Data set construction method and device applied to rare earth iron boron performance prediction model

Abstract

The invention relates to a data set construction method and device applied to a rare earth iron boron performance prediction model. The method comprises the following steps: determining a first intrinsic characteristic of the rare earth iron-boron component according to the number and initial proportion of basic trace elements in the rare earth iron-boron component, predicting by using an initial data set containing the initial proportion and the first intrinsic characteristic to obtain a first predicted value of the rare earth iron-boron performance, and when the first predicted value does not meet a preset performance standard interval, determining the performance of the rare earth iron-boron component according to the first predicted value. Adjusting the number of the basic trace elements and the corresponding initial proportion to obtain the number of updated trace elements and the updating proportion; and based on the number of the updated trace elements and the updating proportion, determining a second intrinsic feature so as to construct an updated data set, and predicting by utilizing the updated data set to obtain that a second predicted value of the rare earth iron boron performance meets a preset performance standard interval. According to the method, the data analysis cost and time for constructing the data set of the rare earth iron boron performance prediction model are saved, and the performance prediction precision of the rare earth iron boron performance prediction model is improved.

Description

Technical Field

[0001] The present application relates to the technical field of machine learning for permanent magnet materials, and in particular to a method and device for constructing a data set for a rare earth iron boron performance prediction model. Background Art

[0002] With the increasing demand for high-performance permanent magnets, key rare earth elements such as Pr, Nd, Dy, and Tb are consumed in large quantities. In order to balance the utilization of rare earth resources and reduce material costs, high-abundance, low-cost non-key rare earth elements such as Ce, La, and Y are considered to be substitutes for the above key rare earth elements. Therefore, the development of permanent magnet materials without key rare earths is an effective way to achieve balanced utilization of rare earth resources.

[0003] Among materials with multiple performance requirements, permanent magnet materials constructed with non-critical rare earth elements such as Ce, La, and Y may cause significant declines in certain properties of permanent magnet materials due to composition adjustments and structural changes. In addition, in multi-component rapid quenching magnets with greater chemical complexity such as La, Ce, and Y, there are too many potential compositions to design that meet the target characteristics. Intuition guided by human intuition or trial and error experiments will have large composition prediction errors and it is difficult to obtain the optimal composition. Existing research has developed some alloys with excellent magnetic properties through trace element substitution, but it is still based on continuous trial and error experiments. Summary of the invention

[0004] Based on this, it is necessary to provide a data set construction method, device, training method of rare earth iron boron performance prediction model, computer equipment and computer-readable storage medium for rare earth iron boron performance prediction model, which can optimize the rare earth iron boron composition and performance more effectively and specifically through machine learning to solve the above technical problems.

[0005] In a first aspect, the present application provides a method for constructing a data set for a rare earth iron boron performance prediction model, the method comprising:

[0006] According to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial ratio, the first intrinsic characteristic of the rare earth iron boron component is determined, and an initial data set is constructed based on the first intrinsic characteristic and each of the initial ratios; wherein the initial data set is input into the rare earth iron boron performance prediction model, and the first predicted value of the rare earth iron boron performance is output;

[0007] When the first predicted value does not meet the preset performance standard interval, based on the first predicted value, the number of the basic trace elements and the corresponding initial ratio are adjusted to obtain the updated number of the trace elements and the corresponding updated ratio;

[0008] Based on the number of updated trace elements and the updated ratio, a second intrinsic characteristic of the rare earth iron boron component is determined, and an updated data set is constructed based on the second intrinsic characteristic and each of the updated ratios; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard range.

[0009] In one embodiment, the adjusting the number of basic trace elements and the corresponding initial ratio to obtain the updated number of trace elements and the corresponding updated ratio includes:

[0010] According to the first predicted value of the rare earth iron boron performance, increasing or decreasing the number of the basic trace elements to obtain the updated trace elements;

[0011] If the updated number of trace elements includes newly added trace elements, the preset occupied sites of the newly added trace elements in the rare earth iron boron are obtained, the preset occupied sites are compared with the occupied sites of other basic trace elements in the composition, and at least one target element is selected from each of the basic trace elements for site replacement to obtain a site replacement result;

[0012] According to the site replacement result, the renewal ratio of each of the renewal trace elements in the rare earth iron boron is calculated.

[0013] In one embodiment, determining the first intrinsic characteristic of the rare earth iron boron component according to the number of basic trace elements in the obtained rare earth iron boron component and the corresponding initial ratio includes:

[0014] Obtaining the intrinsic characteristic factors of each of the basic trace elements; wherein the intrinsic characteristic factors include electronegativity, atomic radius, atomic volume, charge density, number of valence electrons and molar fraction;

[0015] Obtaining an initial ratio of each of the basic trace elements, and obtaining an atomic fraction of each of the basic trace elements according to the initial ratio;

[0016] The first intrinsic characteristic of the rare earth iron boron component is calculated based on the intrinsic characteristic factor and the atomic fraction of each of the basic trace elements.

[0017] In one embodiment, the first intrinsic characteristic of the rare earth iron boron component is calculated according to the intrinsic characteristic factor and the atomic fraction, including:

[0018] Calculating a first atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius;

[0019] Calculating a first molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume;

[0020] Calculating a first electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity;

[0021] Calculating a first mixing entropy of the rare earth iron boron component according to the molar fraction;

[0022] Calculating a first total valence electron number of the rare earth iron boron component according to the valence electron number;

[0023] A first charge density of the rare earth iron boron component is calculated according to the atomic fraction and the electronegativity.

