Refractory high-entropy alloy oxidation behavior prediction and optimization method and device

Abstract

The invention provides a refractory high-entropy alloy oxidation behavior prediction and optimization method and device, and relates to the technical field of metal materials. Comprising the following steps: acquiring oxidation data of a plurality of refractory high-entropy alloys at different temperatures; the oxidation data comprises an oxidation kinetic constant and an oxidation index; constructing a sample set based on the element content, the physical characteristic information and the oxidation data included in the refractory high-entropy alloy; training a machine learning model by adopting the sample set to obtain a plurality of target prediction models, and outputting oxidation data of any refractory high-entropy alloy component based on the target prediction models; the sample set comprises element content and physical characteristic information as input and oxidation data as output; different target prediction models output different oxidation data; and carrying out interpretability analysis on the target prediction models to obtain key influence factors of each target prediction model so as to determine and optimize the refractory high-entropy alloy components according to the key influence factors. According to the scheme, the oxidation dynamic constant and the oxidation index of the refractory high-entropy alloy can be quickly and accurately predicted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of metal materials, in particular to a refractory high-entropy alloy oxidation behavior prediction and optimization method and device. BACKGROUND

[0002] The refractory high-entropy alloy (RHEAs) material system breaks through the design concept of traditional high-temperature alloys. Compared with the 1200℃ temperature limit of traditional Ni / Co-based high-temperature alloys, through the composition solid solution of multiple high-melting-point refractory metal components, the refractory high-entropy alloy can still maintain excellent strength at 1400℃ high-temperature environment. The refractory high-entropy alloy is considered as an ideal candidate material in the fields of aerospace, energy and nuclear industry due to its excellent high-temperature strength, excellent thermal stability and unique microstructure. However, high-temperature oxidation is one of the key factors restricting the application of refractory high-entropy alloys in high-temperature environments, which directly affects the service life and performance stability. For an equal-atom-ratio high-entropy alloy containing n elements, theoretically, 2n different alloy compositions can be formed, so it is a challenge to be solved to efficiently screen out alloy compositions with excellent oxidation resistance from the vast composition space. n-1

[0003] The existing research on the oxidation behavior of refractory high-entropy alloys mainly relies on traditional experimental methods, theoretical prediction methods based on thermodynamic calculations, semi-quantitative prediction methods based on empirical formulas, and numerical simulation methods based on finite element analysis. However, these methods have significant limitations. Among them, the traditional experimental method is time-consuming and costly, and it is difficult to fully understand the interaction between each component in a complex alloy system. The theoretical prediction method based on thermodynamic calculation has high calculation complexity and limited accuracy, and it is difficult to handle multiple element interactions. The semi-quantitative prediction method based on empirical formula has a narrow application range and insufficient prediction accuracy, and cannot capture nonlinear relationships. The numerical simulation based on finite element analysis has high calculation cost, is difficult to batch predict, and has poor practicality. Therefore, these methods cannot effectively meet the engineering demand for rapid and accurate prediction of the oxidation behavior of refractory high-entropy alloys. SUMMARY

[0004] The present application provides a refractory high-entropy alloy oxidation behavior prediction and optimization method and device, which can quickly, accurately and efficiently predict the high-temperature oxidation behavior of refractory high-entropy alloys and determine the key influencing factors affecting the high-temperature oxidation behavior, providing scientific guidance for optimizing oxidation resistance and significantly improving development efficiency.

[0005] In a first aspect, the present application provides a refractory high-entropy alloy oxidation behavior prediction and optimization method, comprising:

[0006] Obtaining oxidation data of a plurality of refractory high-entropy alloys at different temperatures; wherein the oxidation data includes oxidation kinetic constants and oxidation indexes;​

[0007] constructing a sample set based on the element content, the physical characteristic information and the oxidation data at different temperatures of the refractory high-entropy alloy;

[0008] training a machine learning model using the sample set to obtain a plurality of target prediction models to output oxidation data of any refractory high-entropy alloy composition based on the target prediction models; wherein the sample set comprises element content, physical characteristic information and temperature as input and oxidation data as output; and the oxidation data output by different target prediction models is different;

[0009] performing explainability analysis on the target prediction models to obtain key influence factors of each target prediction model, so as to determine an optimized refractory high-entropy alloy composition according to the key influence factors.

[0010] Optionally, the refractory high-entropy alloy composition comprises at least three elements selected from Al, Si, Ti, V, Cr, Zr, Nb, Mo, Hf, Ta and W.

