Amorphous alloy design method based on random forest and SHAP analysis

Abstract

The invention relates to an amorphous alloy design method based on random forest and SHAP analysis, and the method comprises the steps: obtaining a to-be-designed alloy, carrying out SHAP analysis on elements in the to-be-designed alloy, and screening out key elements; the key elements are a plurality of elements which have the greatest influence on the forming ability of the amorphous alloy; screening key elements by utilizing mutual information analysis, obtaining a to-be-optimized element combination, determining the optimal proportion and concentration range of each element in the to-be-optimized element combination according to the binary phase diagram and the SHAP dependency diagram, and obtaining an optimal element combination; inputting the optimal element combination into a random forest model, and predicting an amorphous alloy forming ability value; the random forest model is obtained through training of a training set; wherein the training set comprises element composition and proportion of the alloy and target variables. According to the method, the experimental verification frequency is effectively reduced, the alloy design time is shortened, and a scientific and efficient solution is provided for development of amorphous alloy and multi-component materials.

Description

Technical Field

[0001] The present invention relates to the technical field of material design, and in particular to an amorphous alloy design method based on random forest and SHAP analysis. Background Art

[0002] Amorphous alloys have broad application prospects in aerospace, electronic devices, energy and other fields due to their excellent mechanical properties, corrosion resistance, magnetic properties and other characteristics. However, the design and optimization of amorphous alloys usually face great challenges. This is because the forming ability (GFA, Glass-Forming Ability) of amorphous alloys is significantly affected by the alloy composition, element ratio and complex interactions between elements. The traditional composition optimization process often relies on repeated experiments in the laboratory, which is not only costly but also time-consuming and cannot meet the needs of rapid design.

[0003] With the advancement of the Materials Genome Project, data-driven methods have gradually shown great potential in material design. In particular, machine learning methods can extract patterns from big data and efficiently predict material properties, providing new possibilities for material research and development. However, in the process of applying machine learning to amorphous alloy design, how to accurately identify and screen elements that have a significant impact on GFA, and how to effectively analyze the interactions between different elements, are still difficult problems that need to be solved:

[0004] 1. Low efficiency in alloy composition screening: The glass forming ability (GFA) of amorphous alloys is affected by alloy composition, element ratio and complex interactions between elements. Traditional experimental methods are difficult to quickly and efficiently screen out the best alloy composition.

[0005] 2. Too many experimental verifications: Current alloy designs often rely on a large number of experiments to verify and optimize the composition, which not only requires high costs, but also increases the R&D cycle and cannot meet the needs of rapid design.

[0006] 3. The interactions between elements are complex and difficult to analyze: Traditional methods lack effective analytical means when identifying and processing complex interactions between elements, resulting in low design efficiency and inaccurate optimization process.

[0007] 4. Difficulty in optimizing element ratios: Even if important elements can be identified, how to reasonably optimize the ratios of these elements, especially in multi-element systems, remains a difficult problem that needs to be solved.

[0008] In order to better understand and optimize the alloy composition, the present invention proposes a method for screening and optimizing the composition of amorphous alloys based on machine learning. This method combines the random forest model and SHAP (SHapley AdditiveexPlanations) analysis, which can effectively identify the elements that have the greatest impact on the target performance (such as GFA), and further screen and optimize through mutual information analysis and binary phase diagrams. Specifically, the random forest model provides feature importance evaluation, while the SHAP analysis provides detailed contributions of single elements and element combinations to the model output. In addition, mutual information analysis is used to identify the dependencies between key elements, and the binary phase diagram is combined with the interaction analysis of SHAP to further optimize the element ratio, thereby achieving efficient composition design.

[0009] This method can significantly reduce the number of experimental verifications and design time, providing an effective solution for the design of multi-component alloys such as amorphous alloys. This not only improves the efficiency of alloy development, but also lays a technical foundation for the rapid iteration and optimization of new materials. Summary of the invention

[0010] In order to solve the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide an amorphous alloy design method based on random forest and SHAP analysis, which can effectively identify the elements with the greatest impact on target performance (such as GFA) by combining the random forest model and SHAP analysis, and further screen and optimize through mutual information analysis and binary phase diagram. Specifically, the random forest model provides feature importance evaluation, while the SHAP analysis provides detailed contributions of single elements and element combinations to the model output. In addition, mutual information analysis is used to identify the dependencies between key elements, and the binary phase diagram is combined with the interaction analysis of SHAP to further optimize the element ratio, and realize efficient composition design, thereby significantly reducing the number of experimental verifications and design time, and providing an effective solution for the design of multi-component alloys such as amorphous alloys. This not only improves the development efficiency of alloys, but also lays a technical foundation for the rapid iteration and optimization of new materials.

