Heavy rare earth-based alloy magnetocaloric effect prediction method, system, equipment, medium and product

Abstract

The invention discloses a heavy rare earth-based alloy magnetocaloric effect prediction method, system, equipment, medium and product, and relates to the field of magnetocaloric effect materials.The method comprises the steps that feature data of known heavy rare earth-based alloy magnetocaloric effect materials are obtained, and an overall alloy data set is constructed; the characteristic data comprises chemical attribute characteristics including heavy rare earth-based alloy element components and target attribute characteristics including Curie temperature phase transformation points and magnetic entropy change sizes; classifying the overall alloy data set according to the chemical attribute characteristics to obtain a classified data set; according to the classification data set and the overall alloy data set, a heavy rare earth-based alloy magnetocaloric effect prediction model is obtained; the heavy rare earth-based alloy magnetocaloric effect prediction model is used for outputting target attribute features; and the heavy rare earth-based alloy magnetocaloric effect prediction model is used for heavy rare earth-based alloy magnetocaloric effect prediction. The accuracy of the prediction result of the magnetocaloric effect material under the complex classification method can be improved.

Description

Technical Field

[0001] The present application relates to the field of magnetocaloric effect materials, and in particular to a method, system, equipment, medium and product for predicting the magnetocaloric effect of heavy rare earth-based alloys. Background Art

[0002] The growing global demand for refrigeration has posed new challenges to environmental protection and refrigeration technology. Magnetic refrigeration technology, based on the magnetocaloric effect (MCE), has emerged as a new refrigeration technology, evolving from basic scientific research to applied research in recent years. It holds the potential to replace traditional, energy-intensive and highly polluting gas compression refrigeration technology. Low-temperature magnetic refrigeration materials hold significant application value in aerospace, military defense, healthcare, low-temperature physics, and other fields. However, current research on magnetic refrigeration materials relies primarily on experience and scientific intuition to synthesize materials with varying compositions and analyze their magnetic properties and evaluate their magnetocaloric effects through conventional physical property measurements. This approach is often inefficient and requires significant investment in research funding and effort. This approach can be described as a "cooking-style" material development model, where traditional methods such as element doping and substitution are used to modify the primary material to improve its properties, or a trial-and-error approach is used to investigate material properties. In experiments, varying the "primary ingredient" or "seasoning" yields varying results. This approach often consumes significant human, material, time, and financial resources to achieve new discoveries. Compared to the "cook-and-dish" material development model of exploring new materials through continuous trial and error and adjusting experimental parameters, machine learning can accelerate data analysis and new material development. However, current machine learning research is limited to the simulation and prediction of single-structure materials and lacks universal applicability. Furthermore, the phase transition temperature range of materials is mainly concentrated in the medium-high temperature range, while machine learning research on magnetocaloric effect materials in the medium-low temperature range has yet to be developed.

[0003] The data on medium- and low-temperature magnetocaloric effect materials primarily includes numerous alloy compounds composed of heavy rare earth elements and non-rare earth elements. The dataset encompasses numerous heavy rare earth alloys with varying structures and elements. Because the magnetocaloric effect is not an intrinsic physical property of the material and is affected by measurement noise, the factors influencing magnetocaloric effect performance are complex. Therefore, it is difficult to achieve optimal results by integrating all the data for modeling.

[0004] Based on the above problems, there is an urgent need to provide a method for predicting the magnetocaloric effect of heavy rare earth-based alloys, so as to solve the problem of the single machine learning data structure of magnetocaloric effect materials in the medium and low temperature zones, and to improve the accuracy of the prediction results of magnetocaloric effect materials under complex classification methods. Summary of the Invention

[0005] The purpose of this application is to provide a method, system, equipment, medium and product for predicting the magnetocaloric effect of heavy rare earth-based alloys, which can solve the problem of the single machine learning data structure of magnetocaloric effect materials in the medium and low temperature zones, and improve the accuracy of the prediction results of magnetocaloric effect materials under complex classification methods.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a method for predicting the magnetocaloric effect of a heavy rare earth-based alloy, the method comprising:

[0008] Acquire characteristic data of known heavy rare earth alloy magnetocaloric effect materials and construct an overall alloy data set; the characteristic data includes chemical property characteristics including the elemental composition of the heavy rare earth alloy and target property characteristics including the Curie temperature phase transition point and the magnitude of the magnetic entropy change;

[0009] Classify the entire alloy dataset based on chemical property characteristics to obtain a classified dataset;

[0010] A magnetocaloric effect prediction model for a heavy rare earth-based alloy is obtained based on the classification data set and the overall alloy data set; the magnetocaloric effect prediction model for a heavy rare earth-based alloy is used to output target attribute characteristics;

[0011] The magnetocaloric effect prediction model of heavy rare earth based alloys is used to predict the magnetocaloric effect of heavy rare earth based alloys.