[0011] Optionally, the physical characteristic information comprises atomic size difference, electronegativity difference, average electronegativity, valence electron concentration, melting point, thermal expansion coefficient, molar volume, mixing entropy, mixing enthalpy, Ω parameter and specific heat capacity.

[0012] Optionally, training a machine learning model using the sample set to obtain a plurality of target prediction models comprises:

[0013] For each type of oxidation data, the following is performed:

[0014] inputting the element content, the physical characteristic information and the temperature in the sample set into the machine learning model to output predicted oxidation data of this type at this temperature;

[0015] comparing the predicted oxidation data of this type with the oxidation data of this type included in the sample set to obtain root mean square error and determination coefficient;

[0016] when the root mean square error and the determination coefficient both satisfy a preset threshold, training is completed to obtain a target prediction model for the oxidation data of this type.

[0017] Optionally, performing explainability analysis on the target prediction models to obtain key influence factors of each target prediction model comprises:

[0018] For each target prediction model, the following is performed:

[0019] generating an interpreter and calculating a feature contribution value of each feature included in each sample using the interpreter;

[0020] For each feature, the following is performed: calculating a feature contribution value mean of the feature based on feature contribution values of all samples including the feature;

[0021] determining the feature corresponding to the feature contribution value mean greater than a preset mean threshold as a key impact factor.

[0022] Optionally, the determining of the optimized refractory high-entropy alloy composition according to the key impact factor comprises:

[0023] determining a contribution impact direction of the key impact factor;

[0024] determining a key element as the key impact factor with a positive contribution impact direction and representing element content;

[0025] calculating an interaction intensity index between any two features, and retaining a feature pair corresponding to an interaction intensity index greater than a preset index threshold;

[0026] determining an optimization strategy according to the key impact factor, the feature pair, and the key element;

[0027] determining an optimized refractory high-entropy alloy composition satisfying a preset requirement based on the optimization strategy.

[0028] In a second aspect, the present application further provides a device for predicting and optimizing oxidation behavior of a refractory high-entropy alloy, comprising:

[0029] an acquisition module configured to acquire oxidation data of a plurality of refractory high-entropy alloys at different temperatures; wherein the oxidation data comprises oxidation kinetic constants and oxidation indexes;

[0030] a sample generation module configured to construct a sample set based on element content, physical feature information, and the oxidation data at different temperatures included in the refractory high-entropy alloy;

[0031] a training and prediction module configured to train a machine learning model using the sample set to obtain a plurality of target prediction models, so as to output oxidation data of any refractory high-entropy alloy composition based on the target prediction models; wherein the sample set comprises element content, physical feature information, and temperature as input, and oxidation data as output; and the oxidation data output by different target prediction models is different;

[0032] an optimization design module configured to perform interpretability analysis on the target prediction models to obtain key impact factors of each target prediction model, so as to determine an optimized refractory high-entropy alloy composition according to the key impact factors.

[0033] In a third aspect, the present application also provides a computing device comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method for predicting and optimizing oxidation behavior of refractory high-entropy alloys according to any one of the preceding aspects.

[0034] In a fourth aspect, the present application also provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed in a computer, causes the computer to perform the method according to any one of the preceding aspects.

[0035] In a fifth aspect, the present application also provides a computer program product comprising computer instructions, wherein the computer instructions, when executed by a processor, implement the steps of the method according to any one of the first aspects.

[0036] The present application provides a method and device for predicting and optimizing oxidation behavior of refractory high-entropy alloys. For each type of refractory high-entropy alloy, the method uses the obtained oxidation data of each refractory high-entropy alloy at different temperatures, the element content and physical characteristic information included in each refractory high-entropy alloy as samples to construct a sample set, and then uses the sample set to train a machine learning model to obtain a target prediction model for each type of oxidation data. By inputting the element content, physical characteristic information and temperature into the target prediction model, the corresponding type of oxidation data can be output, so as to predict the oxidation data of any refractory high-entropy alloy composition. Then, based on the target prediction model, the element content, the physical characteristic information and the oxidation data, an explainability analysis is performed to obtain the key influence factors of each target prediction model, so as to determine the optimized refractory high-entropy alloy composition according to the key influence factors. In this way, by establishing a high-precision target prediction model of oxidation data, potential alloy compositions with high oxidation behavior can be screened out without a large number of experiments, and the key influence factors affecting the target mechanical properties can be determined through explainability analysis, thereby significantly improving the efficiency of optimizing the refractory high-entropy alloy composition. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0038] Figure 1 is a flowchart of a method for predicting and optimizing oxidation behavior of refractory high-entropy alloys provided by an embodiment of the present application;

[0039] Figure 2A feature contribution value-feature importance relationship diagram of an oxidation kinetic constant target prediction model is provided by an embodiment of the present application.