[0011] To achieve the above object, the present invention provides the following solutions:

[0012] An amorphous alloy design method based on random forest and SHAP analysis, comprising:

[0013] Obtaining an alloy to be designed, performing SHAP analysis on the elements in the alloy to be designed, and screening out key elements; the key elements are several elements that have the greatest influence on the forming ability of amorphous alloys;

[0014] The key elements are screened by mutual information analysis to obtain the element combination to be optimized, and the optimal ratio and concentration range of each element in the element combination to be optimized are determined according to the binary phase diagram and the SHAP dependency diagram to obtain the optimal element combination;

[0015] The optimal element combination is input into a random forest model to predict the amorphous alloy forming ability value; the random forest model is trained using a training set; wherein the training set includes: element composition and proportion of the alloy and a target variable.

[0016] Optionally, screening out the key elements includes:

[0017] Performing SHAP analysis on the element composition and ratio to quantify the contribution of each element to the amorphous alloy forming ability under different conditions, i.e., Shapley value;

[0018] Based on the Shapley value, the elements in the alloy to be designed are sorted, a screening threshold is set, and the key elements are further screened out.

[0019] Optionally, obtaining the combination of elements to be optimized includes:

[0020] The mutual information analysis is used to evaluate the interdependence between the elements in the key elements. Based on the interdependence, the element redundancy of the key elements is eliminated to obtain the element combination to be optimized; the element combination to be optimized is the element combination that has the greatest impact on the forming ability of the amorphous alloy.

[0021] Optionally, determining the optimal ratio of each element in the combination of elements to be optimized includes:

[0022] According to the eutectic point of the binary phase diagram, the solubility, phase transition temperature and behavior of the liquid alloy of each element in the element combination to be optimized are obtained, and based on the solubility, phase transition temperature and behavior of the liquid alloy of each element, the optimal proportion of each element in the element combination to be optimized is determined.

[0023] Optionally, determining the concentration range of each element in the element combination to be optimized includes:

[0024] According to the SHAP dependency graph, the sensitivity of the concentration of each element in the element combination to be optimized to the amorphous alloy forming ability is obtained, and in combination with the eutectic point, the target relationship between the element concentration and the amorphous alloy forming ability is obtained;

[0025] Based on the target relationship between the element concentration and the amorphous alloy forming ability, the concentration range of each element in the element combination to be optimized is determined.

[0026] Optionally, obtaining the random forest model includes:

[0027] Obtaining the element composition and proportion of the alloy and a target variable; the target variable is a target parameter of the amorphous alloy forming ability;

[0028] The element composition and proportion of the alloy and the target variable are used as input data, the input data are preprocessed, the preprocessed input data are input into the original random forest model, the hyperparameters of the model are optimized, and the NSGA-II algorithm is used to balance multiple performance indicators of the model to obtain the random forest model.

[0029] Optionally, preprocessing the input data includes:

[0030] Data cleaning and standardization methods are used to scale the content of each element in the input data to a target range.

[0031] Optionally, the hyperparameters of the model include: the number of trees, the maximum depth, and the minimum number of samples.

[0032] The beneficial effects of the present invention are:

[0033] Efficient screening of alloy components: The present invention uses a random forest regression model to extract key factors affecting GFA from a large amount of experimental data. Random forest can automatically screen out the elements that have the greatest impact on glass forming ability through feature importance assessment, achieving efficient and accurate alloy component screening, significantly improving the efficiency of alloy design.

[0034] Reduce the number of experimental verifications: Through the trained random forest model, the present invention can directly predict the influence of alloy composition on GFA without a large number of experiments, reducing the limitation of traditional methods that rely on experiments. Through this data-driven prediction model, the alloy design cycle is greatly shortened, the experimental cost is reduced, and the problem of high cost and long cycle in the traditional design process is solved.

[0035] Accurately analyze the interaction between elements: By introducing SHAP analysis, the present invention can deeply explore and explain the contribution of each element to GFA, not only considering the role of a single element, but also revealing the complex interactions between elements. SHAP analysis provides detailed contributions of each element and element combination to the model output, ensuring the accuracy and reliability of the design process.