[0012] Optionally, the heavy rare earth-based alloy element composition includes: heavy rare earth metal elements, transition metal elements, Group III elements and Group IV elements.

[0013] Optionally, the classified data set includes: a ternary RTX alloy data set and a binary RX / RT alloy data set.

[0014] Optionally, obtaining a prediction model for the magnetocaloric effect of heavy rare earth-based alloys based on the classified data set and the overall alloy data set specifically includes:

[0015] Construct a database based on the classification data set and the overall alloy data set;

[0016] The database is divided into training set and test set by randomly extracting data;

[0017] According to the training set, a prediction model for the magnetocaloric effect of heavy rare earth-based alloys was obtained based on the gradient boosting regression tree model.

[0018] Optionally, the method of obtaining a prediction model for the magnetocaloric effect of heavy rare earth-based alloys based on a gradient boosting regression tree model according to the training set specifically includes:

[0019] Use learning rate and number of estimators as hyperparameters for the gradient boosted regression tree model;

[0020] Use grid search algorithm and cross validation to determine the optimal hyperparameter combination;

[0021] Based on the optimal hyperparameter combination, a prediction model for the magnetocaloric effect of heavy rare earth-based alloys was obtained.

[0022] Optionally, a prediction model for the magnetocaloric effect of heavy rare earth-based alloys is obtained based on the optimal hyperparameter combination, and then the following steps are further included:

[0023] The root mean square error and correlation coefficient of determination were used to evaluate the prediction model of magnetocaloric effect of heavy rare earth based alloys.

[0024] In a second aspect, the present application provides a system for predicting the magnetocaloric effect of a heavy rare earth alloy, the system comprising:

[0025] A data set construction module is used to obtain characteristic data of known heavy rare earth alloy magnetocaloric effect materials and construct an overall alloy data set; the characteristic data includes: chemical property characteristics including the elemental composition of the heavy rare earth alloy and target property characteristics including the Curie temperature phase transition point and the magnitude of the magnetic entropy change;

[0026] A classification data set determination module is used to classify the entire alloy data set based on chemical property characteristics to obtain a classification data set;

[0027] A prediction model determination module is used to obtain a magnetocaloric effect prediction model for heavy rare earth-based alloys based on the classification data set and the overall alloy data set; the magnetocaloric effect prediction model for heavy rare earth-based alloys is used to output target attribute characteristics;

[0028] The prediction module is used to predict the magnetocaloric effect of heavy rare earth based alloys using a magnetocaloric effect prediction model of heavy rare earth based alloys.

[0029] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for predicting the magnetocaloric effect of heavy rare earth-based alloys.

[0030] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for predicting the magnetocaloric effect of heavy rare earth-based alloys.

[0031] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the method for predicting the magnetocaloric effect of heavy rare earth-based alloys when executed by a processor.

[0032] According to the specific embodiments provided in this application, this application has the following technical effects:

[0033] The present application provides a method, system, equipment, medium and product for predicting the magnetocaloric effect of heavy rare earth-based alloys. First, an overall alloy data set is obtained. Then, the overall alloy data set is classified by utilizing chemical property characteristics to obtain a classified data set, thereby achieving refined classification of the data set and increasing the physical correlation of non-intrinsic magnetic entropy change performance. Then, a magnetocaloric effect prediction model for heavy rare earth-based alloys is obtained based on the classified data set and the overall alloy data set. This can improve the prediction ability of the magnetocaloric effect prediction model for heavy rare earth-based alloys for magnetic entropy change performance, solve the problems of the current single structure and insufficient universality of magnetocaloric effect material data sets, thereby reducing the cost of manual trial and error and helping to accelerate the development of low-temperature magnetocaloric effect materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 This is a flow chart of a method for predicting the magnetocaloric effect of a heavy rare earth-based alloy in one embodiment of the present application;

