Method and device for predicting hydrogen storage performance of hydrogen storage alloy and vehicle

Abstract

The invention provides a hydrogen storage performance prediction method and device of a hydrogen storage alloy and a vehicle, and belongs to the technical field of solid hydrogen storage. The method for predicting the hydrogen storage performance of the hydrogen storage alloy comprises the steps that a sample database is obtained, the sample database comprises characteristic parameters of various hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, and the various hydrogen storage alloys are different in component and the same in crystal structure; based on the sample database, a plurality of preset machine learning models are adopted for training, and a plurality of initial prediction models are obtained; determining an optimal initial prediction model from the plurality of initial prediction models; based on a sample database, an Ada integration model is established for the optimal initial prediction model, and a performance prediction model is obtained; and performing hydrogen storage prediction on the hydrogen storage alloy to be predicted based on the performance prediction model to obtain a hydrogen storage prediction result. The method improves the prediction accuracy of the prediction model, is higher in efficiency and precision, remarkably reduces the cost of experiment screening, and shortens the experiment period.

Description

Technical Field

[0001] The present invention relates to the field of solid-state hydrogen storage technology, and in particular to a method for predicting the hydrogen storage performance of a hydrogen storage alloy, a device for predicting the hydrogen storage performance of a hydrogen storage alloy, a machine-readable storage medium, an electronic device and a vehicle. Background Art

[0002] As the energy crisis and environmental pollution become increasingly serious, hydrogen energy has received widespread attention due to its renewable, clean, pollution-free and high energy density. The development of safe and efficient hydrogen storage alloys is a key link in its development. For example, Ti-based hydrogen storage alloys are considered to be the most promising hydrogen storage alloys for use in hydrogen fuel cell vehicles due to their good hydrogen absorption and desorption kinetics, activation performance, cycle performance, safety and low cost. However, low hydrogen storage capacity will limit its application and promotion, so the development of high-performance hydrogen storage materials is crucial to the application of hydrogen storage technology.

[0003] The current prediction method of hydrogen storage performance of hydrogen storage alloys generally uses empirical criteria to predict performance, which has low prediction precision and accuracy. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a method for predicting the hydrogen storage performance of a hydrogen storage alloy, a device for predicting the hydrogen storage performance of a hydrogen storage alloy, a machine-readable storage medium, an electronic device and a vehicle. The method for predicting the hydrogen storage performance of a hydrogen storage alloy achieves rapid and accurate prediction of the hydrogen storage performance of a hydrogen storage alloy through cross-validation of a machine learning model, an Ada integrated model and multi-parameter iterative screening and optimization.

[0005] In order to achieve the above-mentioned object, the first aspect of the present application provides a method for predicting the hydrogen storage performance of a hydrogen storage alloy, comprising:

[0006] Acquire a sample database, wherein the sample database includes characteristic parameters of a plurality of hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure;

[0007] Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models;

[0008] Determining an optimal initial prediction model from the multiple initial prediction models;

[0009] Based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model;

[0010] Based on the performance prediction model, hydrogen storage prediction is performed on the hydrogen storage alloy to be predicted to obtain a hydrogen storage prediction result.

[0011] In the embodiment of the present application, the step of obtaining a sample database includes:

[0012] Obtain various hydrogen storage alloys and the corresponding hydrogen storage capacities of various hydrogen storage alloys;

[0013] First-principles calculations were performed on various hydrogen storage alloys to obtain the element crystal parameters of various hydrogen storage alloys;

[0014] Based on the element crystal parameters of the various hydrogen storage alloys, characteristic parameters of the various hydrogen storage alloys are calculated according to preset characteristic parameter calculation formulas;

[0015] Based on the characteristic parameters of the various hydrogen storage alloys and the hydrogen storage capacities corresponding to the various hydrogen storage alloys, a sample database is obtained.

[0016] In the embodiment of the present application, the preset multiple machine learning models are trained based on the sample database to obtain multiple initial prediction models, including:

[0017] Cleaning the data in the sample database to obtain a new sample database;

[0018] Based on the new sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models.

[0019] In the embodiment of the present application, the characteristic parameter includes temperature, and the hydrogen storage capacity corresponding to the various hydrogen storage alloys is the hydrogen storage capacity corresponding to the various hydrogen storage alloys at different temperatures.

[0020] In the embodiment of the present application, the preset multiple machine learning models are trained based on the sample database to obtain multiple initial prediction models, including:

[0021] A1: Divide the sample database into k subsets to obtain k sub-sample sets;

[0022] A2: Select k-1 sub-sample sets from the k sub-sample sets as training sets, and the remaining sub-sample sets as validation sets. Based on the training sets, train each machine learning model respectively, and use the validation sets to test the trained machine learning model.

[0023] A3: Repeat steps A1-A2 until every sample in the sample database is traversed as a validation set to obtain multiple initial prediction models.

[0024] In the embodiment of the present application, determining the optimal initial prediction model from the multiple initial prediction models includes:

[0025] Using each initial prediction model to predict the validation set, respectively, to obtain prediction results corresponding to each initial prediction model;

[0026] Based on the prediction results and actual results corresponding to each of the initial prediction models, respectively calculate the regression performance evaluation index of each initial prediction model;

[0027] Based on the regression performance evaluation indicators of the various initial prediction models, an optimal initial prediction model is determined.

[0028] In the embodiment of the present application, it also includes:

[0029] Obtaining a multi-element hydrogen storage alloy to be predicted;

[0030] Dividing the alloy components in the multi-component hydrogen storage alloy to be predicted according to preset intervals to obtain a plurality of alloy ratio combinations;

[0031] Based on the performance prediction model, hydrogen storage prediction is performed on each alloy ratio combination to obtain hydrogen storage prediction results of multiple alloy ratio combinations;

[0032] Based on the hydrogen storage prediction results of the plurality of alloy ratio combinations, an alloy composition with optimal hydrogen storage performance is obtained.

