CN117272811B

CN117272811B - Iron core process parameter determination method and related equipment thereof

Info

Publication number: CN117272811B
Application number: CN202311242070.3A
Authority: CN
Inventors: 杨泽民; 左思红; 黄学良; 周后勇; 刘新卓; 袁建怀; 郝向博; 邹红; 刘柏林
Original assignee: Chongqing Wangbian Electric Group Corp ltd
Current assignee: Chongqing Wangbian Electric Group Corp ltd
Priority date: 2023-09-25
Filing date: 2023-09-25
Publication date: 2024-07-02
Anticipated expiration: 2043-09-25
Also published as: CN117272811A

Abstract

The application belongs to the technical field of iron core processing, and discloses a method for determining iron core process parameters and related equipment thereof.

Description

Iron core process parameter determination method and related equipment thereof

Technical Field

The application relates to the technical field of iron core processing, in particular to a method for determining iron core process parameters and related equipment thereof.

Background

The selection of proper technological parameters is one of key links in the design process of the transformer iron core, and the improper selection of the technological parameters can lead to the condition that the performance of the iron core is not met or is excessive, wherein the condition that the performance is not met represents that the product is not qualified, and the excessive performance can lead to the increase of production cost. At present, the technological parameters of the iron core are mainly determined manually, and the dependency on the design experience of designers is strong, so that the design efficiency is low and the rationality of the finally determined technological parameters is poor.

Disclosure of Invention

The application aims to provide a method for determining the process parameters of an iron core and related equipment, which can reduce the dependence of the process parameter determination process of the iron core on the design experience of designers and improve the design efficiency and the rationality of the process parameters.

In a first aspect, the present application provides a method for determining a process parameter of an iron core, including the steps of:

A1. dividing the iron core technological parameters into fixed parameters and adjustable parameters;

A2. determining key parameters affecting the no-load loss of the iron core from the adjustable parameters;

A3. Training a neural network based on the fixed parameters and the key parameters in the historical process parameters of the iron core and the corresponding historical no-load loss to obtain a no-load loss estimation model;

A4. acquiring target no-load loss, target fixed parameters and initial key parameters of a target iron core;

A5. according to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by utilizing the no-load loss estimation model to obtain a final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range;

A6. outputting final process parameters including the target fixed parameters and the final key parameters.

By constructing an idle load loss estimation model taking fixed parameters and key parameters of the iron core as input, and utilizing the idle load loss estimation model to adjust the key parameters of the target iron core, the deviation of the idle load loss corresponding to the adjusted key parameters and the target idle load loss is within a preset deviation range, thereby realizing automatic determination of the process parameters, reducing the dependency of the process parameter determination process of the iron core on the design experience of designers, improving the design efficiency, avoiding the condition that the performance of the iron core does not meet the requirement or the performance is excessive, and improving the rationality of the process parameters.

Preferably, the fixed parameters include core type and core overall size.

Preferably, step A2 comprises:

Acquiring historical adjustable parameters and corresponding historical no-load loss of the iron core;

Based on the historical adjustable parameters and the historical no-load loss, constructing a regression model between the adjustable parameters and the no-load loss by adopting a partial least squares PLS method;

And screening key parameters from the adjustable parameters according to the absolute values of regression coefficients of the adjustable parameters in the regression model.

The regression model is constructed by using the partial least square PLS method, the influence of each adjustable parameter on the no-load loss of the iron core can be effectively determined, so that the adjustable parameter with smaller influence is screened out, the adjustable parameter with larger influence is reserved as a key parameter, the accuracy of the output result of the no-load loss estimation model obtained later is guaranteed, the data processing capacity is reduced, and the operation efficiency is improved.

Preferably, the step of screening key parameters from the adjustable parameters according to the absolute values of regression coefficients of the adjustable parameters in the regression model includes: :

and screening out the adjustable parameters with absolute values of the regression coefficients not smaller than a preset coefficient threshold value as key parameters.

Preferably, in step A2, corresponding key parameters are determined for each type of iron core;

in step A3, corresponding no-load loss estimation models are obtained by training for each type of iron cores.