[0036] Figure 2 This is a schematic diagram of the overall process of a method for predicting the magnetocaloric effect of a heavy rare earth-based alloy in one embodiment of the present application;

[0037] Figure 3 This is a schematic diagram of the prediction effect of the Curie temperature phase transition point on the overall alloy data set in one embodiment of the present application;

[0038] Figure 4 Schematic diagram of the predicted effect of the overall alloy data set on the magnitude of magnetic entropy change in one embodiment of the present application;

[0039] Figure 5 Schematic diagram of the prediction effect of the model corresponding to the ternary RTX alloy dataset and the binary RX / RT alloy dataset in one embodiment of the present application. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0041] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0042] In an exemplary embodiment, Figure 1 and Figure 2 As shown, a method for predicting the magnetocaloric effect of a heavy rare earth-based alloy is provided, the method comprising the following S101 to S104.

[0043] S101, obtaining characteristic data of known heavy rare earth alloy magnetocaloric effect materials and constructing an overall alloy data set; the characteristic data includes: chemical property characteristics including the elements of the heavy rare earth alloy and Curie temperature phase transition point (T c ) and the magnitude of magnetic entropy change (-ΔS M ) target property characteristics; the Curie temperature phase transition point of the heavy rare earth-based alloy magnetocaloric effect material from the ordered to paramagnetic state is always selected; the magnetic moment of most heavy rare earth-based alloy magnetocaloric effect materials can reach saturation under a 0-5T magnetic field. In order to predict the effect, the magnetic entropy change value under a large magnetic field of 0-5T is selected.

[0044] As a specific embodiment, characteristic data of known heavy rare earth-based alloy magnetocaloric effect materials are obtained through published SCI literature data.

[0045] In order to ensure the accuracy of the constructed overall alloy data set, the characteristic data in the overall alloy data set are preprocessed, and the preprocessing includes: data deletion; that is, deleting heavy rare earth-based alloy elements whose occurrence frequency is less than a set value;

[0046] As a specific embodiment, 219 materials are retained in heavy rare earth alloying elements, wherein the heavy rare earth alloying elements include: heavy rare earth metal elements R = Gd, Tb, Dy, Ho, Er, Tm), transition metal elements (T = Fe, Co, Ni, Cu, Mn), Group III elements and Group IV elements (X = Al, Si, Ge, Ga). Among them, the above 15 elements are used as 15 features, and the eigenvalue of each feature is the proportion of the corresponding feature in the chemical composition of the alloy. If the feature does not exist, the eigenvalue is zero. Therefore, each data point is a 15-dimensional vector.

[0047] S102, classifying the entire alloy data set based on chemical property characteristics to obtain a classified data set;

[0048] In a specific embodiment, there are 221 sets of overall alloy data sets. In the process of learning the Curie temperature phase transition point of the overall alloy data set, the Curie temperature phase transition point of the collected materials varies from 0 to 350K; -ΔS M The values are mainly concentrated in 5-30J·kg -1 ·K -1 If the entire alloy dataset is used as the research object, the complexity and uncertainty of the prediction results will be greatly increased. Therefore, the entire alloy dataset is classified according to chemical property characteristics to increase the correlation of the data.

[0049] The classified data sets include: a ternary RTX alloy data set and a binary RX / RT alloy data set.

[0050] S103, obtaining a magnetocaloric effect prediction model for heavy rare earth-based alloys based on the classified data set and the overall alloy data set; the magnetocaloric effect prediction model for heavy rare earth-based alloys is used to output target attribute characteristics;

[0051] S103 specifically includes:

[0052] S31, constructing a database based on the classification data set and the overall alloy data set; that is, the database includes the overall alloy data set, the ternary RTX alloy data set, and the binary RX / RT alloy data set.

[0053] In the database, each data set is separated into characteristic targets. One characteristic target is the Curie temperature phase transition point, and the other characteristic target is the magnitude of magnetic entropy change. Figure 3 Part (a) and Figure 4 Part (a) is the separation of the two target attribute features of Curie temperature phase transition point and magnetic entropy change in the entire alloy data set. The target attribute feature is one of the prediction targets.

[0054] S32, randomly extract data to divide the database into a training set and a test set; the test set accounts for 20% of the total data, and the training set accounts for 80%.