[0024] In one embodiment, constructing an initial data set based on the first intrinsic feature and each of the initial proportions comprises:

[0025] Combining the first atomic radius difference, the first molar volume difference, the first electronegativity difference, the first mixing entropy, the first total number of valence electrons, and the first charge density in any number to obtain at least one random data set;

[0026] The initial proportion of the basic trace elements is added to each of the random data sets to obtain a plurality of the initial data sets.

[0027] In one embodiment, after adding the initial proportion of the basic trace elements to each of the random data sets to obtain a plurality of the initial data sets, the method further comprises:

[0028] If the first predicted value of the rare earth iron boron performance output after the initial data set is input into the rare earth iron boron performance prediction model is overfit or underfit, the initial data set and the first predicted value corresponding to the initial data set are discarded.

[0029] In a second aspect, the present application also provides a device for constructing a data set for a rare earth iron boron performance prediction model, the device comprising:

[0030] An initial data set construction module is used to calculate the first intrinsic characteristics of the rare earth iron boron component according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial proportions, and construct an initial data set based on the first intrinsic characteristics and each of the initial proportions; wherein the initial data set is input into the rare earth iron boron performance prediction model, and the first predicted value of the rare earth iron boron performance is output;

[0031] An element adjustment module, used for adjusting the number of the basic trace elements and the corresponding initial ratio when the first predicted value does not meet the preset performance standard value, to obtain the updated number of the trace elements and the corresponding updated ratio;

[0032] An updated data set construction module is used to calculate a second intrinsic characteristic based on the number of updated trace elements and the updated ratio, and to construct an updated data set based on the second intrinsic characteristic and each of the updated ratios; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard value.

[0033] In a third aspect, the present application also provides a training method for a rare earth iron boron performance prediction model, the method comprising:

[0034] Obtaining an initial data set and an updated data set of rare earth iron boron, and constructing a model training database; the model training database includes a training set and a validation set;

[0035] Set the original model parameters of the rare earth iron boron performance prediction model;

[0036] The training set and the verification set are input into the original rare earth iron boron performance prediction model for training and verification. According to the output predicted value of the rare earth iron boron performance, the original model parameters are adjusted until the mean square error between the predicted value of the rare earth iron boron performance and the preset performance value is less than or equal to the set error threshold and the determination coefficient is greater than or equal to the set coefficient threshold, so as to obtain a fully trained rare earth iron boron performance prediction model.

[0037] In a fourth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps corresponding to the method described in the first aspect when executing the computer program.

[0038] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps corresponding to the method described in the first aspect above are implemented.

[0039] The above-mentioned data set construction method and device, computer equipment, and computer-readable storage medium applied to the rare earth iron boron performance prediction model use the initial data set containing the initial proportion of rare earth iron boron and the first intrinsic characteristic to predict the first predicted value of the rare earth iron boron performance, and infer the number and type of adjustable elements on the rare earth iron boron fixed matrix based on the first predicted value to obtain updated trace elements of the rare earth iron boron and the corresponding updated proportions, thereby achieving improved coercive force and remanence and other properties on the rare earth iron boron fixed matrix, greatly saving the data analysis cost and time of the data set for constructing the rare earth iron boron performance prediction model, and is conducive to improving the performance prediction accuracy of the rare earth iron boron performance prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 An application environment diagram of a data set construction method applied to a rare earth iron boron performance prediction model in one embodiment;

[0042] Figure 2 A schematic flow chart of a method for constructing a data set for a rare earth iron boron performance prediction model in one embodiment;

[0043] Figure 3 is a flow chart of step 202 in one embodiment;

[0044] Figure 4 is a flow chart of step 201 in one embodiment;

[0045] Figure 5 is a flow chart of step 403 in one embodiment;

[0046] Figure 6 is a schematic diagram of the flow chart after step 201 in another embodiment;

[0047] Figure 7 A structural block diagram of a data set construction device for a rare earth iron boron performance prediction model in one embodiment;

[0048] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0051] The data set construction method for the rare earth iron boron performance prediction model provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the server 104 through a network. The data storage system 106 can store data that the server 104 needs to process. The data storage system 106 can be integrated on the server 104, or placed on the cloud or other network servers.

[0052] On server 104, a first intrinsic characteristic of the rare earth iron boron component is determined according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial proportions, and an initial data set is constructed based on the first intrinsic characteristic and each of the initial proportions; wherein the initial data set is input into a rare earth iron boron performance prediction model, and a first predicted value of the rare earth iron boron performance is output; when the first predicted value does not meet a preset performance standard interval, the number of the basic trace elements and the corresponding initial proportions are adjusted based on the first predicted value to obtain the number of updated trace elements and the corresponding updated proportions; based on the number of updated trace elements and the updated proportions, a second intrinsic characteristic of the rare earth iron boron component is determined, and an updated data set is constructed based on the second intrinsic characteristic and each of the updated proportions; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard interval.

[0053] The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and IoT devices, and the IoT devices may be smart TVs, projection devices, etc. The server 104 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0054] First, a machine learning algorithm is used to establish a rare earth iron boron performance prediction model. Since the unique properties of permanent magnetic materials constructed by adding elements such as Ce, La, and Y include saturation coercivity, remanence, maximum magnetic energy product, and saturation magnetization, the predicted value of the rare earth iron boron performance output by the rare earth iron boron performance prediction model includes at least one of the coercivity value, remanence value, maximum magnetic energy product value, and saturation magnetization value.