Because the types of the iron cores are various, key parameters of each iron core are different, corresponding key parameters are determined for each iron core type, and corresponding no-load loss estimation models are obtained through training, so that the accuracy of output results of the no-load loss estimation models can be further improved, the corresponding no-load loss estimation models can be selected for carrying out optimization on technological parameters according to the specific types of the target iron cores, and the rationality of the technological parameters can be improved.

Preferably, step A4 comprises:

Acquiring the target no-load loss and the target fixed parameter of the target iron core;

And matching a reference iron core from a reference database according to the target no-load loss and the target fixed parameter of the target iron core, and extracting key parameters of the reference iron core as the initial key parameters.

Preferably, step A5 comprises:

A501. Extracting the iron core type of the target iron core from the target fixed parameters, and determining an empty load loss estimation model corresponding to the target iron core, and marking the empty load loss estimation model as a target empty load loss estimation model;

A502. taking the initial key parameters as alternative key parameters;

A503. Inputting the target fixed parameters and the alternative key parameters of the target iron core into the target no-load loss estimation model to obtain no-load loss output by the target no-load loss estimation model, and recording the no-load loss as first no-load loss;

A504. if the deviation between the first no-load loss and the target no-load loss is within a preset deviation range, the alternative key parameter is used as the final key parameter of the target iron core;

A505. And if the deviation between the first no-load loss and the target no-load loss is not within a preset deviation range, adjusting the alternative key parameters according to a preset rule, and returning to the step A503.

In a second aspect, the present application provides an apparatus for determining a process parameter of an iron core, including:

The division module is used for dividing the iron core technological parameters into fixed parameters and adjustable parameters;

The screening module is used for determining key parameters affecting the no-load loss of the iron core from the adjustable parameters;

The training module is used for training the neural network based on the fixed parameters and the key parameters in the historical technological parameters of the iron core and the corresponding historical no-load loss to obtain a no-load loss estimation model;

the acquisition module is used for acquiring target no-load loss, fixed parameters and initial key parameters of the target iron core;

The adjusting module is used for adjusting the initial key parameters by utilizing the no-load loss estimation model according to the target no-load loss and the target fixed parameters of the target iron core to obtain final key parameters so that the deviation of the no-load loss of the target iron core corresponding to the final key parameters and the target no-load loss is within a preset deviation range;

And the output module is used for outputting final process parameters comprising the target fixed parameters and the final key parameters.

In a third aspect, the present application provides an electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, when executing the computer program, running the steps in the method for determining a process parameter of a core as described above.

In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the core process parameter determination method as described hereinbefore.

The beneficial effects are that: according to the iron core process parameter determination method and the related equipment, the fixed parameters and the key parameters of the iron core are used as input no-load loss estimation models, and the key parameters of the target iron core are adjusted by utilizing the no-load loss estimation models, so that the deviation between the no-load loss corresponding to the adjusted key parameters and the target no-load loss is within the preset deviation range, the automatic determination of the process parameters is realized, the dependence of the process parameter determination process of the iron core on the design experience of design personnel is reduced, the design efficiency is improved, meanwhile, the condition that the performance of the iron core does not meet the requirement or is excessive is avoided, and the rationality of the process parameters is improved.

Drawings

Fig. 1 is a flowchart of a method for determining a process parameter of an iron core according to an embodiment of the present application.

Fig. 2 is a schematic structural diagram of an iron core process parameter determining device according to an embodiment of the present application.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Description of the reference numerals: 1. dividing the module; 2. a screening module; 3. a training module; 4. an acquisition module; 5. an adjustment module; 6. an output module; 301. a processor; 302. a memory; 303. a communication bus.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the application, as presented in the figures, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present application.

It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.

Referring to fig. 1, fig. 1 is a method for determining a process parameter of an iron core according to some embodiments of the present application, including the steps of:

A3. training a neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss to obtain a no-load loss estimation model;

A5. According to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by using the no-load loss estimation model to obtain the final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range;

A6. final process parameters including the target fixed parameters and final key parameters are output.