[0055] S33, according to the training set, based on the gradient boosted regression tree model (GBRT), a prediction model for the magnetocaloric effect of heavy rare earth-based alloys is obtained.

[0056] S33 specifically includes:

[0057] S331, the learning rate and the number of estimators, which have the greatest impact on the training effect, are used as hyperparameters of the gradient boosting regression tree model; it is estimated that the number represents the number of iterations. A smaller learning rate usually requires a larger number of estimators to achieve better performance. The learning rate and the number of iterations are interdependent hyperparameters.

[0058] S332, using grid search algorithm and cross validation to determine the optimal hyperparameter combination;

[0059] Among them, the hyperparameter optimization settings are selected through grid search and k (k is selected as 5) fold cross validation;

[0060] S333: Based on the optimal hyperparameter combination, a prediction model for the magnetocaloric effect of heavy rare earth alloys is obtained. The training results are based on the comprehensive reflection of the training results of all sub-models. By plotting the model performance (learning rate as a function of the number of estimators), the most stable performance point is found.

[0061] In order to achieve the best training effect of the model, the number of iterations and the learning rate need to be adjusted to the appropriate range; then run all parameter combinations, such as Figure 3 As shown in the figure, all models selected learning rates of 0.1-1 and iterations of 50-140, respectively. 100 combinations of two parameter models were trained, and then a block diagram with colors from light to dark was drawn. The lightest block was selected as the optimal parameter combination. The best model parameters were marked with asterisks. The learning rate and number of estimators were selected as 0.3 and 70 respectively. Finally, the best parameter combination selected was input into the gradient boosting regression algorithm for training.

[0062] S33 and later also include:

[0063] The root mean square error (RMSE) and the coefficient of determination (R 2 )Evaluate the prediction model of magnetocaloric effect of heavy rare earth based alloys.

[0064] Among them, RMSE is sensitive to outliers in the data, so outliers in the data are usually deleted. It is used to measure the size of the prediction error, and the smaller the value, the better. RMSE is used to judge the error between the predicted value and the true value. The formula for the root mean square error is as follows:

[0065]

[0066] Among them, n is the total number of samples, y i is the actual value of the input xi sample, is the predicted value.

[0067] R 2 It is a dimensionless indicator used to measure the model's ability to explain the target variable. Its value range is (-∞, 1]. When R 2 The closer it is to 1, the stronger the model fitting ability is and the stronger its ability to explain the target variable is. 2 The formula is as follows:

[0068]

[0069] in, is the mean of the target variable;

[0070] Root mean square error RMSE and R for the training set 2 To comprehensively evaluate the performance of the regression model, the best model was selected to predict data from an unused heavy rare earth alloy test set. This helped understand the model's generalization ability (i.e., its ability to learn from new data), thereby verifying the model's predictive effectiveness for the magnetocaloric effect properties of the classified heavy rare earth alloy dataset. A scatter plot of actual values against predicted values was drawn based on the best model. Specifically, the actual values of the heavy rare earth alloys in the database were plotted as the horizontal axis, and the predicted values of the machine learning model were plotted as the vertical axis. The closer the point is to the y-axis, the closer the actual and predicted values are.

[0071] S104, predicting the magnetocaloric effect of heavy rare earth based alloys using a magnetocaloric effect prediction model for heavy rare earth based alloys.

[0072] The following examples are used to verify the beneficial effects of this application:

[0073] Example 1: Characteristic data of heavy rare earth alloys are collected from literature, and an overall alloy dataset is established. The overall alloy dataset includes 221 groups, which are then subdivided into a ternary alloy RTX dataset and a binary alloy RX / RT dataset. Each dataset is separated into a training set and a test set. The GBRT machine learning method is then used for model training, model evaluation, and model prediction. Figure 2 shown.

[0074] Example 2: In the process of learning the Curie temperature transition point of the entire alloy data set, the Curie temperature transition point of the collected materials ranges from 0 to 350K. A learning rate of 0.1 to 1 and an iteration number of 50 to 140 are selected. It is found that the best combination is an iteration number of 70 and a learning rate of 0.3. Figure 3 Asterisk mark in part (b); Figure 3 As shown in part (c), the R 2The values reached 0.96 and 0.94 respectively, indicating that the model can achieve good training and prediction for the Curie temperature phase transition point; the RMSE values of the training set and test set reached 16.1 and 27.9 respectively, indicating that there is a large deviation between the predicted value and the actual value. It is well known that the RMSE value is very sensitive to large deviations, so when there are outliers in the data, the results will be greatly affected. Figure 3 As shown in part (a) of Figure 3, since the Curie temperature characteristics of each material type in the overall alloy dataset are very different, fusing them together for modeling and prediction is a complex project in itself.