[0040] Figure 3 A feature contribution value-feature importance relationship diagram of an oxidation index target prediction model is provided by an embodiment of the present application.

[0041] Figure 4 An influence mechanism analysis diagram of V element content on oxidation kinetic constant and oxidation index is provided by an embodiment of the present application.

[0042] Figure 5 A hardware architecture diagram of a computing device is provided by an embodiment of the present application.

[0043] Figure 6 A refractory high-entropy alloy oxidation behavior prediction and optimization device structure diagram is provided by an embodiment of the present application. DETAILED DESCRIPTION

[0044] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0045] The concept of the present application will be described below, please refer to Figure 1 The present application provides a refractory high-entropy alloy oxidation behavior prediction and optimization method, comprising:

[0046] In step 100, oxidation data of a plurality of refractory high-entropy alloys at different temperatures is obtained; wherein the oxidation data comprises oxidation kinetic constant and oxidation index.

[0047] In step 102, a sample set is constructed based on element content, physical characteristic information and oxidation data at different temperatures of the refractory high-entropy alloy.

[0048] In step 104, the sample set is used to train a machine learning model to obtain a plurality of target prediction models, so as to output oxidation data of any refractory high-entropy alloy composition based on the target prediction model; wherein the sample set comprises element content, physical characteristic information and temperature as input and oxidation data as output; the oxidation data output by different target prediction models is different.

[0049] In step 106, an explainability analysis is performed on the target prediction model to obtain key influence factors of each target prediction model, so as to determine an optimized refractory high-entropy alloy composition according to the key influence factors.

[0050] In the embodiments of the present application, for each refractory high-entropy alloy, the obtained oxidation data of each refractory high-entropy alloy at different temperatures, the element content and physical characteristic information included in each refractory high-entropy alloy are taken as samples to construct a sample set, and then the sample set is used to train a machine learning model to obtain a target prediction model for each type of oxidation data. By inputting the element content, the physical characteristic information and the temperature into the target prediction model, the corresponding type of oxidation data can be output, so as to predict the oxidation data of any refractory high-entropy alloy composition. Then, based on the target prediction model, the element content, the physical characteristic information and the oxidation data, an explainability analysis is performed to obtain the key influence factors of each target prediction model, so as to determine the optimized refractory high-entropy alloy composition according to the key influence factors. In this way, by establishing a high-precision target prediction model of oxidation data, the alloy composition with potential high oxidation behavior can be screened without a large number of experiments, and the key influence factors affecting the target mechanical properties can be determined through explainability analysis, thereby significantly improving the efficiency of optimization of refractory high-entropy alloy composition.

[0051] The following describes Figure 1 the execution mode of each step shown.

[0052] In a preferred embodiment, the refractory high-entropy alloy composition includes at least three elements among Al, Si, Ti, V, Cr, Zr, Nb, Mo, Hf, Ta and W. For example, the refractory high-entropy alloy composition is composed of Nb, Ti, Zr, Mo and Ta; the refractory high-entropy alloy composition is composed of Al, Si, Ti and Cr; and the like.

[0053] Specifically, the oxidation data of the refractory high-entropy alloy at different temperatures in step 100 includes but is not limited to the experimental data and the literature research. More specifically, the obtained oxidation data is preprocessed, and the oxidation kinetic constant and the oxidation index are obtained by fitting all the oxidation data (including oxidation time, sample weight change within the oxidation time, sample surface area) according to a unified formula: Δm / A = Kp-t^n.

[0054] Wherein, Δm is the sample weight change within the oxidation time (mg); A is the sample surface area (cm 2 ), Kp is the oxidation kinetic constant; t is the oxidation time (h); n is the oxidation index (dimensionless).

[0055] In a preferred embodiment, the physical characteristic information of step 102 includes: atomic size difference, electronegativity difference, average electronegativity, valence electron concentration, melting point, thermal expansion coefficient, molar volume, mixing entropy, mixing enthalpy, Ω parameter and specific heat capacity.

