistent with the widest range of principles and features disclosed in accordance with the embodiments of the present application.

[0056] Embodiments of the present application provide a method for adjusting process parameters of industrial production, an adjustment device, and a computer-readable storage medium.

[0057] Some embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0058] The process parameter adjustment method for industrial production of the embodiment of the present application can be applied to the adjustment equipment in the industrial production system 100, such as Figure 1 As shown, the industrial production system 100 includes an operating device 101 , an adjustment device 102 and a parameter detection device 103 .

[0059] The operating equipment 101 is used to implement specific production operations. The operating equipment 101 is a device that implements a series of operations such as processing, transportation, assembly, and inspection starting from the time the raw materials enter the production site. For example, the operating equipment 101 may include stirring equipment, transmission equipment, assembly equipment, etc.

[0060] The parameter detection device 103 is used to detect the parameter values ​​of various process parameters of the industrial production system 100. For example, it is used to detect the operating temperature and operating speed of the operating equipment 101. The parameter detection device 103 specifically includes a temperature sensor, a pressure sensor, a timer, a speed sensor, etc. The temperature sensor is used to detect the temperature parameter, the pressure sensor is used to detect the pressure parameter, the timer is used to collect the time parameter, and the speed sensor is used to detect the speed parameter.

[0061] The adjustment device 102 is used to adjust the parameter value of the process parameter of the industrial production system 100 to improve the output and quality. The adjustment device 102 can adjust the process parameter according to the parameter value of the process parameter detected by the parameter detection device 103 during the production process of the industrial product (e.g., according to the current parameter value of the target process parameter detected by the parameter detection device 103, compare the current parameter value of the target process parameter with the target parameter value to see if they are equal, and if the current parameter value of the target process parameter is not equal to the target parameter value, adjust the target process parameter to the target parameter value).

[0062] like Figure 2 As shown, Figure 2 The structure schematic block diagram of an adjustment device provided in an embodiment of the present application is shown in FIG. The adjustment device 200 includes a processor 201 and a memory 202, and the processor 201 and the memory 202 are connected via a bus 203, such as an I2C (Inter-integrated Circuit) bus.

[0063] Specifically, the processor 201 is used to provide computing and control capabilities to support the operation of the entire adjustment device 200. The processor 201 can be a central processing unit (CPU), and the processor 201 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0064] Specifically, the memory 202 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0065] Those skilled in the art will understand that Figure 2 The structure shown in the figure is merely a block diagram of a partial structure related to the embodiment scheme of the present application, and does not constitute a limitation on the adjustment device to which the embodiment scheme of the present application is applied. The specific adjustment device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0066] The processor 201 is used to run the computer program stored in the memory 202, and implements any one of the industrial production process parameter adjustment methods provided in the embodiments of the present application when executing the computer program. For example, the processor 201 is used to run the computer program stored in the memory 202, and can implement the following steps when executing the computer program:

[0067] Acquire multiple groups of historical production data of industrial products, wherein each group of historical production data includes historical parameter values ​​of each process parameter and historical index values ​​of production result indicators, and the production result indicators include at least one of output indicators and quality indicators; from each of the process parameters, determine the process parameters related to the production result indicators as target process parameters, wherein the target process parameters include at least two of temperature parameters, pressure parameters, time parameters, and speed parameters, the temperature parameter is detected by the temperature sensor, the pressure parameter is detected by the pressure sensor, the time parameter is collected by the timer, and the speed parameter is detected by the speed sensor; based on the target process parameters and multiple groups of historical production data, generate a parameter prediction model for the industrial product, wherein the parameter prediction model is used to indicate the correlation between the target process parameters and the production result indicators; based on the parameter prediction model, obtain the target parameter value corresponding to the target process parameter when the index value of the production result indicator is the largest; in the production process of the industrial product, adjust the target process parameter based on the target parameter value.

[0068] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working process of the adjustment equipment described above can refer to the corresponding process in the following industrial production process parameter adjustment method embodiment, and will not be repeated here.

[0069] In the following, Figure 2 The adjustment device shown in is used as the execution subject of the process parameter adjustment method for industrial production as an example, and the process parameter adjustment method for industrial production provided by the embodiment of the present application is described in detail. For the sake of simplicity and convenience of description, the execution subject will be omitted in the subsequent method embodiments. It should be noted that Figure 1 The scenario is only used to explain the process parameter adjustment method for industrial production provided in the embodiment of the present application, but does not constitute a limitation on the application scenario of the process parameter adjustment method for industrial production provided in the embodiment of the present application.