[0055] When training the rare earth iron boron performance prediction model, it is necessary to explore the relationship between the intrinsic characteristics of the elements in the rare earth iron boron composition and the performance and deduce it nearby in the periodic table, so as to achieve the best trace element addition to improve the corresponding magnetic properties. Therefore, it is necessary to continuously fine-tune and update the data set representing the rare earth iron boron element structure according to the predicted value of the rare earth iron boron performance, so as to determine the rare earth iron boron structure with the best performance.

[0056] In an exemplary embodiment, Figure 2 As shown, a method for constructing a data set for a rare earth iron boron performance prediction model is provided. Figure 1 The server in is taken as an example to illustrate, including the following steps 201 to 203.

[0057] Step 201, determining the first intrinsic characteristic of the rare earth iron boron component according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial proportions, and constructing an initial data set based on the first intrinsic characteristic and each of the initial proportions.

[0058] The initial data set is input into the rare earth iron boron performance prediction model, and the first prediction value of the rare earth iron boron performance is output.

[0059] Specifically, the basic trace elements in rare earth iron boron include lanthanum La, cerium Ce, yttrium Y, iron Fe, boron B, etc. Collect multiple initial proportions of the basic trace elements of rare earth iron boron as the initial proportions of the elements in the initial data set. Solve the first intrinsic characteristic corresponding to each initial proportion, and combine the first intrinsic characteristic with the corresponding initial proportion to obtain the initial data set. Each initial data set is input into the rare earth iron boron performance prediction model constructed based on the random forest gradient algorithm, and the first predicted value of the rare earth iron boron performance is output, wherein the first predicted value includes at least one of the coercive force value, the remanence value, the maximum magnetic energy product value and the saturation magnetization intensity value.

[0060] Step 202: When the first predicted value does not meet the preset performance standard interval, the number of the basic trace elements and the corresponding initial ratio are adjusted based on the first predicted value to obtain the updated number of trace elements and the corresponding updated ratio.

[0061] The preset performance standard intervals include a coercivity standard interval, a remanence standard interval, a maximum energy product standard interval, and a saturation magnetization standard interval. The updated trace elements include at least one of the basic trace elements, and / or other elements added according to the periodic table.

[0062] Specifically, the first predicted value is compared with a preset performance standard interval. If the first predicted value does not meet the preset performance standard interval, the number, type, or initial ratio of the basic trace elements is adjusted according to the size of the first predicted value to obtain the number, type, and corresponding updated ratio of the updated trace elements. If the first predicted value meets the preset performance standard interval, the initial data set corresponding to the first predicted value is stored in the model training database.

[0063] Step 203, based on the updated number of trace elements and the updated ratio, determine the second intrinsic characteristic of the rare earth iron boron component, and construct an updated data set based on the second intrinsic characteristic and each of the updated ratios.

[0064] The updated data set is input into the rare earth iron boron performance prediction model, and the second prediction value obtained by output meets the preset performance standard range.

[0065] Specifically, the second intrinsic characteristic of the rare earth iron boron is recalculated according to the number, type and update ratio of the updated trace elements. Each updated ratio is combined with the corresponding calculated second intrinsic characteristic to obtain an adjusted updated data set. Each updated initial data set is input into the constructed rare earth iron boron performance prediction model, and the second predicted value of the rare earth iron boron performance is output, wherein the second predicted value includes at least one of the coercive force value, the remanence value, the maximum magnetic energy product value and the saturation magnetization intensity value.

[0066] In the above-mentioned method for constructing a data set for a rare earth iron boron performance prediction model, a first predicted value of the rare earth iron boron performance is predicted by utilizing an initial data set including an initial ratio of rare earth iron boron and a first intrinsic characteristic, and the number and type of adjustable elements on a fixed matrix of rare earth iron boron are inferred based on the first predicted value, and updated trace elements of rare earth iron boron and corresponding updated ratios are obtained, thereby achieving improved coercive force, remanence and other properties on a fixed matrix of rare earth iron boron, greatly saving the data analysis cost and time of a data set for constructing a rare earth iron boron performance prediction model, and being beneficial to improving the performance prediction accuracy of the rare earth iron boron performance prediction model.

[0067] In an exemplary embodiment, Figure 3 As shown, step 202 adjusts the number of basic trace elements and the corresponding initial ratio to obtain the updated number of trace elements and the corresponding updated ratio, which specifically includes the following steps 301 to 303.

[0068] Step 301: according to the first predicted value of the rare earth iron boron performance, increase or decrease the number of the basic trace elements to obtain the updated trace elements.

[0069] Step 302: If the updated number of trace elements includes newly added trace elements, the preset occupied sites of the newly added trace elements in the rare earth iron boron are obtained, the preset occupied sites are compared with the occupied sites of other basic trace elements in the composition, and at least one target element is selected from each of the basic trace elements for site replacement to obtain a site replacement result.

[0070] In detail, the newly added trace elements include rare earth elements and non-rare earth elements. When adding non-rare earth elements according to the first predicted value of the rare earth iron boron performance, the preset occupied sites of the non-rare earth elements and the occupied sites of the non-rare earth elements contained in the basic trace elements in the composition are obtained, and the preset occupied sites are compared with the occupied sites of other basic trace elements in the composition. The occupied sites can be sorted according to the proportion of the same points in different non-rare earth elements, and at least one target element can be selected for site replacement. It is also possible to select at least one target element according to the amount of non-rare earth elements contained in the basic trace elements in the composition, and the preset occupied sites of the non-rare earth elements and the occupied sites of the target elements are replaced to obtain the site replacement result when adding non-rare earth elements. For example, when the basic trace elements include La1.5Ce14.9Y0.9 FebalB5.9, according to the first predicted value, it is necessary to add non-rare earth element aluminum Al, and the content of non-rare earth element iron Fe in the basic trace elements is the highest. Therefore, iron Fe is selected as the target element, and the preset occupied sites of aluminum Al replace the iron Fe on the same site. When rare earth elements are added according to the first predicted value of the rare earth iron boron performance, the selection of target elements is the same as the above-mentioned non-rare earth elements.