[0075] Example 3: This application attempts to train and evaluate the maximum magnetic entropy change under 0T-5T magnetic field changes. -ΔS of the entire rare earth-based alloy M The values are mainly concentrated in 5-30J·kg -1 ·K -1 , the best combination of parameter model is 70 iterations and 0.4 learning rate. The final performance of the model is as follows Figure 4 As shown in part (c), it can be seen that R on the training set 2 The value is 0.95, but the R 2 The model's prediction of MCE in the test set is not ideal, indicating that the model cannot fit the data of magnetic entropy changes well; however, the RMSE values on the training set and test set are 2.0 and 5.3, respectively, indicating that there is a small deviation between the predicted value and the actual value.

[0076] Example 4: In order to enhance the correlation and regularity between different material data, the overall alloy data set is subdivided into a ternary RTX alloy data set and a binary RX / RT alloy data set. The same method is used to predict and obtain the following: Figure 5 The results are shown in Figure 2. Among them, the ternary RTX alloy data set obtained very good fitting results. In the training data, R 2 =0.97, in the test data R 2 =0.80. Compared to the fitting effect of the training set data, the performance of the test set is slightly worse, indicating that the model is slightly overfitting. The main reason may be that the training data is relatively small, and the model has relatively strong learning ability, which has learned some rules that only exist in the training set data. In addition, the RMSE value of the ternary alloy is also very small, indicating that the ternary RTX alloy dataset performs very well under the training of the GBRT model. In contrast, the training results of the binary RX / RT alloy dataset are poor, and the R value of the test set is 0. 2The value is only 0.16. Considering that the ternary RTX alloy dataset significantly outperforms the binary RX / RT alloy dataset when the data and the learning model are similar, it's necessary to analyze the reasons from a physical perspective. In the ternary RTX alloy dataset, substitution or mixing of elements may not significantly alter the crystal environment of the material itself. In contrast, the binary RX / RT alloy dataset has a simpler composition, and substitution of an element significantly impacts the crystal environment.

[0077] This application uses a method based on machine learning to predict the magnetocaloric effect of heavy rare earth-based alloys, classifying all heavy rare earth alloy data sets into ternary alloy RTX data sets and binary RX / RT alloy data sets, and successfully predicts two important magnetic properties of magnetocaloric effect materials, namely the magnetocaloric effect and Curie temperature phase transition point, under a complex classification method.

[0078] This application uses chemical properties as the target attributes, and Curie temperature phase transition point and magnetic entropy change as the target attributes. For the first time, a gradient boosted regression tree (GBRT) model has been used to predict magnetocaloric properties in a ternary RTX alloy dataset. This approach addresses the current limitations of magnetocaloric material datasets, which often lack a single material structure and lack universality. It also contributes to the accelerated development of low-temperature magnetocaloric materials.

[0079] Based on the same inventive concept, the present application also provides a system for predicting the magnetocaloric effect of a heavy rare earth alloy, which is used to implement the aforementioned method for predicting the magnetocaloric effect of a heavy rare earth alloy. The solution provided by this system is similar to the solution described in the aforementioned method. Therefore, the specific limitations in one or more embodiments of the system for predicting the magnetocaloric effect of a heavy rare earth alloy provided below can be found in the aforementioned method for predicting the magnetocaloric effect of a heavy rare earth alloy, and will not be further elaborated here.

[0080] In an exemplary embodiment, a system for predicting magnetocaloric effect of a heavy rare earth-based alloy is provided, comprising:

[0081] A data set construction module is used to obtain characteristic data of known heavy rare earth alloy magnetocaloric effect materials and construct an overall alloy data set; the characteristic data includes: chemical property characteristics including the elemental composition of the heavy rare earth alloy and target property characteristics including the Curie temperature phase transition point and the magnitude of the magnetic entropy change;

[0082] A classification data set determination module is used to classify the entire alloy data set based on chemical property characteristics to obtain a classification data set;

[0083] A prediction model determination module is used to obtain a magnetocaloric effect prediction model for heavy rare earth-based alloys based on the classification data set and the overall alloy data set; the magnetocaloric effect prediction model for heavy rare earth-based alloys is used to output target attribute characteristics;

[0084] The prediction module is used to predict the magnetocaloric effect of heavy rare earth based alloys using a magnetocaloric effect prediction model of heavy rare earth based alloys.