The fixed parameter is an invariable parameter, and is generally determined by the order requirement (or customer requirement) of the iron core, and mainly comprises the type of the iron core and the total size of the iron core. The core type includes, but is not limited to, E-type core, U-type core, I-type core, and "day" type core. The core overall dimensions mainly include the core overall length, overall width, overall thickness, and width of each portion of the core (e.g., the width of the "daily" core, including the left leg, center leg, right leg, upper yoke, and lower yoke). The adjustable parameters are other adjustable process parameters besides fixed parameters, such as material parameters, silicon steel sheet size (the iron core is formed by stacking silicon steel sheets, the silicon steel sheet size refers to the size of a single silicon steel sheet), the number of single silicon steel sheet stacks (when the iron core is assembled, N silicon steel sheets can be stacked on the iron core together after being stacked and aligned each time, and N is the number of single silicon steel sheet stacks), the number of process holes, the size of process holes, the positions of process holes and the like, but the adjustable parameters are not limited to the above.

In some embodiments, step A2 comprises:

AA201 obtaining the historical adjustable parameters and the corresponding historical no-load loss of the iron core;

AA202 constructing a regression model between the adjustable parameter and the no-load loss by adopting a partial least squares PLS method based on the historical adjustable parameter and the historical no-load loss;

AA203 the key parameters are selected from the adjustable parameters according to the absolute values of regression coefficients of the adjustable parameters in the regression model.

Wherein the historical adjustable parameter and corresponding historical no-load loss of the iron core are the adjustable parameter and corresponding no-load loss of the historically produced iron core (the historical no-load loss is an actual measurement value).

The construction of regression models using partial least squares PLS is prior art and will not be described in detail here. In the regression model, the absolute value of the regression coefficient of each adjustable parameter reflects the influence of the corresponding adjustable parameter on the no-load loss of the iron core, and the larger the absolute value of the regression coefficient is, the larger the influence of the corresponding adjustable parameter on the no-load loss of the iron core is.

In some embodiments, step AA203 comprises:

The preset coefficient threshold can be set according to actual needs, key parameters are screened by simply comparing the absolute value of the regression coefficient with the preset coefficient threshold, the logic is simple, and the operation speed is high.

In other embodiments, step AA203 comprises:

according to the absolute value of the regression coefficient, sorting all the adjustable parameters in a descending order;

selecting an adjustable parameter of M before sequencing as a key parameter; m is a preset positive integer.

M can be set according to actual needs, and the quantity of the adjustable parameters is limited, so that the problem that the data processing quantity is overlarge due to the fact that the quantity of the adjustable parameters is too large is solved.

In fact, in step AA203, the adjustable parameter whose absolute value of the regression coefficient is not less than the preset coefficient threshold may be selected as the candidate key parameter, then it is determined whether the number of candidate key parameters exceeds M, if so, the candidate key parameters are sorted in descending order according to the absolute value of the regression coefficient, and the candidate key parameter M before sorting is selected as the effective key parameter, if not, the candidate key parameter is used as the effective key parameter. Therefore, the excessive data processing capacity can be avoided, and the effective key parameters can be ensured to be the process parameters which have great influence on the no-load loss of the iron core.

In other embodiments, step A2 comprises:

ab201 obtaining the historical adjustable parameters and corresponding historical no-load loss of the iron core;

ab202 taking the historical adjustable parameter as an initial first adjustable parameter;

AB203 based on the current first adjustable parameter and the historical no-load loss, constructing a regression model between the adjustable parameter and the no-load loss by adopting a partial least squares PLS method, and calculating the root mean square error of the regression model under the initial first adjustable parameter through cross validation (the root mean square error of the regression model calculated through cross validation is the prior art and is not described in detail here);

AB204 eliminating the first adjustable parameter with the minimum absolute value of the regression coefficient in the regression model;

Ab205 if the number of the remaining first adjustable parameters is greater than K (K is a preset positive integer, which can be set according to actual needs), returning to step AB203; if the number of the remaining first adjustable parameters is not greater than K, step AB206 is performed;

AB206 selecting the first adjustable parameter corresponding to the regression model with minimum root mean square error as the key parameter.