[0056] In particular, the atomic size difference (δ r ) is:

[0057] The electronegativity difference (δ χ ) is:

[0058] The average electronegativity (χ a ) is χ a =∑c i ×χ i

[0059] The valence electron concentration (VEC) is VEC =∑c i ×(VEC) i

[0060] The melting point (Tm) is Tm =∑c i ×(Tm) i

[0061] The thermal expansion coefficient (Tep) is Tep =∑c i ×(Tep) i

[0062] The molar volume (Vm) is Vm =∑c i ×V i

[0063] The mixing entropy (ΔSmix) is ΔSmix =-R×∑c i ×ln(c i )

[0064] The mixing enthalpy (ΔHmix) is:

[0065] The Ω parameter is Ω =Tm×ΔSmix / |ΔHmix|

[0066] The specific heat capacity (Cv) is Cv =∑c i ×(Cv) i

[0067] wherein c i is the concentration (i.e. atomic percentage) of element i, r i is the atomic radius of element i, is the average value of the atomic radius in the refractory high-entropy alloy composition (i.e. the sum of all atomic radii in the refractory high-entropy alloy composition divided by the total number of elements included), χ i is the electronegativity of element i, is the average value of the electronegativity in the refractory high-entropy alloy composition; is the enthalpy between element i and element j; R is the gas constant; (VEC) i is the valence electron concentration of element i; (Tm)i Tep is the melting point of element i; i Cv is the thermal expansion coefficient of element i; i Ci is the specific heat capacity of element i.

[0068] In step 104, the sample set is used to train the machine learning model to obtain a plurality of target prediction models, including:

[0069] For each type of oxidation data, the following is performed:

[0070] The element content, physical characteristic information and temperature in the sample set are input into the machine learning model, and the predicted oxidation data of this type at this temperature is output;

[0071] The predicted oxidation data of this type is compared with the oxidation data of this type included in the sample set to obtain the root mean square error and the determination coefficient;

[0072] When the root mean square error and the determination coefficient both satisfy the preset threshold, the training is completed, and the target prediction model for the oxidation data of this type is obtained.

[0073] It should be noted that before training, a multi-dimensional feature correlation heat map is also constructed based on the physical characteristic information, oxidation data and temperature. Based on the multi-dimensional feature correlation heat map, the highest correlation coefficient between each feature is about 0.8, which ensures the independence of the features. At the same time, these features and the oxidation kinetic constant and the oxidation index present a complex nonlinear relationship, so the prediction model is needed to more accurately predict the oxidation kinetic constant and the oxidation index of each refractory high-entropy alloy.

[0074] In the present application, each prediction model only outputs one type of oxidation data, for example, the oxidation kinetic constant target prediction model is used to output the oxidation kinetic constant, and the oxidation index target prediction model is used to output the oxidation index. In this way, by training the machine learning model with the sample set to obtain the target training model, the oxidation data of any refractory high-entropy alloy composition can be determined without experiments. Subsequently, various types of oxidation data can be analyzed one by one for interpretability to obtain an optimized refractory high-entropy alloy composition.

[0075] It should be noted that, for example, the sample set of step 102 is used to train and optimize the machine learning model, and the 5-fold cross-validation method is used to evaluate the model performance. By comparing with existing regression machine learning algorithms, the CatBoost model with the best performance is selected to predict the oxidation kinetic constant and the oxidation index, respectively, to obtain the optimized oxidation kinetic constant target prediction model and the oxidation index target prediction model, respectively. The root mean square error of the oxidation kinetic constant target prediction model is 0.804 mg·cm-2 h-n, the determination coefficient is 0.9888; the root mean square error of the oxidation index target prediction model is 0.063, and the determination coefficient is 0.9296.

[0076] In step 108, the target prediction model is subjected to an explainability analysis, and the key influencing factors of each target prediction model are obtained, including:

[0077] For each target prediction model, the following is performed:

[0078] An interpreter is generated, and the interpreter is used to calculate the feature contribution value of each feature included in each sample;

[0079] For each feature, the following is performed: the feature contribution value mean of the feature is calculated based on the feature contribution values of all samples including the feature;

[0080] The feature corresponding to the feature contribution value mean greater than the preset mean threshold value is determined as the key influencing factor.

[0081] Specifically, the features include elements, element content, physical characteristic information, and temperature. The feature contribution value mean is determined by the following formula:

[0082]

[0083] wherein, I j is the feature contribution value mean of feature j; n is the number of samples in the sample set; is the feature contribution value of feature j in the i-th sample;

[0084] The feature contribution value means are sorted from high to low, the features with feature contribution value means greater than the preset mean threshold value are retained, and the features are determined as key influencing factors.