[0036] Optimize element ratio: The present invention identifies the interdependence between elements through mutual information analysis, so as to further screen out element combinations with high synergy. Combining the analysis of binary phase diagram and SHAP dependency diagram, the composition ratio of the alloy can be accurately optimized, thus improving the effect of ratio optimization.

[0037] Improve alloy design efficiency: With the help of machine learning technology, especially random forest and SHAP analysis, the present invention effectively reduces the number of experiments and significantly improves the efficiency of alloy composition screening and optimization. Through data-driven methods, the alloy design process is optimized and the R&D cycle is greatly shortened.

[0038] Reduced experimental cost and time: The present invention can avoid tedious experimental processes through model prediction, thereby saving a lot of experimental resources and reducing overall R&D costs. At the same time, the alloy design cycle is also greatly shortened, meeting the needs of rapid R&D.

[0039] Accurately identify key elements and interactions: This paper uses SHAP analysis to provide a detailed explanation of the contribution of each element, clearly demonstrates the interaction between elements, optimizes the alloy composition ratio, and improves the design accuracy. This not only improves the quality of amorphous alloys, but also provides a reference for the design of other alloys.

[0040] Efficient multi-element system optimization: The present invention is capable of processing complex multi-element systems. Through precise ratio optimization, it improves the design efficiency and reliability of multi-component alloys and reduces experimental errors and trial-and-error costs.

[0041] Improving alloy design efficiency: This invention uses machine learning technology, especially random forest and SHAP analysis, to effectively reduce the number of experiments, significantly improve the efficiency of alloy composition screening and optimization, and further optimize the alloy design process through data-driven methods, greatly shortening the R&D cycle.

[0042] Reduced experimental cost and time: The present invention can avoid tedious experimental processes through model prediction, thereby saving a lot of experimental resources and reducing overall R&D costs. At the same time, the alloy design cycle is also greatly shortened, meeting the needs of rapid R&D.

[0043] Efficient multi-element system optimization: The present invention is capable of processing complex multi-element systems. Through precise ratio optimization, it improves the design efficiency and reliability of multi-component alloys and reduces experimental errors and trial-and-error costs.

[0044] Promote the progress of the Material Genome Project: By combining machine learning and data analysis methods, this invention conforms to the development trend of the Material Genome Project, provides new technical support for the rapid research and development of materials, and promotes the digital transformation in the field of material design.

[0045] Expanding the industrial application of amorphous alloys: This invention reduces the R&D cost of amorphous alloys and improves the feasibility of their industrialization by reducing experimental verification and design cycles. This method not only has important application prospects in the fields of aerospace, electronic devices, etc., but also provides new ideas for the design of other high-end materials.

[0046] Providing solutions for other material systems: Although the present invention focuses on the design of amorphous alloys, its method has broad versatility. Whether it is other metal alloys, ceramic materials, or multi-component composite materials, this method can be used for efficient screening and optimization.

[0047] Enhance the intelligence and automation level of alloy design: The present invention introduces intelligent algorithms and automated optimization processes, reduces manual intervention, improves the automation level of the alloy design process, and further improves design efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0049] Figure 1 A flow chart of an amorphous alloy design method based on random forest and SHAP analysis according to an embodiment of the present invention;

[0050] Figure 2 is the R of the model in the embodiment of the present invention on the training set and the test set 2 and mean square error (MSE) indicator diagram;

[0051] Figure 3 This is a performance diagram of the model of an embodiment of the present invention in predicting the formation energy (Dmax) of an amorphous alloy;

[0052] Figure 4 It is an element feature importance diagram based on SHAP analysis according to an embodiment of the present invention;

[0053] Figure 5 Schematic diagram of weighted mutual information values ​​of different alloy systems according to an embodiment of the present invention;

[0054] Figure 6 A histogram of the interaction strength between element pairs according to an embodiment of the present invention;

[0055] Figure 7 is a binary phase diagram of each element system of an embodiment of the present invention;

[0056] Figure 8 A schematic diagram of the distribution of the contribution of different elements in the amorphous alloy composition design based on SHAP analysis as the concentration changes according to an embodiment of the present invention;

[0057] Fig. 9 A parallel coordinate diagram of an embodiment of the present invention;