The set of key parameters selected by the mode is an adjustable parameter set with the largest influence on the no-load loss of the iron core, and the no-load loss estimation model constructed by the adjustable parameter set has high accuracy on no-load loss estimation of the iron core.

The no-load loss estimation model may be a BP neural network model, but is not limited thereto. In step A3, a fixed parameter and a key parameter in a set of historical process parameters of the iron core are used as input data of a sample, a historical no-load loss (which is an actual measurement value) corresponding to the set of historical process parameters is used as a label value of the sample, so that a sample is obtained, a training data set is formed by using samples corresponding to a plurality of sets of historical process parameters, and then a pre-built BP neural network model is trained by using the training data set, so that a no-load loss estimation model is obtained. The obtained no-load loss estimation model takes fixed parameters and key parameters of the iron core as input data, and takes no-load loss of the iron core as output data.

Preferably, in step A2, corresponding key parameters are determined for each type of core (i.e. steps AA201-AA203 are performed for each type of core, or steps AB201-AB206 are performed);

The target no-load loss is the upper limit value of no-load loss which is required to be achieved by the target iron core, the no-load loss of the produced target iron core cannot be higher than the target no-load loss, the target no-load loss is determined by the order requirement of the target iron core, and the target no-load loss can be directly extracted from the order information of the target iron core; for convenience of description, the fixed parameter of the target iron core is referred to as a target fixed parameter, and the target fixed parameter can be directly extracted from order information of the target iron core.

Wherein the initial key parameters of the target core may be set manually, or in some preferred embodiments, step A4 includes:

acquiring target no-load loss and target fixed parameters of a target iron core (for example, extracting from order information of the target iron core);

And matching the reference iron core from a reference database according to the target no-load loss and the target fixed parameter of the target iron core, and extracting the key parameter of the reference iron core as an initial key parameter.

The reference database records actual no-load loss and corresponding process parameters (including fixed parameters and adjustable parameters) of a plurality of historical iron cores (i.e. iron cores produced in a historical manner), when the actual no-load loss and the corresponding process parameters are matched, the historical iron cores with the same iron core type as the target iron cores and absolute value deviation of the actual no-load loss and the target no-load loss within a preset tolerance range (which can be set according to actual needs) can be used as candidate iron cores, then the candidate iron cores with the total iron core size closest to the total iron core size of the target iron cores are selected as reference iron cores, and then key parameters are extracted from the process parameters of the reference iron cores as initial key parameters.

When the candidate iron core with the total iron core size closest to that of the target iron core is selected as the reference iron core, the deviation between the total iron core sizes (for example) of the candidate iron cores and the target iron core can be calculated and recorded as a first deviation, then the square sum of the first deviations corresponding to the candidate iron cores is calculated and used as the deviation coefficient between the candidate iron cores and the target iron core, and the candidate iron core with the smallest deviation coefficient is used as the reference iron core.

The initial key parameters are obtained in a matching mode, so that the initial key parameters are relatively close to the final key parameters of the target iron core, good initial conditions are provided for obtaining the final key parameters, and the efficiency of adjusting the initial key parameters to obtain the final key parameters is improved.

Preferably, step A5 comprises:

A501. extracting the core type of the target core from the target fixed parameters, determining an empty load loss estimation model corresponding to the target core, and marking the model as the target empty load loss estimation model;

A502. taking the initial key parameters as alternative key parameters;

A503. Inputting the target fixed parameters and the alternative key parameters of the target iron core into a target no-load loss estimation model to obtain no-load loss output by the target no-load loss estimation model, and recording the no-load loss as first no-load loss;

A504. if the deviation between the first no-load loss and the target no-load loss is within a preset deviation range, taking the alternative key parameter as the final key parameter of the target iron core;

A505. If the deviation between the first no-load loss and the target no-load loss is not within the preset deviation range, the alternative key parameters are adjusted according to the preset rule, and the step A503 is returned.