[0071] Step 303, according to the site replacement result, calculate the update ratio of each of the updated trace elements in the rare earth iron boron.

[0072] In this embodiment, the initial ratio of the basic trace elements of rare earth iron boron is adjusted by inferring different types of newly added trace elements and preset occupied sites, and the updated ratio of the updated trace elements is obtained, thereby achieving the optimal trace element adjustment to improve the corresponding magnetic properties of rare earth iron boron.

[0073] In an exemplary embodiment, Figure 4 As shown, step 201 determines the first intrinsic characteristic of the rare earth iron boron component according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial ratio, which specifically includes the following steps 401 to 403.

[0074] Step 401, obtaining the intrinsic characteristic factors of each of the basic trace elements.

[0075] The intrinsic characteristic factors include electronegativity, atomic radius, atomic volume, charge density, number of valence electrons and molar fraction.

[0076] Step 402, obtaining the initial ratio of each of the basic trace elements, and obtaining the atomic fraction of each of the basic trace elements according to the initial ratio.

[0077] Step 403: Calculate the first intrinsic characteristic of the rare earth iron boron component according to the intrinsic characteristic factor and the atomic fraction of each of the basic trace elements.

[0078] In this embodiment, by obtaining the intrinsic characteristic factor and atomic fraction of each basic trace element, the first intrinsic characteristic corresponding to a certain initial ratio of rare earth iron boron components is calculated, so as to achieve accurate acquisition of the intrinsic characteristics and improve the accuracy and practicality of the data set.

[0079] Among them, Figure 5 As shown, step 403 calculates the first intrinsic characteristic of the rare earth iron boron component according to the intrinsic characteristic factor and the atomic fraction, and specifically includes the following steps 501 to 506.

[0080] Step 501: Calculate the first atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius.

[0081] Step 502: Calculate the first molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume.

[0082] Step 503: Calculate the first electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0083] Step 504: Calculate the first mixing entropy of the rare earth iron boron component according to the molar fraction.

[0084] Step 505: Calculate a first total valence electron number of the rare earth iron boron component according to the valence electron number.

[0085] Step 506: Calculate the first charge density of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0086] In this embodiment, six first intrinsic characteristics corresponding to each initial proportion of the basic trace elements in the rare earth iron boron component are calculated through six intrinsic characteristic factors of the initial proportion.

[0087] Preferably, if Figure 6 As shown, in step 201, an initial data set is constructed based on the first intrinsic feature and each of the initial proportions, which specifically includes the following steps 601 to 602.

[0088] Step 601 , performing any number of combinations of the first atomic radius difference, the first molar volume difference, the first electronegativity difference, the first mixing entropy, the first total number of valence electrons, and the first charge density to obtain at least one random data set.

[0089] Step 602, adding the initial proportion of the basic trace elements to each of the random data sets to obtain a plurality of the initial data sets.

[0090] Specifically, after calculating the six first intrinsic characteristics corresponding to each initial ratio, the six first intrinsic characteristics are combined in any number to obtain a random data set, for example, the random data set can be {first atomic radius difference, first molar volume difference, first electronegativity difference}, or {first atomic radius difference, first electronegativity difference, first mixing entropy, first total valence electron number, first charge density}, etc., which are not fully expanded here. At the same time, the initial ratio of the corresponding basic trace elements is also added to each random data set to obtain multiple initial data sets, which reduces the data analysis cost and time while expanding the data volume of the initial data set.

[0091] Wherein, in step 602, the initial proportion of the basic trace elements is added to each of the random data sets. After obtaining a plurality of the initial data sets, the method includes the following contents:

[0092] If the first predicted value of the rare earth iron boron performance output after the initial data set is input into the rare earth iron boron performance prediction model is overfit or underfit, the initial data set and the first predicted value corresponding to the initial data set are discarded.

[0093] In this embodiment, by discarding the overfitting or underfitting initial data set, the practicality of the data set stored in the model training database is improved, and the effectiveness of the updated trace element number and updated ratio obtained by subsequent adjustment using the first predicted value is improved.

[0094] In an exemplary embodiment, step 203 determines the second intrinsic characteristic of the rare earth iron boron component based on the number of updated trace elements and the updated ratio, and specifically includes the following steps 701 to 703.

[0095] Step 701, obtaining the intrinsic characteristic factor of each updated trace element.

[0096] Step 702, obtaining an update ratio of each of the updated trace elements, and obtaining an atomic fraction of each of the updated trace elements according to the update ratio.

[0097] Step 703: Calculate the second intrinsic characteristic of the rare earth iron boron component according to the intrinsic characteristic factor and atomic fraction of each updated trace element.

[0098] In detail, the intrinsic characteristic factors of the updated trace elements include electronegativity, atomic radius, atomic volume, charge density, number of valence electrons and molar fraction. By obtaining the intrinsic characteristic factor and atomic fraction of each updated trace element, the second intrinsic characteristic corresponding to the updated ratio is calculated to achieve accurate acquisition of the intrinsic characteristics, thereby improving the accuracy and practicality of the data set.

[0099] Among them, step 703 calculates the second intrinsic characteristic of the rare earth iron boron component according to the intrinsic characteristic factor and atomic fraction of each updated trace element, and specifically includes the following steps 801 to 806.