[0070] See also Figure 3 , Figure 3 301 to 304, wherein:

[0071] 301. Obtain multiple sets of historical production data for industrial products.

[0072] Each set of historical production data includes historical parameter values ​​of each process parameter and historical indicator values ​​of production result indicators, and the production result indicators include at least one of a yield indicator and a quality indicator.

[0073] Among them, industrial products can be light or heavy industrial products such as paints and paper towels.

[0074] Among them, process parameters are parameters that directly affect product production results (such as product characteristics, output and quality) during the manufacturing process. Process parameters may include temperature parameters, pressure parameters, time parameters, speed parameters, current and voltage parameters, etc.

[0075] Specifically, M groups of historical production data may be obtained, wherein the kth group of historical production data in the M groups of historical production data includes process parameters Historical parameter values ​​and production result indicators y k For example, the process parameters of industrial products include three parameters (process parameters x 1 、x 2 、x 3 )” as an example, the historical production data of the k=1th group of historical production data in M=2 groups of historical production data is The historical production data of the k=2th group of historical production data is in, is the process parameter x in the k=1th group of historical production data 1 The historical parameter values ​​of is the process parameter x in the k=1th group of historical production data 2 The historical parameter values ​​of is the process parameter x in the k=1th group of historical production data 3 The historical parameter value, y 1 is the historical indicator value of the production result indicator of the k=1th group of historical production data, and so on.

[0076] 302. Determine, from among the process parameters, a process parameter related to the production result indicator as a target process parameter.

[0077] The target process parameters are selected from the process parameters of the industrial products, are considered to affect the production result index, and are adjusted in the subsequent production process of the industrial products. Specifically, P target process parameters can be determined from the process parameters of the industrial products, where P is a positive integer ≥2.

[0078] In some embodiments, process parameters related to the production result indicators can be directly screened according to manual experience and used as target process parameters.

[0079] In some embodiments, in order to screen out target process parameters that potentially affect production results (such as yield, quality) to improve the index value of production results (such as yield, etc.), the target process parameters can also be determined by the following steps 3021 to 3024:

[0080] 3021. Traverse multiple groups of historical production data, and divide each group of historical production data into positive and negative samples. If the historical indicator value of the production result indicator of the current traversal data group is greater than the preset indicator threshold, the current traversal data group is divided into a positive sample; if the historical indicator value of the production result indicator of the current traversal data group is less than or equal to the preset indicator threshold, the current traversal data group is divided into a negative sample, and the label value of each group of historical production data is obtained.

[0081] For example, "the process parameters of industrial products include 5 (process parameters x 1 、x 2 、x 3 、x 4 、x 5 )” as an example, the kth group of historical production data in the M groups of historical production data If y k > preset indicator threshold, the kth group of historical production data is taken as a positive sample (a positive sample can be represented by a label value of 1); if y k ≤ the preset indicator threshold, the kth group of historical production data is taken as a negative sample (the positive sample can be represented by a label value of 0), thereby obtaining the label value of each group of historical production data in the M groups of historical production data.

[0082] 3022. Based on the label value of each group of the historical production data and the historical parameter values ​​of each process parameter corresponding to each group of the historical production data, the preset Xgboot model is trained to obtain a trained Xgboot model.

[0083] Among them, the idea of ​​the Xgboost model algorithm is to integrate many weak classifiers together to form a strong classifier, and each weak classifier is a decision tree. Exemplarily, the open source Xgboost with default parameters can be used as the preset Xgboot model, and each process parameter is used as the decision attribute of a single decision tree to perform binary classification on each group of historical production data (i.e., divide each group of historical production data into positive samples or negative samples). Then, let the preset Xgboot model perform binary classification (i.e., divide each group of historical production data into positive samples or negative samples) according to the historical parameter values ​​of each process parameter corresponding to each group of historical production data, and obtain the predicted value of each group of historical production data, and construct the loss function of the preset Xgboot model based on the predicted value of each group of historical production data and the label value of each group of historical production data, with the goal of minimizing the loss function of the preset Xgboot model, and adjust the model parameters of the preset Xgboot model; finally, the preset Xgboot model after the model parameters are adjusted is obtained as the trained Xgboot model.