The preset deviation range is marked as [ a, b ], a < b <0, a is the lower limit value of the preset deviation range, b is the upper limit value of the preset deviation range, wherein the b value smaller than 0 is properly set so as to have enough redundancy, and further ensure that the actual no-load loss of the target iron core meets the condition of not more than the target no-load loss, and particularly, a and b can be set according to actual needs.

In step a505, the preset rule is:

S1, substituting the current alternative key parameters into a regression model which only retains key parameter items to obtain corresponding no-load loss estimated values, and recording the estimated values as second no-load loss;

s2, calculating a difference value between the second no-load loss and the first no-load loss as a compensation value delta e;

S3, calculating a theoretical no-load loss target value according to the following formula:

e=（a+b）/2+c+△e；

Wherein e is a theoretical no-load loss target value, and c is a target no-load loss of a target iron core; wherein, (a+b)/2+c is the sum of the median of the preset deviation range and the target no-load loss, and represents the median of the allowable no-load loss range of the target core, and the theoretical no-load loss target value e represents the target output value required to be reached by the regression model of only the key parameter item when the output value of the target no-load loss estimation model is made to approach the median;

S4, respectively adjusting each key parameter (except the adjusted key parameters, other key parameters are kept to be original values) in the alternative key parameters, so that the absolute value deviation between the output value of the regression model which only keeps the key parameter items and the theoretical no-load loss target value is not larger than a preset deviation threshold (which can be set according to actual needs), and obtaining at least one group of adjusted whole alternative key parameters;

s5, obtaining adjustment costs of all adjustment results;

S6, reserving a group of adjusted whole candidate key parameters with the minimum adjustment cost, deleting other adjusted whole candidate key parameters, and completing adjustment of the candidate key parameters.

The regression model only retaining key parameter items refers to the regression model corresponding to the key parameters finally determined in the step A2; for example, when step A2 determines key parameters through steps AA201-AA203, the regression model that only retains key parameter items refers to a model obtained by excluding other parameter items than the key parameters from the regression model constructed in step AA 202; when step A2 determines the key parameters through steps AB201-AB206, the regression model that only retains the key parameter items refers to the regression model that has the smallest root mean square error and is selected in step AB 206.

In step S4, assuming that the candidate key parameters are (A0, B0, and C0), A0, B0, and C0 are original values of three candidate key parameters, a set of adjusted candidate key parameters obtained by adjusting A0 is (A1, B0, and C0), a set of adjusted candidate key parameters obtained by adjusting B0 is (A0, B1, and C0), a set of adjusted candidate key parameters obtained by adjusting C0 is (A0, B0, and C1), and A1, B1, and C1 are adjusted candidate key parameters.

In step S4, when each key parameter in the candidate key parameters is adjusted respectively, the adjustment is performed within the adjustable range of each key parameter, the adjustable range of each key parameter is set according to actual needs, if one candidate key parameter cannot enable the absolute value deviation between the output value of the regression model only retaining the key parameter item and the theoretical no-load loss target value to be not greater than a preset deviation threshold value within the adjustable range of each candidate key parameter, the adjusted whole candidate key parameter corresponding to the candidate key parameter is not acquired; if all the alternative key parameters cannot enable the absolute value deviation between the output value of the regression model which only keeps the key parameter item and the theoretical no-load loss target value to be not larger than a preset deviation threshold value in the adjustable range, sequencing the alternative key parameters according to the absolute value of the regression coefficient of each alternative key parameter, sequentially adjusting each alternative key parameter from front to back according to the sequencing result, and enabling the absolute value deviation between the output value of the regression model which only keeps the key parameter item and the theoretical no-load loss target value to be minimum when each alternative key parameter is adjusted until the absolute value deviation between the output value of the regression model which only keeps the key parameter item and the theoretical no-load loss target value is not larger than the preset deviation threshold value or traversing all the alternative key parameters.