[0100] Step 801, calculating the second atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius.

[0101] Step 802: Calculate the second molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume.

[0102] Step 803: Calculate the second electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0103] Step 804: Calculate the second mixing entropy of the rare earth iron boron component according to the molar fraction.

[0104] Step 805: Calculate a second total valence electron number of the rare earth iron boron component according to the valence electron number.

[0105] Step 806: Calculate the second charge density of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0106] In this embodiment, six second intrinsic characteristics corresponding to each updated ratio in the rare earth iron boron component are calculated through six intrinsic characteristic factors of the updated trace elements in the updated ratio.

[0107] Preferably, in step 203, constructing an updated data set based on the second intrinsic feature and each of the update ratios specifically includes the following steps 901 to 902.

[0108] Step 901, combining the second atomic radius difference, the second molar volume difference, the second electronegativity difference, the second mixing entropy, the second total number of valence electrons, and the second charge density in any number to obtain at least one updated random data set;

[0109] Step 902, adding the initial proportion of the basic trace elements to each of the updated random data sets to obtain a plurality of the updated data sets.

[0110] Wherein, in step 902, after adding the initial proportion of the basic trace elements to each of the updated random data sets to obtain a plurality of the updated data sets, the method further includes the following contents:

[0111] If the second predicted value of the rare earth iron boron performance output after the initial data set is input into the rare earth iron boron performance prediction model is overfitting or underfitting, the updated data set and the second predicted value corresponding to the updated data set are discarded.

[0112] In this embodiment, by combining any number of the six second intrinsic features, the data volume of the updated data set is expanded, the data analysis cost and time are reduced, and the overfitted or underfitted updated data set is discarded, thereby improving the practicality and effectiveness of the data set stored in the model training database.

[0113] In a preferred embodiment, a method for constructing a data set for a rare earth iron boron performance prediction model is provided, which specifically includes the following contents:

[0114] Step 1: Collect multiple initial proportions of basic trace elements of rare earth iron boron, and obtain the intrinsic characteristic factors and atomic fractions of each basic trace element. The intrinsic characteristic factors include electronegativity, atomic radius, atomic volume, charge density, valence electron number and molar fraction.

[0115] Step 2: According to the atomic fraction and the atomic radius, the first atomic radius difference δ1 of the rare earth iron boron component is calculated:

[0116]

[0117] Among them, c in the formula i and r i denote the atomic fraction and atomic radius of the ith element, respectively. It represents the average atomic radius of the rare earth iron boron alloy, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0118] According to the atomic fraction and the atomic volume, the first molar volume difference V1 of the rare earth iron boron component is calculated:

[0119]

[0120] Among them, c in the formula i and d i They represent the atomic fraction and atomic volume of the ith element respectively, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0121] According to the atomic fraction and the electronegativity, the first electronegativity difference ΔX1 of the rare earth iron boron component is calculated:

[0122]

[0123] Among them, X i refers to the Pauling electronegativity, the c in the formula i represents the atomic fraction of the ith element, It represents the average Pauling electronegativity of rare earth iron boron alloy, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0124] According to the molar fraction, the first mixing entropy ΔS of the rare earth iron boron component is calculated. mix 1:

[0125]

[0126] Where R is the gas constant, Q i It refers to the molar fraction of the i-th element, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0127] According to the valence electron number, the first total valence electron number VEC1 of the rare earth iron boron component is calculated:

[0128]

[0129] Among them, C i It represents the atomic percentage of the i-th element, VEC i It refers to the number of valence electrons of the i-th element, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0130] According to the atomic fraction and the electronegativity, the first charge density Δσ1 of the rare earth iron boron component is calculated:

[0131]

[0132] Among them, e i refers to the charge density, c i represents the atomic fraction of the ith element, and N is the total number of elements that make up the rare earth iron boron alloy system.

[0133] Step 3: Substitute the first atomic radius difference δ1, the first molar volume difference V1, the first electronegativity difference ΔX1, and the first mixing entropy ΔS mix 1. The first total valence electron number VEC1 and the first charge density Δσ1 are combined in any number to obtain multiple random data sets. The initial proportions of the basic trace elements are added to the corresponding random data sets to obtain multiple initial data sets.

[0134] Step 4: Input multiple initial data sets into the rare earth iron boron performance prediction model, and output a first prediction value of the rare earth iron boron performance. If the first prediction value is overfitting or underfitting, the corresponding initial data set is discarded. Otherwise, the initial data set corresponding to the first prediction value is retained. Determine whether the retained first prediction value meets the preset performance standard interval.

[0135] Step 5: When the first predicted value does not meet the preset performance standard interval, the number of the basic trace elements is increased or decreased according to the first predicted value. If it is inferred that additional trace elements need to be added, the preset occupied sites of the additional trace elements are obtained, the element with the largest content is selected from the basic trace elements as the target element, the preset occupied sites of the non-rare earth elements are replaced with the occupied sites of the target elements, and the updated ratio of each updated trace element in the rare earth iron boron is calculated according to the site replacement result.

[0136] Step 6: According to the updated proportions of each updated trace element, recalculate the second intrinsic characteristic of rare earth iron boron. The second intrinsic characteristic includes the second atomic radius difference δ2, the second molar volume difference V2, the second electronegativity difference ΔX2, the second mixing entropy ΔS mix 2. The second total valence electron number VEC2 and the second charge density Δσ2. The calculation process of the second intrinsic characteristic is the same as step 2.