[0084] In this embodiment, the training of the preset Xgboot model is similar to the existing training method. For details, please refer to the relevant instructions of the existing preset Xgboot model training, which will not be repeated here.

[0085] 3023. Calculate the importance score of each of the process parameters using the trained Xgboot model.

[0086] 3024. From the process parameters, extract the process parameters whose importance scores are greater than a preset score threshold as the target process parameters.

[0087] Specifically, the trained Xgboot model can automatically calculate the importance of features (i.e., process parameters), so the trained Xgboot model can be used to calculate the importance score of each process parameter. For example, for the five process parameters of an industrial product, the trained Xgboot model can calculate the process parameter x 1 、x 2 、x 3 、x 4 、x 5 The importance scores of are: N1, N2, N3, N4, N5. Since the importance scores N1, N2, N3, N4 are greater than the preset score threshold, and the importance score N5 is less than or equal to the preset score threshold, the process parameter x 1 、x 2 、x 3 、x 4 As the target process parameter, discard the process parameter x 5 As the target process parameter (i.e. process parameter x 5 Not used as target process parameter).

[0088] Therefore, by training the Xgboot model to calculate the importance score of each process parameter, the degree of influence of each process parameter on the production result index can be determined, so that more important process parameters can be selected as target process parameters and the target process parameters can be adjusted during the production process of industrial products, thereby effectively improving the output and quality of industrial products.

[0089] 303. Generate a parameter prediction model for the industrial product based on the target process parameters and the multiple groups of historical production data.

[0090] The parameter prediction model is used to indicate the correlation between the target process parameter and the production result indicator.

[0091] Specifically, Figure 4 As shown, step 303 may specifically include the following steps 3031 to 3034:

[0092] 3031. Based on the target process parameters, a preliminary multivariate nonlinear regression model is established.

[0093] The expression of the multivariate nonlinear regression model is shown in the following formula (1):

[0094]

[0095] In the expression of the multivariate nonlinear regression model shown in formula (1), is the predicted value of the production result index, P is the total number of target process parameters, β 0 is the regression constant term to be determined, β i is the partial regression coefficient of the first-order term to be determined, b i is the partial regression coefficient of the determined quadratic term, x i is the i-th target process parameter, and α is the random error.

[0096] 3032. Based on the expression of the nonlinear regression model, construct a loss function of the multivariate nonlinear regression model.

[0097] The loss function is expressed as follows:

[0098]

[0099] In the loss function shown in formula (2), is the predicted value of the production result index corresponding to the kth group of historical production data, y k is the true value of the production result indicator corresponding to the kth group of historical production data, and m is the number of groups of historical production data.

[0100] 3033. With the goal of minimizing the loss function, the target process parameter x in each group of the historical production data is i The historical parameter values ​​and the historical index values ​​of the production result index values ​​are substituted into the loss function, and the regression constant term, the linear term partial regression coefficient and the quadratic term partial regression coefficient are solved by the least squares method to obtain the solved multivariate nonlinear regression model.

[0101] 3034. The solved multivariate nonlinear regression model is used as the parameter prediction model.

[0102] For example, the target process parameters include three (process parameters x 1 、x 2 、x 3 )” as an example, assuming that the historical production data of the kth group of historical production data in M=20 groups is ..., the historical production data of the k=20th group of historical production data is First, based on the three target process parameters, a preliminary expression of the multivariate nonlinear regression model can be constructed, as shown in the following formula (3).

[0103]

[0104] Then, the historical index value of the production result index value corresponding to each group of historical production data in M=20 groups of historical production data, as well as formula (3), are substituted into the loss function shown in formula (2) to obtain the loss function shown in formula (4); with the goal of minimizing the loss function, the least squares method is used to solve the regression constant term, the linear partial regression coefficient and the quadratic partial regression coefficient, and then the solved regression constant term, the linear partial regression coefficient and the quadratic partial regression coefficient are substituted into the expression of the multivariate nonlinear regression model shown in formula (1) to obtain the solved multivariate nonlinear regression model; the solved multivariate nonlinear regression model is used as the parameter prediction model of industrial products.