[0085] In one preferred embodiment, in step 108, the optimized refractory high-entropy alloy composition is determined according to the key influencing factors, including:

[0086] The contribution influence direction of the key influencing factor is determined;

[0087] The key influencing factor with a positive contribution influence direction and representing element content is determined as a key element;

[0088] The interaction intensity index between any two features is calculated, and the feature pair corresponding to the interaction intensity index greater than the preset index threshold value is retained;

[0089] The optimization strategy is determined according to the key influencing factors, feature pairs, and key elements;

[0090] The optimized refractory high-entropy alloy composition satisfying the preset requirements is determined based on the optimization strategy.

[0091] Specifically, a slope of the key influencing factor is calculated to determine a contribution influence direction of the key influencing factor; wherein the slope is greater than 0, the contribution influence direction of the key influencing factor is a positive direction; otherwise, the contribution influence direction of the key influencing factor is a negative direction; the slope is determined by the following formula:

[0092]

[0093] wherein β j is the slope of the key influencing factor j; x j is the numerical value (i.e. eigenvalue) of the key influencing factor j; cov(φ j , x j ) is the covariance of the eigen-contribution value of the key influencing factor j and the numerical value of the key influencing factor j; var(x j ) is the variance of the numerical value of the key influencing factor j;

[0094] a key influencing factor with a positive contribution influence direction and representing an element content is determined as a key element;

[0095] By detecting the joint effect of any two features, an interaction intensity index between the two features is calculated; the interaction intensity index is determined by the following formula:

[0096]

[0097] Int jk is the interaction intensity index between feature j and feature k; n is the number of samples in the sample set; is the eigen-contribution value of feature j in the i-th sample; is the eigen-contribution value of feature k in the i-th sample; is the eigen-contribution value of feature j and feature k in the i-th sample; j≠k;

[0098] A feature pair with an interaction intensity index greater than a preset index threshold is screened, and the feature pair includes two features with synergistic effect.

[0099] In a more preferred embodiment, step 108 of determining the optimized refractory high-entropy alloy composition according to the key influencing factor further comprises:

[0100] According to the key element, the interaction intensity index, the feature pair, the mean eigen-contribution value of each feature and the slope, a multidimensional evaluation value of any refractory high-entropy alloy is calculated; the multidimensional evaluation value is determined by the following formula:

[0101]

[0102] ΔI jk = I jk - (I j + Ik )

[0103] wherein C is a multi-dimensional evaluation value; σ() represents a Sigmoid function; β j is a slope of feature j; p is a total number of features; λ is an amplification coefficient, λ>1; P represents a set of feature pairs; Int jk is an interaction intensity index between feature j and feature k in the feature pair; ΔI jk is a synergistic effect amount between feature j and feature k; I j , I k , I jk respectively are a feature contribution value mean of feature j, feature k, and feature j and feature k existing simultaneously; q is a number of key elements; c m is a current content of element m; is a key content threshold of element m; δ m is a tolerance coefficient of element m, that is, a window width of the content range interval of element m; it should be noted that the absolute value of the change rate of the feature contribution value is maximum at

[0104] inputting the multi-dimensional evaluation value greater than the preset evaluation threshold to the target prediction model to obtain predicted oxidation data;

[0105] determining an optimized refractory high-entropy alloy composition meeting the preset requirement based on the predicted oxidation data.

[0106] In the embodiments of the present application, the first term of the above formula strengthens the features having a positive directional contribution to the oxidation data, and suppresses the features having a negative directional contribution; the second term is used to reward synergistic effect combinations and punish antagonistic effect combinations; and the third term focuses on giving element compositions deviating from the optimal concentration. In this way, through the interaction of each term, the complex material relationship is converted into an operable optimization path, thereby greatly improving the design efficiency of the refractory high-entropy alloy. The higher the multi-dimensional evaluation value is, the more excellent the oxidation data of the corresponding refractory high-entropy alloy is; and the refractory high-entropy alloy composition with the highest predicted oxidation data is determined as the optimized refractory high-entropy alloy composition.