[0058] Fig.10 The Al with different thicknesses in the embodiment of the present invention 39 Co 17 Ce 22 La 22 X-ray diffraction (XRD) patterns of materials. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] like Figure 1 As shown, this embodiment discloses an amorphous alloy design method based on random forest and SHAP analysis, including: obtaining an alloy to be designed, performing SHAP analysis on the elements in the alloy to be designed, and screening out key elements; the key elements are several elements that have the greatest impact on the forming ability of the amorphous alloy; using mutual information analysis to screen the key elements, obtaining the element combination to be optimized, and determining the optimal proportion and concentration range of each element in the element combination to be optimized according to the binary phase diagram and the SHAP dependency diagram, and obtaining the optimal element combination; inputting the optimal element combination into the random forest model to predict the amorphous alloy forming ability value; the random forest model is obtained by training using a training set; wherein the training set includes: the element composition and proportion of the alloy and the target variable.

[0062] Specifically:

[0063] This embodiment discloses an amorphous alloy design method based on random forest and SHAP analysis, including:

[0064] Data acquisition and preprocessing: Collect alloy compositions and their corresponding glass forming ability (GFA) parameters from experimental data, databases or literature, clean the data, remove outliers, and use standardization techniques to adjust the content of each element to the [0,1] interval;

[0065] Model training and evaluation: A random forest regression model was used to train the preprocessed data to predict the maximum forming capacity (Dmax) of the alloy and the R 2 The performance of the model was evaluated by the mean square error (MSE);

[0066] Element screening: Determine the contribution of each element in the alloy to be designed to Dmax through SHAP analysis, and select the key elements that have a significant impact on GFA;

[0067] Combination screening: mutual information analysis is performed on the selected key elements to evaluate the synergy between element combinations and select element combinations with higher synergy;

[0068] Composition ratio optimization: Based on highly synergistic element combinations, combined with binary phase diagram and SHAP dependency diagram analysis, the optimal concentration range of each element in the alloy is determined to improve the amorphous alloy forming ability;

[0069] Experimental verification: Based on the optimized element combination, experimental verification was carried out to synthesize the alloy and verify its amorphous structure through X-ray diffraction (XRD) analysis.

[0070] Furthermore, obtaining the random forest model includes: obtaining the elemental composition and proportion of the alloy and the target variable; the target variable is the target parameter of the amorphous alloy forming ability; using the elemental composition and proportion of the alloy and the target variable as input data, preprocessing the input data, inputting the preprocessed input data into the original random forest model, optimizing the hyperparameters of the model, and using the NSGA-II algorithm to balance multiple performance indicators of the model to obtain the random forest model.

[0071] Furthermore, preprocessing the input data includes: using data cleaning and standardization methods to scale the content of each element in the input data to a target range.

[0072] Specifically:

[0073] Data collection and preprocessing: Alloy composition and GFA parameters (maximum forming ability Dmax) were collected from experimental data, databases or relevant literature. Data cleaning and standardization methods (MinMaxScaler) were used to scale the content of each element to the interval [0,1] to ensure the consistency of feature weights during model training. The data set was divided into training and test sets in a ratio of 8:2 to ensure model training and verification.

[0074] Random Forest Model Training and Performance Evaluation: A Random Forest Regressor model was used to learn the training set to predict the Dmax of the alloy and obtain the feature importance. 2 ) and mean square error (MSE) indicators to evaluate the model performance and ensure that its prediction ability on the training set and test set has good generalization.

[0075] The training of the random forest model includes: input data: elemental composition and proportion of the alloy, such as the content and proportion of elements such as Ni, Cu, Fe, etc. in the alloy. Each alloy sample is characterized by the percentage of these elements. Output data: glass forming ability (GFA) of the alloy, that is, the predicted Dmax value. The present invention learns the training set through the random forest regression model (Random Forest Regressor). The goal of the training is to establish a relationship between the alloy composition and its glass forming ability so that the model can predict Dmax (GFA).

[0076] The random forest regression model has been optimized, mainly in the following two aspects:

[0077] Hyperparameter optimization: Optuna automatically optimizes the model's hyperparameters (such as the number of trees, maximum depth, minimum number of samples, etc.), significantly improving the model's prediction accuracy and stability. In this way, the present invention achieves more efficient alloy composition design and reduces the risk of overfitting the model.