[0137] Step 7: Substitute the second atomic radius difference δ2, the second molar volume difference V2, the second electronegativity difference ΔX2, and the second mixing entropy ΔS mix2. The second total valence electron number VEC2 and the second charge density Δσ2 are combined in any number to obtain multiple updated random data sets. The updated proportion of the basic trace elements is added to the corresponding updated random data sets to obtain multiple updated data sets.

[0138] Step 8: Input multiple updated data sets into the rare earth iron boron performance prediction model, and output a second predicted value of the rare earth iron boron performance. If the second predicted value is overfitting or underfitting, the corresponding updated data set is discarded. Otherwise, the updated data set corresponding to the second predicted value is retained.

[0139] Among them, through model fitting, it is found that the difference in atomic radius (δ), molar volume difference (V), electronegativity difference (ΔX), mixing entropy (ΔS mix ), the sum of the number of valence electrons (VEC), these five intrinsic features as random data sets can effectively avoid underfitting and overfitting of model outputs.

[0140] In an exemplary embodiment, a training method for a rare earth iron boron performance prediction model is provided, and the method includes the following steps 1001 to 1004.

[0141] Step 1001, obtain an initial data set and an updated data set of rare earth iron boron, and construct a model training database; the model training database includes a training set and a validation set.

[0142] Specifically, the content of the model training database is randomly sorted, and the fixed data in the shuffled database is taken as the test set, and the remaining data is used as the training set and the validation set. Among them, the training set and the validation set are divided using the ten-fold cross-validation method, and the data set in the database is divided into ten subsets, one subset is used as the validation set, and the remaining nine subsets are used as the training set to complete the division of the training set and the validation set. The data of the training set is used to train the model, and then the error on the test set is used as the generalization error of the final model in dealing with real-world scenarios. Finally, the data of the validation set is used to verify the model.

[0143] Step 1002, setting the original model parameters of the rare earth iron boron performance prediction model.

[0144] The original model parameters include hyperparameters and other parameters. The hyperparameters can be adjusted manually, and the other parameters are adaptively changed with the external input data set and hyperparameters.

[0145] Step 1003, input the training set and the verification set into the original rare earth iron boron performance prediction model for training and verification, and adjust the original model parameters according to the output predicted value of the rare earth iron boron performance until the mean square error between the predicted value of the rare earth iron boron performance and the preset performance value is less than or equal to the set error threshold and the determination coefficient is greater than or equal to the set coefficient threshold, so as to obtain a fully trained rare earth iron boron performance prediction model.

[0146] Specifically, the training set is input into the original rare earth iron boron performance prediction model for training, and the best model is trained by adjusting the value of the hyperparameter according to the predicted value of the rare earth iron boron performance obtained by the output, and then the model is evaluated by random search cross-validation strategy using the validation set, and the root mean square error RMSE and the determination coefficient R are calculated respectively. 2 , the smaller the root mean square error, the better the model performance, and the value of the determination coefficient is between 0 and 1. The larger the value, the stronger the explanatory power of the model.

[0147] In detail, the formula for the root mean square error is as follows:

[0148]

[0149] Where RMSE is the root mean square error, N' is the total number of data sets, and y i is the performance prediction value of the model output for the i-th input data set, y′ i is the preset performance standard value of the i-th input data set.

[0150] The formula for the coefficient of determination is as follows:

[0151]

[0152] In the formula, R 2 is the coefficient of determination, n is the total number of data sets, y i is the performance prediction value of the model output for the i-th input data set, y′ i is the preset performance standard value of the i-th input data set, Yes i The mean value of

[0153] Furthermore, according to the experimental results, the rare earth iron boron performance prediction model obtained by training based on the random forest algorithm is the best model for predicting the coercivity and remanence of rare earth iron boron. The best hyperparameter combination for coercivity and the best hyperparameter combination for remanence are as follows:

[0154] Args Br ={n: 432, md: 10, mss: 9, msl: 4};

[0155] Args Hcj={n:388, md:50, mss:14, msl:1};

[0156] In the formula, Args Br represents the best hyperparameter combination for coercivity, Args Hcj represents the optimal hyperparameter combination of remanence, n represents the number of trees in the random forest, md represents the maximum depth of each tree, mss represents the minimum number of samples required for internal node repartitioning, and msl represents the minimum number of samples required for leaf nodes.

[0157] Fifty thousand data sets containing the initial proportions of basic trace elements were generated through random functions, and five intrinsic eigenvalues ​​were calculated. These ten independent variables were used as the input of the rare earth iron boron performance prediction model. The predicted coercive force and remanence were used as the model outputs. The element proportion with the highest predicted performance was selected, and experiments were carried out in combination with the process. During the experiment, the process was adjusted and three strip-spinning rate processes were used. The magnetic properties under the optimal process were tested respectively, as shown in Table 1 below.

[0158] Table 1

[0159]

[0160] Compare the model predicted values ​​of coercive force and remanence with the preset performance standard values ​​to determine whether the model prediction is accurate. If the model predicted value is less than the preset performance standard value under the optimal process and the difference between the two is not greater than the set value (15% of the preset performance standard value), the model prediction is accurate. At the same time, the experimental data is added to the model training database to obtain an updated database, which is used as the initial database content for the next model training.

[0161] Optionally, if the model prediction value is greater than the preset performance standard value and the difference between the two is greater than the set value (15% of the preset performance standard value), the experiment is repeated. If, after repeating the experiment multiple times, the model prediction value is greater than the preset performance standard value and the difference between the two is greater than the set value, it means that the prediction value output by the model can reach a better level than the preset performance standard value. The experimental data set is added to the initial data set to obtain an updated database, and this is used as the initial database content for the next model training.