[0105]

[0106] For example, the target process parameters include four parameters (temperature parameter x 1 , pressure parameter x 2 , time parameter x 3 , speed parameter x 4 )” as an example, assuming that the historical production data of the k=1th group of historical production data in M=30 groups of historical production data is ..., the historical production data of the k=30th group of historical production data is First, based on the four target process parameters, a preliminary expression of the multivariate nonlinear regression model can be constructed, as shown in the following formula (5).

[0107]

[0108] In formula (5), x 1 is the temperature parameter, x 2 is the pressure parameter, x 3 is the time parameter, x 4 is the speed parameter, β 1 is the partial regression coefficient of the temperature parameter to be determined, β 2 is the partial regression coefficient of the pressure parameter to be determined, β 3 is the partial regression coefficient of the time parameter to be determined, β 4 is the partial regression coefficient of the velocity parameter to be determined, b 1 is the partial regression coefficient of the temperature parameter to be determined, b 2 is the partial regression coefficient of the quadratic term to be determined for the pressure parameter, b 3is the partial regression coefficient of the quadratic term to be determined for the time parameter, b 4 is the partial regression coefficient of the quadratic term of the speed parameter to be determined, and α is the random error.

[0109] Then, the historical index values ​​of the production result index values ​​corresponding to each group of historical production data in M=30 groups of historical production data, as well as formula (5), are substituted into the loss function shown in formula (2) to obtain the loss function shown in formula (6); with the goal of minimizing the loss function, the least squares method is used to solve the regression constant term, the linear partial regression coefficient and the quadratic partial regression coefficient, and then the solved regression constant term, the linear partial regression coefficient and the quadratic partial regression coefficient are substituted into the expression of the multivariate nonlinear regression model shown in formula (1) to obtain the solved multivariate nonlinear regression model; the solved multivariate nonlinear regression model is used as the parameter prediction model of industrial products.

[0110]

[0111] Since there is no simple linear relationship between process parameters and production result indicators, in order to avoid the overfitting problem of the multivariate linear regression model, in this embodiment, a multivariate nonlinear regression model is constructed with the first-order term of the target process parameters and the second-order term of the target process parameters. A parameter prediction model is trained using the multivariate nonlinear regression model, which can reduce the overfitting problem of the parameter prediction model, thereby improving the reliability of the maximum indicator value of the production result indicator given by the parameter prediction model and improving the output and quality of industrial products.

[0112] 304. Based on the parameter prediction model, obtain a target parameter value corresponding to the target process parameter when the index value of the production result index is maximum.

[0113] The target parameter value refers to the parameter value corresponding to the target process parameter when the index value of the production result index solved based on the parameter prediction model is the maximum.

[0114] Exemplarily, step 304 may specifically include: obtaining the target process parameter x i The value constraint range of the target process parameter x i The parameter prediction model is solved for the maximum value within the value constraint range to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is the maximum. Specifically, the target process parameter x is defined by taking the production result index in the parameter prediction model as the objective function. i The value constraint range is used as the constraint condition, and MATLAB is used to find the maximum value of the production result index of the parameter prediction model (that is, the solved multivariate nonlinear regression model), as well as the target parameter value corresponding to the target process parameter when the production result index reaches the maximum value.

[0115] Among them, MATLAB is a programming and numerical computing platform that supports data analysis, algorithm development, and modeling.

[0116] By using the target process parameter x i The maximum value of the parameter prediction model is solved within the value constraint range. On the one hand, the data processing amount for solving the maximum value of the production result indicator can be reduced, and the speed of solving the maximum value of the parameter prediction model can be improved; on the other hand, the target parameter value can be restricted to the value range that effectively affects the production result indicator, avoiding the problem that the target parameter value cannot really affect the production result indicator, improving the accuracy of the target parameter value, and then effectively improving the output and quality of process product production.

[0117] (1) In some embodiments, the constraint conditions may be defined according to the artificial experience value range of the target process parameter, and MATLAB may be used to find the maximum value of the production result index of the parameter prediction model (i.e., the solved multivariate nonlinear regression model), as well as the target parameter value corresponding to the target process parameter when the production result index reaches the maximum value. In this case, step 304 may specifically include: obtaining the target process parameter x according to the artificial experience value range of the target process parameter i The value constraint range of the target process parameter x i The parameter prediction model is solved for the maximum value within the value constraint range to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is maximum.