For example, assuming that the candidate key parameters are sequentially K, L, M from front to back according to the sorting result, firstly adjusting K in an adjustable range of K, if k=k1, only keeping the absolute value deviation between the output value of the regression model of the key parameter item and the theoretical no-load loss target value to be minimum but still greater than a preset deviation threshold, at this time, on the basis of adjusting K to K1, adjusting L in an adjustable range of L again, if l=l1, only keeping the absolute value deviation between the output value of the regression model of the key parameter item and the theoretical no-load loss target value to be minimum but still greater than the preset deviation threshold, on the basis of adjusting K to K1 and adjusting L to L1, and on the basis of adjusting M to be M in an adjustable range of M again, assuming that when M is in a range of M1-M2 (M1, M2 is a value in an adjustable range of M), the absolute value deviation between the output value of the regression model of the key parameter item and the theoretical no-load loss target value is not greater than the preset deviation threshold, and on the basis of M1 is taken, and at this time, one or more values, for example, M1 and M2 are taken as a final set of key parameters, and the final parameters are obtained: (k 1, l1, m 1) and (k 1, l1, m 2).

The adjustment cost of an adjustment result refers to an increase of the production cost when the production of the target iron core is performed according to the adjusted whole candidate key parameter corresponding to the adjustment result relative to the production cost when the production of the target iron core is performed according to the adjusted whole candidate key parameter.

In some embodiments, in step A6, the output final process parameter may be based on the parameters corresponding to the reference core (i.e., the other process parameters of the target core are made equal to the process parameters corresponding to the reference core) in addition to the target fixed parameter and the final key parameter.

The method for determining the iron core technological parameters divides the iron core technological parameters into fixed parameters and adjustable parameters; determining key parameters affecting the no-load loss of the iron core from the adjustable parameters; training a neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss to obtain a no-load loss estimation model; acquiring target no-load loss, target fixed parameters and initial key parameters of a target iron core; according to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by using the no-load loss estimation model to obtain the final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range; outputting final process parameters including the target fixed parameters and final key parameters; therefore, the dependence of the process parameter determination process of the iron core on the design experience of the designer can be reduced, and the design efficiency and the rationality of the process parameter are improved.

Referring to fig. 2, the present application provides an iron core process parameter determining apparatus, including:

the division module 1 is used for dividing the iron core technological parameters into fixed parameters and adjustable parameters;

a screening module 2, configured to determine key parameters affecting no-load loss of the iron core from the adjustable parameters;

the training module 3 is used for training the neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and the corresponding historical no-load loss to obtain a no-load loss estimation model;

The acquisition module 4 is used for acquiring target no-load loss, fixed parameters and initial key parameters of the target iron core;

The adjusting module 5 is used for adjusting the initial key parameters by utilizing the no-load loss estimation model according to the target no-load loss and the target fixed parameters of the target iron core to obtain final key parameters so that the deviation of the no-load loss of the target iron core corresponding to the final key parameters and the target no-load loss is within a preset deviation range;

and the output module 6 is used for outputting final process parameters comprising the target fixed parameters and final key parameters.

In some embodiments, the screening module 2 performs, when determining key parameters affecting the no-load loss of the core from the adjustable parameters:

In some embodiments, the screening module 2 performs, when screening out key parameters from the adjustable parameters according to the absolute magnitudes of regression coefficients of the adjustable parameters in the regression model:

In other embodiments, the screening module 2 performs when screening key parameters from the adjustable parameters according to the absolute values of the regression coefficients of the adjustable parameters in the regression model:

In practice, when the screening module 2 screens out the key parameters from the adjustable parameters according to the absolute values of the regression coefficients of the adjustable parameters in the regression model, the adjustable parameters with the absolute values of the regression coefficients not smaller than the preset coefficient threshold value can be screened out as candidate key parameters, whether the number of the candidate key parameters exceeds M is judged, if so, the candidate key parameters are sorted in descending order according to the absolute values of the regression coefficients, and the candidate key parameters of M before sorting are selected as effective key parameters, if not, the candidate key parameters are used as effective key parameters. Therefore, the excessive data processing capacity can be avoided, and the effective key parameters can be ensured to be the process parameters which have great influence on the no-load loss of the iron core.