[0078] Multi-objective optimization: In order to balance multiple performance indicators of the model (such as R 2 And mean square error MSE), the present invention adopts NSGA-II algorithm (non-dominated sorting genetic algorithm II) for multi-objective optimization, so that the model can find the optimal balance between prediction accuracy and error, and further enhance the generalization ability of the model.

[0079] Random Forest Model Training and Performance Evaluation: In the present invention, a Random Forest Regressor model is used to learn the training set to predict the glass forming ability (GFA, abbreviated as Dmax) of the alloy. Specifically, the input data of the training set is the elemental composition and proportion of the alloy, which is obtained by analyzing the chemical composition of the alloy, and the output data is the GFA value (i.e., Dmax) of the alloy. The goal of the random forest model is to predict the glass forming ability of the alloy based on its elemental composition.

[0080] Furthermore, screening out key elements includes: performing SHAP analysis on element composition and proportion to quantify the contribution of each element to the ability to form amorphous alloys under different conditions, namely, the Shapley value; based on the Shapley value, sorting the elements in the designed alloy, setting a screening threshold, and further screening out key elements.

[0081] Specifically:

[0082] In order to evaluate the influence of input features (i.e., element composition and ratio) on the forming ability of amorphous alloys, the present invention adopts the SHAP analysis (SHapley Additive exPlanations) method to perform feature importance analysis. SHAP is an interpretation method based on cooperative game theory, which can provide a Shapley value for each feature to measure the specific contribution of each input feature to the model output.

[0083] SHAP provides the precise contribution of each feature to the model output by considering all possible permutations of feature values, and is able to quantify the interaction effects between features. Specifically, SHAP analysis calculates the marginal contribution of each feature under all possible feature combinations, thereby evaluating its effect on the prediction of the target variable (Dmax).

[0084] SHAP analysis is used for element screening: the elements of the alloy to be designed are introduced into the SHAP value analysis to quantify the contribution of each element to the amorphous alloy forming ability under different conditions. The application of SHAP value enables the present invention to not only evaluate the importance of elements, but also explain the specific effects of each element and its combination on the amorphous alloy forming ability, providing higher explainability and scientific basis.

[0085] In order to accurately quantify the contribution of each element in the prediction of the glass forming ability (GFA, Dmax) of the alloy, the present invention introduces SHAP analysis (SHapley Additive exPlanations). SHAP analysis is based on the idea of ​​game theory and evaluates the influence of each element (feature) on the glass forming ability by calculating the Shapley value. This process not only reveals the marginal contribution of each element to the glass forming ability, but also identifies the elements with important influence, thereby guiding the alloy design.

[0086] The specific analysis process is as follows:

[0087] Calculate the SHAP value:

[0088] Use the SHAP explainer (shap.TreeExplainer(model)) to calculate the contribution of each element of the alloy to be designed to Dmax.

[0089] The Shapley value is based on the idea of ​​cooperative game theory and measures the marginal contribution of each element under all possible feature arrangements. For each alloy sample, the SHAP value provides a specific value for each element, indicating the positive or negative impact of the element on the glass-forming ability.

[0090] Calculation of Shapley Values: By considering all possible feature orders, the SHAP interpreter calculates the marginal contribution of each element in the glass-forming ability. This process takes into account the interactions between features, ensuring that the assessment of feature contributions is both comprehensive and accurate.

[0091] Quantifying feature contribution: Global feature importance: By calculating the Shapley value of each element (feature), the present invention obtains a global feature importance ranking. The importance of a feature is quantified by its average Shapley value. The larger the Shapley value, the greater the contribution of the element to the amorphous alloy forming ability. The present invention ranks the average contribution of each element to screen out the elements that have the greatest impact on the amorphous alloy forming ability.

[0092] Element screening process: According to the SHAP value sorting, the present invention can screen out the elements that play the most important role in alloy design. Specifically, elements with greater contribution will be given priority because they have the greatest impact on the amorphous alloy forming ability in the model. Through the SHAP summary plot, the present invention can intuitively display the average contribution of all elements to the amorphous alloy forming ability. This allows designers to quickly identify key elements and optimize the alloy composition based on the impact of these elements.

[0093] Furthermore, obtaining the element combination to be optimized includes: using mutual information analysis to evaluate the interdependence between each element in the key elements, eliminating the element redundancy of the key elements based on the interdependence, and obtaining the element combination to be optimized; the element combination to be optimized is the element combination that has the greatest impact on the ability to form amorphous alloys.