[0118] (2) In some embodiments, in order to improve the maximum prediction accuracy of the production result indicator, a functional relationship between the target process parameter and the production result indicator can be constructed and the target process parameter x can be obtained by using the functional relationship between the target process parameter and the production result indicator. i The value constraint range of is used to determine the target parameter value corresponding to the target process parameter when the production result indicator reaches the maximum value. Figure 5 As shown, at this time, step 304 may specifically include the following steps 3041 to 3044:

[0119] 3041. Based on the target process parameter, establish a target process parameter x i The basic Gaussian function.

[0120] Among them, the target process parameter x i The expression of the basic Gaussian function is shown in formula (7):

[0121]

[0122] In formula (7), f(x i) is the production result index, a is the peak height of the Gaussian function curve, μ is the parameter value corresponding to the peak of the Gaussian function curve, σ is the standard deviation of the Gaussian function, and x i is the i-th target process parameter.

[0123] For example, the target process parameters include P = 4 (temperature parameters x 1 , pressure parameter x 2 , time parameter x 3 , speed parameter x 4 ) as an example, the temperature parameters x can be obtained respectively 1 , pressure parameter x 2 , time parameter x 3 , speed parameter x 4 The basic Gaussian functions are shown in the following formulas (8), (9), (1) and (11):

[0124]

[0125]

[0126]

[0127]

[0128] 3042. The target process parameter x in each group of the historical production data is i The historical parameter values ​​and the historical index values ​​of the production result index values ​​are substituted into the target process parameter x i The target process parameter x is obtained by solving the basic Gaussian function of i The solved Gaussian function.

[0129] 3043. Based on the target process parameter x i The standard deviation of the Gaussian function after solution is used to obtain the target process parameter x i The value constraint range of .

[0130] like Figure 6 As shown, μ represents f(x i ) is the maximum corresponding target process parameter x i The parameter value (i.e. the target process parameter x i The peak of the Gaussian function curve after solution corresponds to the parameter value).

[0131] In some embodiments, Figure 6 As shown, [μ-2σ,μ+2σ] can be used as the target process parameter x i The value constraint range of σ is the target process parameter x iThe standard deviation of the Gaussian function after solution, μ is the target process parameter x i The peaks of the solved Gaussian function curve correspond to the parameter values.

[0132] In other embodiments, Figure 6 As shown, in order to make the target process parameter x have a positive impact on the production result index i The parameter value of the target process parameter x i The value constraint range of the target parameter value is set to maximize the production result index as much as possible, and to improve the influence of the determined target parameter value on the production result index. [μ-3σ,μ+3σ] can also be used as the target process parameter x i The value constraint range of σ is the target process parameter x i The standard deviation of the Gaussian function after solution, μ is the target process parameter x i The peak of the Gaussian function curve after solution corresponds to the parameter value. i The value range of the target process parameter x i The target process parameter x can be improved by i The probability that there is an optimal parameter value within the value constraint range of , thereby increasing the probability that the determined target parameter value will have the maximum impact on the production result indicator.

[0133] For example, the target process parameters include four parameters (temperature parameter x 1 , pressure parameter x 2 , time parameter x 3 , speed parameter x 4 )” as an example, we can use the temperature parameter x 1 The standard deviation σ of the Gaussian function after solution 1 , and the temperature parameter x 1 The peak of the Gaussian function curve after solving corresponds to the parameter value μ 1 , and [μ 1 -3σ 1 ,μ 1 +3σ 1 ] as the temperature parameter x 1 Similarly, according to the pressure parameter x 2 The standard deviation σ of the Gaussian function after solution 2 , and the pressure parameter x 2 The peak of the Gaussian function curve after solving corresponds to the parameter value μ 2 , and [μ 2 -3σ 2 ,μ 2 +3σ 2 ] as the pressure parameter x 2Similarly, according to the time parameter x 3 The standard deviation σ of the Gaussian function after solution 3 , and the time parameter x 3 The peak of the Gaussian function curve after solving corresponds to the parameter value μ 3 , and [μ 3 -3σ 3 ,μ 3 +3σ 3 ] as the time parameter x 3 Similarly, according to the speed parameter x 4 The standard deviation σ of the Gaussian function after solution 4 , and the speed parameter x 4 The peak of the Gaussian function curve after solving corresponds to the parameter value μ 4 , and [μ 4 -3σ 4 ,μ 4 +3σ 4 ] as the speed parameter x 4 The value constraint range of .