In other embodiments, the screening module 2 performs, when determining, from the adjustable parameters, key parameters affecting the no-load loss of the core:

The no-load loss estimation model may be a BP neural network model, but is not limited thereto. The training module 3 trains the neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss, when an no-load loss estimation model is obtained, takes the fixed parameters and the key parameters in a group of the historical process parameters of the iron core as input data of a sample, takes the historical no-load loss (as an actual measurement value) corresponding to the group of the historical process parameters as a label value of the sample, so as to obtain a sample, forms a training data set by using samples corresponding to a plurality of groups of the historical process parameters, and trains a pre-constructed BP neural network model by using the training data set, thereby obtaining the no-load loss estimation model. The obtained no-load loss estimation model takes fixed parameters and key parameters of the iron core as input data, and takes no-load loss of the iron core as output data.

Preferably, the screening module 2 determines the corresponding key parameters for each type of core (i.e. perform steps AA201-AA203, or steps AB201-AB206 for each type of core);

the training module 3 trains and obtains a corresponding no-load loss estimation model aiming at each type of iron core.

Wherein, the initial key parameters of the target core may be set manually, or in some preferred embodiments, the obtaining module 4 performs, when obtaining the target no-load loss, the fixed parameters, and the initial key parameters of the target core:

Preferably, the adjusting module 5 is configured to adjust the initial key parameter according to the target no-load loss and the target fixed parameter of the target iron core by using the no-load loss estimation model, and when obtaining the final key parameter, perform:

A502. taking the initial key parameters as alternative key parameters;

The preset rules are as follows:

e=（a+b）/2+c+△e；

s5, obtaining adjustment costs of all adjustment results;

The regression model only retaining key parameter items refers to a regression model corresponding to the key parameters finally determined by the screening module 2; for example, when the screening module 2 determines the key parameters through steps AA201 to AA203, the regression model that only retains the key parameter items refers to a model obtained by excluding other parameter items than the key parameters from the regression model constructed in step AA 202; when the screening module 2 determines the key parameters through steps AB201-AB206, the regression model that only retains the key parameter items refers to the regression model that has the smallest root mean square error and is selected in step AB 206.

In some embodiments, the final process parameters output by the output module 6 may be based on parameters corresponding to the reference core (i.e., such that the other process parameters of the target core are equal to the process parameters corresponding to the reference core) in addition to the target fixed parameters and the final critical parameters.

From the above, the iron core process parameter determining device divides the iron core process parameters into fixed parameters and adjustable parameters; determining key parameters affecting the no-load loss of the iron core from the adjustable parameters; training a neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss to obtain a no-load loss estimation model; acquiring target no-load loss, target fixed parameters and initial key parameters of a target iron core; according to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by using the no-load loss estimation model to obtain the final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range; outputting final process parameters including the target fixed parameters and final key parameters; therefore, the dependence of the process parameter determination process of the iron core on the design experience of the designer can be reduced, and the design efficiency and the rationality of the process parameter are improved.

Referring to fig. 3, fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application, where the electronic device includes: processor 301 and memory 302, the processor 301 and memory 302 being interconnected and in communication with each other by a communication bus 303 and/or other form of connection mechanism (not shown), the memory 302 storing a computer program executable by the processor 301, the computer program being executable by the processor 301 when the electronic device is running to perform the core process parameter determination method in any of the alternative implementations of the above embodiments to implement the following functions: dividing the iron core technological parameters into fixed parameters and adjustable parameters; determining key parameters affecting the no-load loss of the iron core from the adjustable parameters; training a neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss to obtain a no-load loss estimation model; acquiring target no-load loss, target fixed parameters and initial key parameters of a target iron core; according to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by using the no-load loss estimation model to obtain the final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range; final process parameters including the target fixed parameters and final key parameters are output.