[0094] Specifically:

[0095] Element combination screening and mutual information analysis: Mutual information analysis is used to perform combination screening on the elements screened by SHAP to evaluate their correlation and synergy.

[0096] SHAP analysis is used for element screening and mutual information analysis is used for alloy system screening:

[0097] After quantifying the contribution of each element to the glassy forming ability (GFA) (i.e., Dmax) through SHAP analysis and screening out the key elements, the next step is to further screen these selected element combinations based on mutual information analysis to determine the optimal alloy system.

[0098] Mutual information analysis to screen alloy element combinations: After screening out the key elements, the next step is to further screen the key element combinations based on mutual information analysis to identify the optimal element combination. Mutual information analysis is mainly used to evaluate the interdependence between two or more variables, helping to identify which element combinations can most effectively affect the amorphous alloy forming ability (GFA).

[0099] Objective of mutual information analysis: The main objective of mutual information analysis is to identify those element combinations that have the largest information gain and can work together effectively. By evaluating the mutual information value of each pair or each multi-element combination, the present invention can find out which element combinations have the greatest impact on the amorphous alloy forming ability and can reduce the redundant impact of other elements.

[0100] Mutual information analysis can quantify the dependencies between elements and thus identify the element combinations with the best synergistic effects. This process ensures that the selected element combinations have the potential to improve the amorphous alloy forming ability GFA.

[0101] The specific process of mutual information analysis:

[0102] Define the input features of mutual information analysis and select the key elements screened out from SHAP analysis as the input features of mutual information analysis.

[0103] For each pair of elements, mutual information regression analysis (mutual_info_regression) is used to calculate the mutual information value between them and Dmax, that is, the contribution of these element combinations to the amorphous alloy forming ability. The higher the mutual information value, the greater the influence of these element combinations on the amorphous alloy forming ability, and the higher the predictive ability.

[0104] Screening the synergistic effect of the combination: By calculating the mutual information value of each element combination, the present invention ranks the element combinations. The element combinations with higher mutual information values ​​are selected as the preferred combinations. The element combinations with high mutual information values ​​have stronger synergistic effects. Element combinations with large information gain: Based on the sorting results, the element combinations that can maximize the information gain are screened. These combinations help reduce the impact of redundant features and may have stronger synergistic effects in the alloy.

[0105] Screening high mutual information combinations: Screen out high mutual information element combinations from mutual information analysis. These combinations are usually elements with strong synergistic effects on each other, which can effectively improve the amorphous alloy forming ability (Dmax). By selecting element combinations with high mutual information values, designers can effectively optimize the alloy composition to ensure that the element combinations in the alloy system can play the greatest role and improve the overall performance.

[0106] Furthermore, determining the optimal ratio of each element in the element combination to be optimized includes: obtaining the solubility, phase transition temperature and behavior of the liquid alloy of each element in the element combination to be optimized according to the eutectic point of the binary phase diagram, and determining the optimal ratio of each element in the element combination to be optimized based on the solubility, phase transition temperature and behavior of the liquid alloy of each element.

[0107] Furthermore, determining the concentration range of each element in the element combination to be optimized includes: obtaining the sensitivity of the concentration of each element in the element combination to be optimized to the amorphous alloy forming ability according to the SHAP dependency graph, and obtaining the target relationship between the element concentration and the amorphous alloy forming ability in combination with the eutectic point; based on the target relationship between the element concentration and the amorphous alloy forming ability, determining the concentration range of each element in the element combination to be optimized.

[0108] Specifically:

[0109] Combining the ratio optimization of binary phase diagram: Based on the determined element combination, binary phase diagram analysis is performed to study the physical interaction between elements, especially the eutectic point and the amorphous forming area. Combined with the SHAP dependency diagram, the effects of individual elements and their combinations on GFA at different concentrations are analyzed to optimize the specific composition ratio of the elements to improve the amorphous forming ability of the alloy.

[0110] Optimizing the composition ratio of amorphous alloys by combining the eutectic point of the binary phase diagram with the SHAP dependency diagram: After determining the key elements and screening out the elements with high contribution, the next step is to optimize the composition ratio of the alloy through the eutectic point criterion of the binary phase diagram and SHAP analysis to improve the glass forming ability (GFA) (i.e., Dmax). Specifically, by combining the eutectic point analysis and the SHAP dependency diagram, the present invention can analyze the effects of single elements and element combinations on GFA at different concentrations, thereby obtaining the specific composition ratio of amorphous alloy formation.