[0107] In one specific embodiment, as shown in the foregoing examples, taking the optimized oxidation kinetic constant target prediction model and the oxidation index target prediction model obtained by the CatBoost model as examples, first, the feature contribution values of each feature are calculated, Figure 2 shows the feature contribution value honeycomb diagram and the feature importance of the oxidation kinetic constant target prediction model, Figure 3 ​The feature contribution value honeycomb chart and feature importance of the oxidation index target prediction model are shown, and then by screening the first feature and the reserved feature pairs, the key influencing factors are determined as temperature, V element content, electronegativity difference, average electronegativity, and mixing entropy. Among them, temperature is the primary influencing factor, which is consistent with the principle of thermal activation process; V element content has a strong negative effect, forming volatile V2O5 oxide; electronegativity difference affects the local electrochemical potential and accelerates oxidation; average electronegativity affects the stability of the oxide; and mixing entropy has an optimal interval effect. At the same time, through further analysis, the following element influence mechanisms are found: Figure 4 As shown in the figure, for V element: the oxidation kinetic constant Kp value increases exponentially with the V element content, and the n value increases significantly; for Nb element: when the Nb element content is greater than 25 at.%, the "pest" oxidation phenomenon occurs; for Mo element: low content can improve short-term oxidation resistance, and high content promotes linear oxidation; for Al element: when the Al element content is in the range of 15-30 at.%, a dense Al2O3 protective film is formed; for Si element: when the Si element content is in the range of 15-25 at.%, a glassy SiO2 protective layer is formed. Among them, Figure 4 Fig. (a) is an influence mechanism analysis chart of V element content on oxidation kinetic constant, and the vertical coordinate is the feature contribution value of V element content on oxidation kinetic constant; Fig. (b) is an influence mechanism analysis chart of V element content on oxidation index, and the vertical coordinate is the feature contribution value of V element content on oxidation index.

[0108] Based on the above key influencing factors and element influence mechanisms, the following four refractory high-entropy alloys are designed: 50 Cr5Al5Ti 22.5 Si 17.5 , Nb 50 Cr5Al5Ti 17.5 Si 22.5 , Nb 50 Cr5Al5Ta5Ti 17.5 Si 17.5 , Nb 50 Cr5Al5Ta5Ti 12.5 Si 22.5 These four refractory high-entropy alloys are tested at 1100℃ and 1200℃, respectively, to obtain experimental oxidation kinetic constants and experimental oxidation indexes. The oxidation kinetic constant target prediction model and the oxidation index target prediction model are used to predict the predicted oxidation kinetic constants and the predicted oxidation indexes of the four refractory high-entropy alloys at 1100℃ and 1200℃. By comparing the predicted data and the experimental data, it is found that the average prediction error of the predicted oxidation kinetic constants at different temperatures is less than 10%, the average prediction error of the predicted oxidation indexes is less than 5%, and all the predicted data can capture the temperature dependence trend and the element influence effect.

[0109] In the embodiment of the present application, the target prediction model of the oxidation behavior of refractory high-entropy alloys realizes ultra-high precision prediction, with a determination coefficient >0.92; at the same time, the target prediction model has a thorough mechanism understanding and can provide all-round guidance from element selection to parameter adjustment, thereby improving the efficiency of composition optimization of refractory high-entropy alloys. Moreover, compared with the traditional experimental method, the prediction speed is seconds, supporting large-scale screening, and having significant practicality.

[0110] As shown in Figure 5 , Figure 6 , the embodiment of the present application provides a refractory high-entropy alloy oxidation behavior prediction and optimization device. The device embodiment can be realized by software, or realized by hardware or a combination of software and hardware. From the hardware layer, as shown in Figure 5 , it is a hardware architecture diagram of a computing device where the refractory high-entropy alloy oxidation behavior prediction and optimization device provided by the embodiment of the present application is located. In addition to the processor, memory, network interface, and non-volatile memory shown in Figure 3 , the computing device where the device in the embodiment is usually also includes other hardware, such as a forwarding chip responsible for processing packets, etc. Taking the software implementation as an example, as shown in Figure 6 , as a logically meaningful device, it is formed by the CPU of the computing device where it is located reading the corresponding computer program in the non-volatile memory into the memory for running. The refractory high-entropy alloy oxidation behavior prediction and optimization device provided by the embodiment includes:

[0111] The acquisition module 600 is configured to acquire oxidation data of a plurality of refractory high-entropy alloys at different temperatures; wherein the oxidation data includes oxidation kinetic constants and oxidation indexes;

[0112] The sample generation module 602 is configured to construct a sample set based on element contents, physical characteristic information, and oxidation data at different temperatures of the refractory high-entropy alloys;

[0113] The training and prediction module 604 is configured to train a machine learning model using the sample set to obtain a plurality of target prediction models, so as to output oxidation data of any refractory high-entropy alloy composition based on the target prediction models; wherein the sample set includes element contents, physical characteristic information, and temperatures as inputs, and oxidation data as an output; the oxidation data output by different target prediction models is different;

[0114] The optimization design module 606 is configured to perform explainability analysis on the target prediction models to obtain key influence factors of each target prediction model, so as to determine an optimized refractory high-entropy alloy composition according to the key influence factors.