[0162] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0163] Based on the same inventive concept, the embodiment of the present application also provides a device for constructing a data set applied to a rare earth iron boron performance prediction model for realizing the method for constructing a data set applied to a rare earth iron boron performance prediction model involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more devices for constructing a data set applied to a rare earth iron boron performance prediction model provided below can be referred to the limitations of the method for constructing a data set applied to a rare earth iron boron performance prediction model above, and will not be repeated here.

[0164] In an exemplary embodiment, Figure 7 As shown, a device for constructing a data set applied to a rare earth iron boron performance prediction model is provided, comprising: an initial data set construction module 71, an element adjustment module 72 and an update data set construction module 73, wherein:

[0165] The initial data set construction module 71 is used to calculate the first intrinsic characteristic of the rare earth iron boron component according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial proportion, and construct an initial data set based on the first intrinsic characteristic and each of the initial proportions; wherein the initial data set is input into the rare earth iron boron performance prediction model, and the first predicted value of the rare earth iron boron performance is output.

[0166] The element adjustment module 72 is used to adjust the number of the basic trace elements and the corresponding initial ratio when the first predicted value does not meet the preset performance standard value, so as to obtain the updated number of the trace elements and the corresponding updated ratio.

[0167] The updated data set construction module 73 is used to calculate the second intrinsic characteristic based on the number of updated trace elements and the updated ratio, and to construct an updated data set based on the second intrinsic characteristic and each updated ratio; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard value.

[0168] In one embodiment, the element adjustment module 72 is also used to: increase or decrease the number of the basic trace elements according to the first predicted value of the rare earth iron boron performance to obtain the updated trace elements; if the updated number of trace elements includes newly added trace elements, obtain the preset occupied sites of the newly added trace elements in the rare earth iron boron, compare the preset occupied sites with the occupied sites of other basic trace elements in the composition, select at least one target element from each of the basic trace elements for site replacement, and obtain a site replacement result; based on the site replacement result, calculate the updated ratio of each of the updated trace elements in the rare earth iron boron.

[0169] In one embodiment, the initial data set construction module 71 is also used to: obtain the intrinsic characteristic factors of each of the basic trace elements; wherein the intrinsic characteristic factors include electronegativity, atomic radius, atomic volume, charge density, number of valence electrons and molar fraction; obtain the initial proportions of each of the basic trace elements, and based on the initial proportions, obtain the atomic fraction of each of the basic trace elements; and calculate the first intrinsic characteristics of the rare earth iron boron component based on the intrinsic characteristic factors and the atomic fractions of each of the basic trace elements.

[0170] In one embodiment, the initial data set construction module 71 is also used to: calculate the first atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius; calculate the first molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume; calculate the first electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity; calculate the first mixing entropy of the rare earth iron boron component according to the molar fraction; calculate the first total valence electron number of the rare earth iron boron component according to the valence electron number; calculate the first charge density of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0171] In one embodiment, the initial data set construction module 71 is also used to: perform any number of combinations of the first atomic radius difference, the first molar volume difference, the first electronegativity difference, the first mixing entropy, the first total number of valence electrons and the first charge density to obtain at least one random data set; and add the initial proportion of the basic trace elements to each of the random data sets to obtain multiple initial data sets.

[0172] In one embodiment, the initial data set construction module 71 is also used to: if after the initial data set is input into the rare earth iron boron performance prediction model, the first predicted value of the rare earth iron boron performance output is overfit or underfit, then the initial data set and the first predicted value corresponding to the initial data set are discarded.

[0173] In one embodiment, the updated data set construction module 73 is also used to: obtain the intrinsic characteristic factor of each of the updated trace elements; obtain the updated ratio of each of the updated trace elements, and obtain the atomic fraction of each of the updated trace elements based on the updated ratio; calculate the second intrinsic characteristic of the rare earth iron boron component based on the intrinsic characteristic factor and atomic fraction of each of the updated trace elements.

[0174] In one embodiment, the update data set construction module 73 is also used to: calculate the second atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius. Calculate the second molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume. Calculate the second electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity. Calculate the second mixing entropy of the rare earth iron boron component according to the molar fraction. Calculate the second total valence electron number of the rare earth iron boron component according to the valence electron number. Calculate the second charge density of the rare earth iron boron component according to the atomic fraction and the electronegativity.

[0175] In one embodiment, the updated data set construction module 73 is also used to: perform any number of combinations of the second atomic radius difference, the second molar volume difference, the second electronegativity difference, the second mixing entropy, the second total number of valence electrons and the second charge density to obtain at least one updated random data set; and add the initial proportion of the basic trace elements to each of the updated random data sets to obtain multiple updated data sets.

[0176] In one embodiment, the updated data set construction module 73 is also used to: if after the initial data set is input into the rare earth iron boron performance prediction model, the second predicted value of the rare earth iron boron performance output is overfit or underfit, then the updated data set and the second predicted value corresponding to the updated data set are discarded.

[0177] Each module in the above-mentioned data set construction device for the rare earth iron boron performance prediction model can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0178] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data of rare earth iron boron trace elements. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for constructing a data set applied to a rare earth iron boron performance prediction model is implemented.

[0179] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0180] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps corresponding to the methods described in the above embodiments are implemented.

[0181] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps corresponding to the methods described in the above embodiments are implemented.

[0182] In one embodiment, a computer program product is provided, including a computer program, which implements the steps corresponding to the methods described in the above embodiments when executed by a processor.