[0111] The eutectic point criterion is a very important theoretical basis in the binary phase diagram, which describes the phase transition behavior between the liquid phase and the solid phase of two elements or alloys at a certain temperature and element ratio. In alloy design, the use of the eutectic point can help the present invention understand the solubility of different elements, the phase transition temperature and the behavior of liquid alloys. When designing the alloy composition, the eutectic point criterion helps to determine the optimal ratio of elements in the alloy, so that it can effectively avoid the formation of crystals and promote the formation of amorphous phases during the cooling process. In a binary alloy system, the eutectic point is often located at the boundary of the amorphous formation region, so the element ratio in this region has a higher glass forming ability GFA.

[0112] SHAP analysis is used to quantify the contribution of each element to the glass forming ability Dmax. After the key elements are screened out through SHAP analysis, the present invention can use the SHAP dependency diagram to further analyze the specific effects of single elements and element combinations on Dmax at different concentrations. By analyzing the SHAP dependency diagram, designers can identify the sensitivity of each element concentration to the amorphous alloy forming ability, and which concentration ranges can maximize Dmax. Using the eutectic point criterion and the SHAP dependency diagram, the composition ratio of the alloy can be further optimized. Through the SHAP dependency diagram, it can be clearly pointed out that near the eutectic point, the concentration changes of certain elements have a significant effect on GFA. By combining this information with the amorphous forming zone in the phase diagram, designers can determine the optimal element ratio and concentration range, thereby improving the glass forming ability of the alloy.

[0113] Through SHAP interaction analysis and physical interaction research, the present invention can analyze the influence of element combinations on Dmax (glass forming ability). Certain element combinations may have a strong synergistic effect near the eutectic point, which can optimize the performance of amorphous alloys. Combined with SHAP analysis, designers can identify the element combination with the greatest synergistic effect and adjust its concentration to optimize the Dmax of the alloy. Effect of concentration changes on GFA: Using the SHAP dependency graph, the contribution of each element to GFA at different concentrations can be analyzed, and the concentration distribution of these elements can be ensured to be in the eutectic point region, that is, the optimal concentration interval between the liquid phase and the solid phase of the alloy, to enhance the amorphous alloy forming ability.

[0114] Combining SHAP analysis, phase diagram analysis and eutectic point criteria, the optimal concentration range of key elements in the alloy can be determined. By optimizing the concentration of these elements, they are ensured to be located in the amorphous formation region, that is, the eutectic point region or near it, thereby promoting the formation of amorphous phase in the alloy during rapid cooling.

[0115] The trained random forest model is used to predict the optimal combination of designed elements and evaluate its forming capacity (Dmax). Finally, experimental verification is performed to ensure the feasibility of the design. The verification results should be consistent with the model predictions, confirming the reliability of the optimization method in practical applications.

[0116] like Figure 2 As shown, the R of the model on the training set and the test set 2 And mean square error (MSE) indicators are used to measure the fitting performance and predictive ability of the model.

[0117] like Figure 3As shown in the figure, the performance of the model in predicting the formation energy (Dmax) of amorphous alloys. The horizontal axis represents the measured Dmax, and the vertical axis represents the Dmax predicted by the model. The blue asterisks in the figure represent the training set data, the red circles represent the test set data, and the black dotted line is the ideal "perfect fit" line, that is, the state where the predicted value is completely consistent with the measured value.

[0118] like Figure 4 As shown in the figure, the element feature importance diagram based on SHAP analysis shows the influence weight of different elements on the model output. This figure is used to illustrate the importance ranking of each element in the design of amorphous alloys.

[0119] like Figure 5 As shown in the figure, the weighted mutual information values ​​of different alloy systems. Through mutual information analysis, the correlation between elements can be identified and the element combination with the largest information gain can be screened out.

[0120] like Figure 6 As shown in the figure, the bar graph of the interaction strength between element pairs shows the mutual influence of different element combinations in the design of amorphous alloy composition. This graph is used to evaluate the synergistic effect of element combinations, where the higher the interaction strength, the more significant the interaction between these element pairs, which may have a positive impact on the target performance of the alloy.

[0121] like Figure 7 As shown, the binary phase diagrams of various element systems show the eutectic points of different alloy systems.