An embodiment of the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the method for determining a process parameter of an iron core in any of the alternative implementations of the foregoing embodiment, so as to implement the following functions: dividing the iron core technological parameters into fixed parameters and adjustable parameters; determining key parameters affecting the no-load loss of the iron core from the adjustable parameters; training a neural network based on fixed parameters and key parameters in the historical process parameters of the iron core and corresponding historical no-load loss to obtain a no-load loss estimation model; acquiring target no-load loss, target fixed parameters and initial key parameters of a target iron core; according to the target no-load loss and the target fixed parameter of the target iron core, the initial key parameter is adjusted by using the no-load loss estimation model to obtain the final key parameter, so that the deviation of the no-load loss of the target iron core corresponding to the final key parameter and the target no-load loss is within a preset deviation range; final process parameters including the target fixed parameters and final key parameters are output. The computer readable storage medium may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM for short), electrically erasable Programmable Read-Only Memory (ELECTRICALLY ERASABLE PROGRAMMABLE READ-Only Memory, EEPROM for short), erasable Programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM for short), programmable Read-Only Memory (PROM for short), read-Only Memory (ROM for short), magnetic Memory, flash Memory, magnetic disk, or optical disk.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.

Further, the units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Furthermore, functional modules in various embodiments of the present application may be integrated together to form a single portion, or each module may exist alone, or two or more modules may be integrated to form a single portion.

In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and variations will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. The method for determining the technical parameters of the transformer core is used for determining the technical parameters of the transformer core in the design process of the transformer core and is characterized by comprising the following steps:

A1. dividing the iron core technological parameters into fixed parameters and adjustable parameters; the fixed parameters comprise the type of the iron core and the total size of the iron core;

A6. outputting final process parameters including the target fixed parameters and the final key parameters;

The step A2 comprises the following steps:

screening key parameters from the adjustable parameters according to the absolute values of regression coefficients of the adjustable parameters in the regression model;

in the step A2, corresponding key parameters are determined for each type of iron cores;

in the step A3, corresponding no-load loss estimation models are obtained by training each type of iron cores;

step A5 includes:

A502. taking the initial key parameters as alternative key parameters;

A505. If the deviation between the first no-load loss and the target no-load loss is not within the preset deviation range, the alternative key parameters are adjusted according to a preset rule, and the step A503 is returned;

The preset rule is as follows:

e=（a+b）/2+c+△e；

wherein e is a theoretical no-load loss target value, c is a target no-load loss of a target iron core, a is a lower limit value of a preset deviation range, and b is an upper limit value of the preset deviation range;

S4, respectively adjusting each key parameter in the alternative key parameters so that absolute value deviation between the output value of the regression model which only keeps the key parameter item and the theoretical no-load loss target value is not larger than a preset deviation threshold value, and obtaining at least one group of adjusted whole alternative key parameters;

s5, obtaining adjustment costs of all adjustment results;

2. The method of claim 1, wherein the step of screening key parameters from the adjustable parameters according to the absolute values of regression coefficients of the adjustable parameters in the regression model comprises:

3. The method of determining a process parameter of an iron core according to claim 1, wherein step A4 comprises:

4. An iron core process parameter determining device for determining a process parameter of a transformer iron core in a design process of the transformer iron core, comprising:

the division module is used for dividing the iron core technological parameters into fixed parameters and adjustable parameters; the fixed parameters comprise the type of the iron core and the total size of the iron core;

The output module is used for outputting final process parameters comprising the target fixed parameters and the final key parameters;

the screening module performs when determining key parameters affecting the no-load loss of the iron core from the adjustable parameters:

The screening module determines corresponding key parameters for each type of iron cores;

The training module trains each type of iron core to obtain a corresponding no-load loss estimation model;

The adjusting module is used for adjusting the initial key parameters by utilizing the no-load loss estimation model according to the target no-load loss and the target fixed parameters of the target iron core to obtain final key parameters, so that when the no-load loss of the target iron core corresponding to the final key parameters and the deviation of the target no-load loss are within a preset deviation range, the adjusting module is used for executing the following steps:

A502. taking the initial key parameters as alternative key parameters;

The preset rule is as follows:

e=（a+b）/2+c+△e；

s5, obtaining adjustment costs of all adjustment results;

5. An electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, when executing the computer program, running the steps of the method for determining the process parameters of the core of any one of claims 1-3.

6. A computer readable storage medium having stored thereon a computer program, which when executed by a processor performs the steps of the method for determining the process parameters of the core of any of claims 1-3.