[0085] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. 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 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 via a network connection. When the computer program is executed by the processor, a method for predicting the magnetocaloric effect of a heavy rare earth-based alloy is implemented.

[0086] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0087] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0088] 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.

[0089] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented 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 memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may 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 may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0090] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.

[0091] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0092] The technical features of the above embodiments can 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 specification.

[0093] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for predicting the magnetocaloric effect of heavy rare earth alloys, characterized in that: The method for predicting the magnetocaloric effect of heavy rare earth-based alloys comprises: Acquire characteristic data of known heavy rare earth alloy magnetocaloric effect materials and construct an overall alloy data set; the characteristic data includes chemical property characteristics including the elemental composition of the heavy rare earth alloy and target property characteristics including the Curie temperature phase transition point and the magnitude of the magnetic entropy change; Classify the entire alloy dataset based on chemical property characteristics to obtain a classified dataset; A magnetocaloric effect prediction model for a heavy rare earth-based alloy is obtained based on the classification data set and the overall alloy data set; the magnetocaloric effect prediction model for a heavy rare earth-based alloy is used to output target attribute characteristics; The magnetocaloric effect prediction model of heavy rare earth based alloys is used to predict the magnetocaloric effect of heavy rare earth based alloys.

2. The method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to claim 1, wherein: The heavy rare earth-based alloy element components include heavy rare earth metal elements, transition metal elements, group III elements and group IV elements.

3. The method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to claim 1, wherein: The classification data sets include: a ternary RTX alloy data set and a binary RX / RT alloy data set.

4. The method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to claim 1, wherein: The magnetocaloric effect prediction model of heavy rare earth-based alloys is obtained based on the classification data set and the overall alloy data set, specifically including: Construct a database based on the classification data set and the overall alloy data set; The database is divided into training set and test set by randomly extracting data; According to the training set, a prediction model for the magnetocaloric effect of heavy rare earth-based alloys was obtained based on the gradient boosting regression tree model.

5. The method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to claim 4, characterized in that: The magnetocaloric effect prediction model of heavy rare earth alloys is obtained based on the training set and the gradient boosting regression tree model, which specifically includes: Use learning rate and number of estimators as hyperparameters for the gradient boosted regression tree model; Use grid search algorithm and cross validation to determine the optimal hyperparameter combination; Based on the optimal hyperparameter combination, a prediction model for the magnetocaloric effect of heavy rare earth-based alloys was obtained.

6. The method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to claim 5, characterized in that: Based on the optimal hyperparameter combination, a prediction model for the magnetocaloric effect of heavy rare earth alloys is obtained, which also includes: The root mean square error and correlation coefficient of determination were used to evaluate the prediction model of magnetocaloric effect of heavy rare earth based alloys.

7. A system for predicting magnetocaloric effect of heavy rare earth alloys, characterized in that: The heavy rare earth-based alloy magnetocaloric effect prediction system comprises: A data set construction module is used to obtain characteristic data of known heavy rare earth alloy magnetocaloric effect materials and construct an overall alloy data set; the characteristic data includes: chemical property characteristics including the elemental composition of the heavy rare earth alloy and target property characteristics including the Curie temperature phase transition point and the magnitude of the magnetic entropy change; A classification data set determination module is used to classify the entire alloy data set based on chemical property characteristics to obtain a classification data set; A prediction model determination module is used to obtain a magnetocaloric effect prediction model for heavy rare earth-based alloys based on the classification data set and the overall alloy data set; the magnetocaloric effect prediction model for heavy rare earth-based alloys is used to output target attribute characteristics; The prediction module is used to predict the magnetocaloric effect of heavy rare earth based alloys using a magnetocaloric effect prediction model of heavy rare earth based alloys.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to any one of claims 1 to 6 is realized.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the magnetocaloric effect of heavy rare earth-based alloys according to any one of claims 1 to 6 is realized.

Images