[0122] like Figure 8 As shown in the figure, the contribution of different elements in the amorphous alloy composition design based on SHAP analysis is distributed as the concentration changes. These figures are used to illustrate the contribution pattern of each high-priority element combination in different concentration ranges to help optimize the alloy composition design.

[0123] like Fig. 9 As shown in the figure, the parallel coordinate diagram shows the influence of the concentration of four elements, Al, Co, Ce and La, on the amorphous alloy forming energy (Dmax). The larger the Dmax value, the stronger the amorphous alloy forming ability. The figure shows the contribution of different component combinations to Dmax through a color mapping path, with the color gradually changing from blue (indicating high Dmax) to red (indicating low Dmax).

[0124] like Fig.10 As shown, different thicknesses of Al 39 Co 17 Ce 22 La 22 X-ray diffraction (XRD) spectra of materials. Used to analyze the crystal structure characteristics of materials at different casting diameters (4mm, 5mm and 6mm).

[0125] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for designing amorphous alloys based on random forest and SHAP analysis, characterized in that: include: Obtaining an alloy to be designed, performing SHAP analysis on elements in the alloy to be designed, and screening out key elements; The key elements are the elements that have the greatest impact on the forming ability of amorphous alloys; The key elements are screened by mutual information analysis to obtain the element combination to be optimized, and the optimal ratio and concentration range of each element in the element combination to be optimized are determined according to the binary phase diagram and the SHAP dependency diagram to obtain the optimal element combination; Inputting the optimal element combination into a random forest model to predict the amorphous alloy forming ability value; The random forest model is obtained by training with a training set, wherein the training set includes: element composition and proportion of the alloy and a target variable.

2. The amorphous alloy design method based on random forest and SHAP analysis according to claim 1, characterized in that: The key elements screened out include: Performing SHAP analysis on the element composition and ratio to quantify the contribution of each element to the amorphous alloy forming ability under different conditions, i.e., Shapley value; Based on the Shapley value, the elements in the alloy to be designed are sorted, a screening threshold is set, and the key elements are further screened out.

3. The amorphous alloy design method based on random forest and SHAP analysis according to claim 1, characterized in that: Obtaining the combination of elements to be optimized includes: The mutual information analysis is used to evaluate the interdependence between the elements in the key elements. Based on the interdependence, the element redundancy of the key elements is eliminated to obtain the element combination to be optimized; the element combination to be optimized is the element combination that has the greatest impact on the forming ability of the amorphous alloy.

4. The amorphous alloy design method based on random forest and SHAP analysis according to claim 1, characterized in that: Determining the optimal ratio of each element in the combination of elements to be optimized includes: According to the eutectic point of the binary phase diagram, the solubility, phase transition temperature and behavior of the liquid alloy of each element in the element combination to be optimized are obtained, and based on the solubility, phase transition temperature and behavior of the liquid alloy of each element, the optimal proportion of each element in the element combination to be optimized is determined.

5. The amorphous alloy design method based on random forest and SHAP analysis according to claim 4, characterized in that: Determining the concentration range of each element in the element combination to be optimized includes: According to the SHAP dependency graph, the sensitivity of the concentration of each element in the element combination to be optimized to the amorphous alloy forming ability is obtained, and in combination with the eutectic point, the target relationship between the element concentration and the amorphous alloy forming ability is obtained; Based on the target relationship between the element concentration and the amorphous alloy forming ability, the concentration range of each element in the element combination to be optimized is determined.

6. The amorphous alloy design method based on random forest and SHAP analysis according to claim 1, characterized in that: Obtaining the random forest model includes: Obtaining the element composition and proportion of the alloy and a target variable; the target variable is a target parameter of the amorphous alloy forming ability; The element composition and proportion of the alloy and the target variable are used as input data, the input data are preprocessed, the preprocessed input data are input into the original random forest model, the hyperparameters of the model are optimized, and the NSGA-II algorithm is used to balance multiple performance indicators of the model to obtain the random forest model.

7. The amorphous alloy design method based on random forest and SHAP analysis according to claim 6, characterized in that: Preprocessing the input data includes: Data cleaning and standardization methods are used to scale the content of each element in the input data to a target range.

8. The amorphous alloy design method based on random forest and SHAP analysis according to claim 6, characterized in that: The hyperparameters of the model include: number of trees, maximum depth, and minimum number of samples.

Images