[0134] 3044. Based on the target process parameter x i The parameter prediction model is solved for the maximum value within the value constraint range to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is maximum.

[0135] For ease of understanding, the above step 3043 is continued, for example, "the target process parameters include 4 (temperature parameters x 1 , pressure parameter x 2 , time parameter x 3 , speed parameter x 4 )” as an example, taking the production result index in the maximization parameter prediction model as the objective function, the temperature parameter x 1 The value constraint range [μ 1 -3σ 1 ,μ 1 +3σ 1 ], pressure parameter x 2 The value constraint range [μ 2 -3σ 2 ,μ 2 +3σ 2 ], time parameter x 3 The value constraint range [μ 3 -3σ 3 ,μ 3 +3σ 3 ], speed parameter x 4 The value constraint range [μ 4 -3σ4 ,μ 4 +3σ 4 ] as constraints, use MATLAB to find the maximum value of the production result index of the parameter prediction model (i.e. the solved multivariate nonlinear regression model), as well as the target parameter value corresponding to the target process parameter when the production result index reaches the maximum value.

[0136] First, as Figure 6 As shown, due to the target process parameter x i Too large or too small will affect the production results and reduce the index, so the target process parameter x i Only within a certain range of values ​​(such as the target process parameter x i The target process parameter x is constructed by taking the Gaussian function relationship between the target process parameter and the production result indicator as a priori condition. i The basic Gaussian function of the target process parameter x i The standard deviation of the Gaussian function after solution is used to obtain the target process parameter x i The value constraint range is used to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is maximum, which can increase the probability that the determined target parameter value will maximize the impact on the production result index.

[0137] Secondly, by combining a single target process parameter x i Gaussian function and multiple target process parameters x i The constructed multivariate nonlinear regression model can not only learn the influence of a single target process parameter itself on the production result index through Gaussian function, but also fit the influence of the combination of multiple target process parameters on the production result index through fitting. i Gaussian function and multiple target process parameters x i The constructed multivariate nonlinear regression model obtains the target parameter value corresponding to the target process parameter when the index value of the production result index is the largest. The target process parameter is adjusted based on the target parameter value, which can maximize the production result index and improve the output and quality of industrial products.

[0138] 305. During the production process of the industrial product, adjust the target process parameter based on the target parameter value.

[0139] Specifically, in the production process of industrial products, the current parameter value of the target process parameter is compared with the target parameter value to see whether they are equal. If the current parameter value of the target process parameter is not equal to the target parameter value, the target process parameter is adjusted to the target parameter value.

[0140] For example, the target process parameters include temperature parameters, pressure parameters, time parameters and speed parameters, and the current parameter value of the temperature parameter can be detected by the temperature sensor; if the current parameter value of the temperature parameter is not equal to the target parameter value corresponding to the temperature parameter, the temperature parameter is adjusted to the target parameter value corresponding to the temperature parameter; if the current parameter value of the temperature parameter is equal to the target parameter value corresponding to the temperature parameter, the temperature parameter is not adjusted.

[0141] Similarly, the current parameter value of the pressure parameter is detected by the pressure sensor; if the current parameter value of the pressure parameter is not equal to the target parameter value corresponding to the pressure parameter, the pressure parameter is adjusted to the target parameter value corresponding to the pressure parameter; if the current parameter value of the pressure parameter is equal to the target parameter value corresponding to the pressure parameter, the pressure parameter is not adjusted.

[0142] Similarly, the current parameter value of the time parameter is collected by the timer; if the current parameter value of the time parameter is not equal to the target parameter value corresponding to the time parameter, the time parameter is adjusted to the target parameter value corresponding to the time parameter; if the current parameter value of the time parameter is equal to the target parameter value corresponding to the time parameter, the time parameter is not adjusted.

[0143] Similarly, the current parameter value of the speed parameter is detected by the speed sensor; if the current parameter value of the speed parameter is not equal to the target parameter value corresponding to the speed parameter, the speed parameter is adjusted to the target parameter value corresponding to the speed parameter. If the current parameter value of the speed parameter is equal to the target parameter value corresponding to the speed parameter, the speed parameter is not adjusted.