[0183] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0184] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0185] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0186] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a data set for a rare earth iron boron performance prediction model, characterized in that: The method comprises: According to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial ratio, the first intrinsic characteristic of the rare earth iron boron component is determined, and an initial data set is constructed based on the first intrinsic characteristic and each of the initial ratios; wherein the initial data set is input into the rare earth iron boron performance prediction model, and the first predicted value of the rare earth iron boron performance is output; When the first predicted value does not meet the preset performance standard interval, based on the first predicted value, the number of the basic trace elements and the corresponding initial ratio are adjusted to obtain the updated number of the trace elements and the corresponding updated ratio; Based on the number of updated trace elements and the updated ratio, a second intrinsic characteristic of the rare earth iron boron component is determined, and an updated data set is constructed based on the second intrinsic characteristic and each of the updated ratios; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard range.

2. The method for constructing a data set for a rare earth iron boron performance prediction model according to claim 1, characterized in that: The step of adjusting the number of basic trace elements and the corresponding initial ratio to obtain the number of updated trace elements and the corresponding updated ratio comprises: According to the first predicted value of the rare earth iron boron performance, increasing or decreasing the number of the basic trace elements to obtain the updated trace elements; If the updated number of trace elements includes newly added trace elements, the preset occupied sites of the newly added trace elements in the rare earth iron boron are obtained, the preset occupied sites are compared with the occupied sites of other basic trace elements in the composition, and at least one target element is selected from each of the basic trace elements for site replacement to obtain a site replacement result; According to the site replacement result, the renewal ratio of each of the renewal trace elements in the rare earth iron boron is calculated.

3. The method for constructing a data set for a rare earth iron boron performance prediction model according to claim 1, characterized in that: Determining the first intrinsic characteristic of the rare earth iron boron component according to the number of basic trace elements in the rare earth iron boron component and the corresponding initial ratio includes: Obtaining the intrinsic characteristic factors of each of the basic trace elements; wherein the intrinsic characteristic factors include electronegativity, atomic radius, atomic volume, charge density, number of valence electrons and molar fraction; Obtaining an initial ratio of each of the basic trace elements, and obtaining an atomic fraction of each of the basic trace elements according to the initial ratio; The first intrinsic characteristic of the rare earth iron boron component is calculated based on the intrinsic characteristic factor and the atomic fraction of each of the basic trace elements.

4. The method for constructing a data set for a rare earth iron boron performance prediction model according to claim 3, characterized in that: The first intrinsic characteristic of the rare earth iron boron component is calculated according to the intrinsic characteristic factor and the atomic fraction, comprising: Calculating a first atomic radius difference of the rare earth iron boron component according to the atomic fraction and the atomic radius; Calculating a first molar volume difference of the rare earth iron boron component according to the atomic fraction and the atomic volume; Calculating a first electronegativity difference of the rare earth iron boron component according to the atomic fraction and the electronegativity; According to the molar fraction, a first mixing entropy of the rare earth iron boron component is calculated; Calculating a first total valence electron number of the rare earth iron boron component according to the valence electron number; A first charge density of the rare earth iron boron component is calculated according to the atomic fraction and the electronegativity.

5. The method for constructing a data set for a rare earth iron boron performance prediction model according to claim 4, characterized in that: The constructing of the initial data set based on the first intrinsic feature and each of the initial proportions comprises: Combining the first atomic radius difference, the first molar volume difference, the first electronegativity difference, the first mixing entropy, the first total number of valence electrons, and the first charge density in any number to obtain at least one random data set; The initial proportion of the basic trace elements is added to each of the random data sets to obtain a plurality of the initial data sets.

6. The method for constructing a data set for a rare earth iron boron performance prediction model according to claim 5, characterized in that: After adding the initial proportion of the basic trace elements to each of the random data sets to obtain a plurality of the initial data sets, the method further comprises: If the first predicted value of the rare earth iron boron performance output after the initial data set is input into the rare earth iron boron performance prediction model is overfit or underfit, the initial data set and the first predicted value corresponding to the initial data set are discarded.

7. A device for constructing a data set for a rare earth iron boron performance prediction model, characterized in that: The device comprises: An initial data set construction module is used to calculate the first intrinsic characteristics of the rare earth iron boron component according to the number of basic trace elements in the acquired rare earth iron boron component and the corresponding initial proportions, and construct an initial data set based on the first intrinsic characteristics and each of the initial proportions; wherein the initial data set is input into the rare earth iron boron performance prediction model, and the first predicted value of the rare earth iron boron performance is output; An element adjustment module, used for adjusting the number of the basic trace elements and the corresponding initial ratio when the first predicted value does not meet the preset performance standard value, to obtain the updated number of the trace elements and the corresponding updated ratio; An updated data set construction module is used to calculate a second intrinsic characteristic based on the number of updated trace elements and the updated ratio, and to construct an updated data set based on the second intrinsic characteristic and each of the updated ratios; wherein the updated data set is input into the rare earth iron boron performance prediction model, and the second predicted value output meets the preset performance standard value.

8. A training method for a rare earth iron boron performance prediction model, characterized in that: The method comprises: Obtaining an initial data set and an updated data set of rare earth iron boron, and constructing a model training database; the model training database includes a training set and a validation set; Set the original model parameters of the rare earth iron boron performance prediction model; The training set and the verification set are input into the original rare earth iron boron performance prediction model for training and verification. According to the output predicted value of the rare earth iron boron performance, the original model parameters are adjusted until the mean square error between the predicted value of the rare earth iron boron performance and the preset performance value is less than or equal to the set error threshold and the determination coefficient is greater than or equal to the set coefficient threshold, so as to obtain a fully trained rare earth iron boron performance prediction model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images