[0115] In some specific embodiments, the acquisition module 600 can be configured to perform the step 100, the sample generation module 602 can be configured to perform the step 102, the training prediction module 604 can be configured to perform the step 104, and the optimization design module 606 can be configured to perform the step 106.

[0116] In some specific embodiments, the refractory high-entropy alloy composition includes at least three elements from Al, Si, Ti, V, Cr, Zr, Nb, Mo, Hf, Ta, and W.

[0117] In some specific embodiments, the physical characteristic information includes atomic size difference, electronegativity difference, average electronegativity, valence electron concentration, melting point, thermal expansion coefficient, molar volume, mixing entropy, mixing enthalpy, Ω parameter, and specific heat capacity.

[0118] In some specific embodiments, the training prediction module 604 is further configured to perform the following operations:

[0119] For each type of oxidation data, the following operations are performed:

[0120] The element content, the physical characteristic information, and the temperature in the sample set are input into the machine learning model, and the predicted oxidation data of the type at the temperature is output;

[0121] The predicted oxidation data of the type is compared with the oxidation data of the type included in the sample set, and the root mean square error and the determination coefficient are obtained;

[0122] When the root mean square error and the determination coefficient both satisfy the preset threshold, the training is completed, and the target prediction model for the type of oxidation data is obtained.

[0123] In some specific embodiments, the optimization design module 606 is further configured to perform the following operations:

[0124] For each target prediction model, the following operations are performed:

[0125] An interpreter is generated, and the interpreter is used to calculate the feature contribution value of each feature included in each sample;

[0126] For each feature, the following operation is performed: the mean value of the feature contribution value of the feature is calculated based on the feature contribution values of all samples including the feature;

[0127] The feature corresponding to the mean value of the feature contribution value greater than the preset mean value threshold is determined as a key impact factor.

[0128] In some specific embodiments, the optimization design module 606 is further configured to perform the following operations:

[0129] The contribution impact direction of the key impact factor is determined;

[0130] The key element is determined as a key element with a positive contribution influence direction and a key influence factor representing the element content;

[0131] An interaction intensity index between any two features is calculated, and a feature pair corresponding to an interaction intensity index greater than a preset index threshold is reserved;

[0132] An optimization strategy is determined according to the key influence factor, the feature pair and the key element;

[0133] An optimized refractory high-entropy alloy composition meeting a preset requirement is determined based on the optimization strategy.

[0134] It can be understood that the structure of the embodiment of the present application does not constitute a specific limitation on the refractory high-entropy alloy oxidation behavior prediction and optimization device. In other embodiments of the present application, a refractory high-entropy alloy oxidation behavior prediction and optimization device can include more or fewer components than the illustration, or combine certain components, or split certain components, or different component arrangements. The illustrated components can be implemented in hardware, software, or a combination of software and hardware.

[0135] The information interaction, execution process and the like between the modules in the above device are based on the same concept as the method embodiments of the present application, and the specific content can be referred to the description in the method embodiments of the present application, which will not be described here.

[0136] The embodiment of the present application also provides a computing device including a memory and a processor, the memory stores a computer program, and the processor implements the refractory high-entropy alloy oxidation behavior prediction and optimization method in any of the embodiments of the present application when executing the computer program.

[0137] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program makes the processor execute the refractory high-entropy alloy oxidation behavior prediction and optimization method in any of the embodiments of the present application when being executed by the processor.

[0138] The embodiment of the present application also provides a computer program product including a computer program, and the processor of the computer device reads the computer program from the computer readable storage medium, and the processor executes the computer program to make the computer device execute the refractory high-entropy alloy oxidation behavior prediction and optimization method in any of the above embodiments.

[0139] Specifically, a system or device equipped with a storage medium can be provided, and the storage medium stores a software program code for realizing the functions of any of the above embodiments, and the computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0140] In this case, the program code read from the storage medium can itself implement the functions of any of the above-described embodiments, and thus the program code and the storage medium that stores the program code can constitute the present application.

[0141] Embodiments of the storage medium that provide the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk such as a CD-ROM, a CD-R, a CD-RW, a DVD-ROM, a DVD-RAM, a DVD- RW, a DVD+RW, a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer through a communication network.

[0142] Furthermore, it will be appreciated that, besides being executed by the computer, the program code read from the storage medium can cause an operating system or the like operating on the computer to perform part or all of the actual operations based on the instructions of the program code, thereby implementing the functions of any of the above-described embodiments.