[0144] From the above content, it can be seen that by obtaining the historical production data of industrial products; determining the target process parameters related to the production result indicators; generating a parameter prediction model for industrial products based on the target process parameters and multiple sets of historical production data; based on the parameter prediction model, obtaining the target parameter value corresponding to the target process parameter when the indicator value of the production result indicator is maximum; in the production process of industrial products, adjusting the target process parameters based on the target parameter value, thereby finding the target process parameters that affect the production result indicators and the control range of the target process parameters, which are used to control the process parameters of industrial production, avoiding direct reliance on experience to set the parameter values ​​of the process parameters, which makes it difficult to maximize the production result indicators, thereby improving the quality and output of industrial production.

[0145] A person of ordinary skill in the art can understand that all or part of the steps in the above-mentioned industrial production process parameter adjustment method can be completed by instructions, or by controlling related hardware through instructions. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0146] To this end, an embodiment of the present application provides a computer-readable storage medium, in which multiple computer programs are stored. The computer programs can be loaded by a processor to execute any industrial production process parameter adjustment method provided in the embodiment of the present application.

[0147] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0148] In the above-mentioned adjustment device and computer-readable storage medium embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process and beneficial effects of the above-described computer-readable storage medium, adjustment device and its corresponding units can refer to the description of the process parameter adjustment method for industrial production in the above embodiment, and will not be repeated here.

[0149] The above is a detailed introduction to an industrial production process parameter adjustment method, adjustment equipment and computer-readable storage medium provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for adjusting process parameters of industrial production, characterized in that: The method is applied to a regulating device in an industrial production system, wherein the industrial production system includes an operating device, a regulating device, and a parameter detection device, wherein the parameter detection device includes at least two of a temperature sensor, a pressure sensor, a timer, and a speed sensor, and the method includes: Acquire multiple groups of historical production data of industrial products, wherein each group of historical production data includes historical parameter values ​​of each process parameter and historical indicator values ​​of production result indicators, wherein the production result indicators include at least one of a yield indicator and a quality indicator; Determine, from each of the process parameters, a process parameter related to the production result indicator as a target process parameter, wherein the target process parameter includes at least two of a temperature parameter, a pressure parameter, a time parameter, and a speed parameter, the temperature parameter is detected by the temperature sensor, the pressure parameter is detected by the pressure sensor, the time parameter is collected by the timer, and the speed parameter is detected by the speed sensor; Based on the target process parameters and the multiple groups of historical production data, a parameter prediction model of the industrial product is generated, wherein the parameter prediction model is used to indicate the correlation relationship between the target process parameters and the production result indicators; Based on the parameter prediction model, obtaining a target parameter value corresponding to the target process parameter when the index value of the production result index is maximum; In the industrial product production process, the target process parameter is adjusted based on the target parameter value; wherein the target process parameter includes P, and the parameter prediction model is generated based on the historical parameter value of the target process parameter and the historical indicator value, including: Based on the target process parameters, a preliminary multivariate nonlinear regression model is established, wherein the expression of the multivariate nonlinear regression model is: in, is the predicted value of the production result index, P is the total number of target process parameters, β0 is the regression constant term to be determined, β i is the partial regression coefficient of the first-order term to be determined, b i is the partial regression coefficient of the determined quadratic term, x i is the i-th target process parameter, α is the random error; Based on the expression of the nonlinear regression model, a loss function of the multivariate nonlinear regression model is constructed, wherein the expression of the loss function is: in, is the predicted value of the production result index corresponding to the kth group of historical production data, y k is the true value of the production result indicator corresponding to the kth group of historical production data, and m is the number of groups of historical production data; With the goal of minimizing the loss function, the target process parameter x in each group of the historical production data is i Substitute the historical parameter values ​​and the historical index values ​​of the production result index values ​​into the loss function, and use the least squares method to solve the regression constant term, the linear term partial regression coefficient and the quadratic term partial regression coefficient to obtain the solved multivariate nonlinear regression model; The solved multivariate nonlinear regression model is used as the parameter prediction model.

2. The process parameter adjustment method for industrial production according to claim 1, characterized in that: The step of obtaining, based on the parameter prediction model, a target parameter value corresponding to the target process parameter when the index value of the production result index is maximum, comprises: Get the target process parameter x i The value constraint range of Based on the target process parameter x i The parameter prediction model is solved for the maximum value within the value constraint range to obtain the target parameter value corresponding to the target process parameter when the index value of the production result index is maximum.

3. The process parameter adjustment method for industrial production according to claim 2, characterized in that: The target process parameter x is obtained. i The value constraints of include: Based on the target process parameters, establish the target process parameters x i A basic Gaussian function, where the target process parameter x i The expression of the basic Gaussian function is: Among them, f(x i ) is the production result index, a is the peak height of the Gaussian function curve, μ is the parameter value corresponding to the peak of the Gaussian function curve, σ is the standard deviation of the Gaussian function, and x i is the i-th target process parameter; The target process parameter x in each group of historical production data i The historical parameter values ​​and the historical index values ​​of the production result index values ​​are substituted into the target process parameter x i The target process parameter x is obtained by solving the basic Gaussian function of i The solved Gaussian function of ; Based on the target process parameter x i The standard deviation of the Gaussian function after solution is used to obtain the target process parameter x i The value constraint range of .

4. The process parameter adjustment method for industrial production according to claim 3, characterized in that: The target process parameter x i The value constraint range of is [μ-3σ,μ+3σ], where σ is the target process parameter x i The standard deviation of the Gaussian function after solution, μ is the target process parameter x i The peaks of the solved Gaussian function curve correspond to the parameter values.

5. The process parameter adjustment method for industrial production according to claim 1, characterized in that: The P target process parameters are respectively a temperature parameter, a pressure parameter, a time parameter, and a speed parameter; The expression of the multivariate nonlinear regression model is: Among them, x1 is the temperature parameter, x2 is the pressure parameter, x3 is the time parameter, x4 is the speed parameter, β1 is the partial regression coefficient of the temperature parameter to be determined, β2 is the partial regression coefficient of the linear term of the pressure parameter to be determined, β3 is the partial regression coefficient of the linear term of the time parameter to be determined, β4 is the partial regression coefficient of the linear term of the speed parameter to be determined, b1 is the partial regression coefficient of the temperature parameter to be determined, b2 is the partial regression coefficient of the quadratic term of the pressure parameter to be determined, b3 is the partial regression coefficient of the quadratic term of the time parameter to be determined, b4 is the partial regression coefficient of the quadratic term of the speed parameter to be determined, and α is the random error.

6. The process parameter adjustment method for industrial production according to claim 1, characterized in that: The step of determining, from among the process parameters, a process parameter related to the production result indicator as a target process parameter, includes: Traversing multiple groups of the historical production data, dividing each group of the historical production data into positive and negative samples, if the historical index value of the production result index of the current traversal data group is greater than the preset index threshold, the current traversal data group is divided into a positive sample; if the historical index value of the production result index of the current traversal data group is less than or equal to the preset index threshold, the current traversal data group is divided into a negative sample, and the label value of each group of the historical production data is obtained; Based on the label value of each group of the historical production data and the historical parameter value of each process parameter corresponding to each group of the historical production data, the preset Xgboot model is trained to obtain a trained Xgboot model; Calculating the importance score of each of the process parameters using the trained Xgboot model; From the process parameters, process parameters whose importance scores are greater than a preset score threshold are extracted as the target process parameters.

7. The process parameter adjustment method for industrial production according to claim 1, characterized in that: The target process parameters include temperature parameters, pressure parameters, time parameters and speed parameters; In the production process of the industrial product, adjusting the target process parameter based on the target parameter value includes: Detecting a current parameter value of the temperature parameter by the temperature sensor; Detecting a current parameter value of the pressure parameter by the pressure sensor; Collecting the current parameter value of the time parameter by the timer; Detecting a current parameter value of the speed parameter by the speed sensor; If the current parameter value of the temperature parameter is not equal to the target parameter value corresponding to the temperature parameter, adjusting the temperature parameter to the target parameter value corresponding to the temperature parameter; If the current parameter value of the pressure parameter is not equal to the target parameter value corresponding to the pressure parameter, adjusting the pressure parameter to the target parameter value corresponding to the pressure parameter; If the current parameter value of the time parameter is not equal to the target parameter value corresponding to the time parameter, adjusting the time parameter to the target parameter value corresponding to the time parameter; If the current parameter value of the speed parameter is not equal to the target parameter value corresponding to the speed parameter, the speed parameter is adjusted to the target parameter value corresponding to the speed parameter.

8. An adjustment device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method for adjusting process parameters of industrial production according to any one of claims 1 to 7 is executed.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is loaded by a processor to execute the process parameter adjustment method for industrial production according to any one of claims 1 to 7.

Images