[0143] Furthermore, it will be appreciated that, besides being executed by the computer, the program code read from the storage medium can cause an operating system or the like operating on the computer to perform part or all of the actual operations based on the instructions of the program code, thereby implementing the functions of any of the above-described embodiments.

[0144] It should be noted that the terms "first" and "second" and the like in this text are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between such entities or operations. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0145] It will be appreciated by those skilled in the art that all or part of the steps of the above-described method embodiments can be completed by program instructions related to hardware, and the aforementioned program can be stored in a computer-readable storage medium, which, when executed, performs steps including the above-described method embodiments; and the aforementioned storage medium includes various media that can store program codes, such as ROM, RAM, magnetic disk, or optical disk.

[0146] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit the same; and although the present application has been described in detail with reference to the foregoing embodiments, it should be appreciated by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features thereof can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting and optimizing the oxidation behavior of refractory high entropy alloys, characterized in that: include: Obtaining oxidation data of a plurality of refractory high entropy alloys at different temperatures; wherein the oxidation data includes oxidation kinetic constants and oxidation indices; Constructing a sample set based on the element content, physical characteristic information and oxidation data at different temperatures included in the refractory high entropy alloy; The sample set is used to train a machine learning model to obtain a plurality of target prediction models, and oxidation data of any refractory high-entropy alloy component is output based on the target prediction model; wherein the sample set includes element content, physical characteristic information, and temperature as input and oxidation data as output; different target prediction models output different oxidation data; An interpretability analysis is performed on the target prediction model to obtain the key influencing factors of each target prediction model, so as to determine the optimized refractory high entropy alloy composition according to the key influencing factors.

2. The method according to claim 1, characterized in that The refractory high entropy alloy composition includes at least ternary elements of Al, Si, Ti, V, Cr, Zr, Nb, Mo, Hf, Ta and W.

3. The method according to claim 1, characterized in that The physical characteristic information includes: atomic size difference, electronegativity difference, average electronegativity, valence electron concentration, melting point, thermal expansion coefficient, molar volume, mixing entropy, mixing enthalpy, Ω parameter and specific heat capacity.

4. The method according to claim 1, wherein The sample set is used to train the machine learning model to obtain several target prediction models, including: For each type of oxidation data, perform: Inputting the element content, physical characteristic information and temperature of the sample set into the machine learning model, and outputting predicted oxidation data of the type at the temperature; comparing the predicted oxidation data of the type with the oxidation data of the type included in the sample set to obtain a root mean square error and a coefficient of determination; When both the root mean square error and the determination coefficient meet preset thresholds, the training is completed and a target prediction model for this type of oxidation data is obtained.

5. The method according to any one of claims 1 to 4, characterized in that The interpretability analysis of the target prediction model is performed to obtain the key influencing factors of each target prediction model, including: For each target prediction model, execute: Generate an interpreter, and use the interpreter to calculate the feature contribution value of each feature included in each sample; For each feature, the following steps are performed: calculating the mean feature contribution value of the feature based on the feature contribution values ​​of all samples including the feature; The feature corresponding to the feature contribution value mean greater than a preset mean threshold is determined as a key influencing factor.

6. The method according to claim 5, characterized in that Determining and optimizing the composition of the refractory high entropy alloy according to the key influencing factors includes: Determine the contribution and impact direction of the key influencing factors; The key influencing factors with positive contribution direction and representing element content are determined as key elements; Calculate the interaction strength index between any two features, and retain the feature pairs corresponding to the interaction strength index greater than the preset index threshold; Determining an optimization strategy based on the key influencing factors, the feature pairs, and the key elements; An optimized refractory high entropy alloy composition that meets preset requirements is determined based on the optimization strategy.

7. A device for predicting and optimizing the oxidation behavior of refractory high entropy alloys, characterized in that: include: An acquisition module, configured to acquire oxidation data of a plurality of refractory high entropy alloys at different temperatures; wherein the oxidation data includes oxidation kinetic constants and oxidation indices; A sample generation module, configured to construct a sample set based on the element content, physical characteristic information, and oxidation data at different temperatures of the refractory high entropy alloy; a training prediction module, configured to train a machine learning model using the sample set to obtain a plurality of target prediction models, and output oxidation data of any refractory high-entropy alloy component based on the target prediction model; wherein the sample set includes element content, physical characteristic information, and temperature as input and oxidation data as output; different target prediction models output different oxidation data; The optimization design module is used to perform interpretability analysis on the target prediction model to obtain the key influencing factors of each target prediction model, so as to determine the optimized refractory high entropy alloy composition according to the key influencing factors.

8. A computing device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6.

Images