[0033] In the embodiment of the present application, it also includes:

[0034] Obtaining the experimental hydrogen storage results corresponding to the alloy composition with the best hydrogen storage performance;

[0035] Based on the experimental hydrogen storage results, the hydrogen storage prediction results corresponding to the alloy composition with the optimal hydrogen storage performance are verified to obtain verification results.

[0036] A second aspect of the present application provides a device for predicting hydrogen storage performance of a hydrogen storage alloy, comprising:

[0037] An acquisition module, used to acquire a sample database, wherein the sample database includes characteristic parameters of a plurality of hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure;

[0038] A training module, used to train multiple preset machine learning models based on the sample database to obtain multiple initial prediction models;

[0039] A determination module, used to determine an optimal initial prediction model from the multiple initial prediction models;

[0040] An establishment module is used to establish an Ada integrated model corresponding to the optimal initial prediction model based on the sample database to obtain a performance prediction model;

[0041] The prediction module is used to perform hydrogen storage prediction on the hydrogen storage alloy to be predicted based on the performance prediction model to obtain a hydrogen storage prediction result.

[0042] A third aspect of the present application provides an electronic device, the electronic device comprising:

[0043] at least one processor;

[0044] a memory connected to the at least one processor;

[0045] The memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the above-mentioned method for predicting the hydrogen storage performance of the hydrogen storage alloy by executing the instructions stored in the memory.

[0046] A fourth aspect of the present application provides a machine-readable storage medium having instructions stored thereon, which, when executed by a processor, configure the processor to execute the above-mentioned method for predicting the hydrogen storage performance of hydrogen storage alloys.

[0047] A fifth aspect of the present application provides a vehicle, comprising the above-mentioned hydrogen storage performance prediction device of the hydrogen storage alloy or the above-mentioned electronic device or the above-mentioned machine-readable storage medium.

[0048] Through the above technical scheme, by obtaining a sample database, the sample database includes characteristic parameters of various hydrogen storage alloys and the corresponding hydrogen storage capacity of various hydrogen storage alloys, and the various hydrogen storage alloys have different compositions but the same crystal structure; based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models; the optimal initial prediction model is determined from the multiple initial prediction models; based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model; based on the performance prediction model, hydrogen storage prediction is performed on the predicted hydrogen storage alloy to obtain a hydrogen storage prediction result. By selecting hydrogen storage alloys with different compositions and the same crystal structure as the sample database, the optimal initial prediction model is selected from multiple initial prediction models trained by multiple machine learning models through cross-validation of the machine learning model, and the Ada integrated model is further used for parameter iterative screening, thereby improving the accuracy of the prediction model prediction. Compared with the empirical criterion method, linear fitting method and traditional machine learning method in the prior art, the method is more efficient and accurate, significantly reducing the cost of experimental screening and shortening the experimental cycle. The sample database contains hydrogen storage alloys of different compositions with the same crystal structure, which can eliminate the influence of crystal structure changes caused by external factors of alloy composition on the formation of hydrogen storage performance, further improving the prediction accuracy.

[0049] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0051] Figure 1 A schematic diagram of a process flow of a method for predicting hydrogen storage performance of a hydrogen storage alloy according to an embodiment of the present application is schematically shown;

[0052] Figure 2 Schematically showing the microscopic crystal structure diagram of different metals according to the embodiments of the present application;

[0053] Figure 3 A diagram schematically showing a correlation analysis diagram of variables in a database according to an embodiment of the present application;

[0054] Figure 4 Schematically showing different initial prediction model accuracy evaluation diagrams according to embodiments of the present application;

[0055] Figure 5 The MSE graph of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown;

[0056] Figure 6 The MAE diagram of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown;

[0057] Figure 7 The PC diagram of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown;

[0058] Figure 8 The ada_XGB integrated prediction model accuracy index R according to the embodiment of the present application is schematically shown. 2 picture;

[0059] Fig. 9 A diagram schematically showing a comparison of predicted values ​​and actual values ​​of the ada_XGB integrated prediction model according to an embodiment of the present application;

[0060] Fig.10 The Ti according to the embodiment of the present application is schematically shown. 0.75 Zr 0.35 Mn 2-y-z Cr y Ni z Cu i Schematic diagram of hydrogen storage prediction results;

[0061] Fig.11 A schematic diagram of a Ti-based hydrogen storage alloy design flow chart according to an embodiment of the present application is shown;

[0062] Fig.12 A schematic diagram of the structure of a hydrogen storage performance prediction device for a hydrogen storage alloy according to an embodiment of the present application is shown;

[0063] Fig.13 The internal structure diagram of a computer device according to an embodiment of the present application is schematically shown.

[0064] Description of Reference Numerals

[0065] 410 - acquisition module; 420 - training module; 430 - determination module; 440 - establishment module; 450 - prediction module; A01 - processor; A02 - network interface; A03 - internal memory; A04 - display screen; A05 - input device; A06 - non-volatile storage medium; B01 - operating system; B02 - computer program. DETAILED DESCRIPTION

[0066] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0067] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0068] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0069] Please see Figure 1 , Figure 1The schematic diagram of the flow chart of a method for predicting the hydrogen storage performance of a hydrogen storage alloy according to an embodiment of the present application is shown schematically. This embodiment provides a method for predicting the hydrogen storage performance of a hydrogen storage alloy. By selecting hydrogen storage alloys with the same crystal structure of different components as data sources, the Ada integrated model and parameter iterative screening are used to improve the accuracy of prediction, which is more efficient and accurate than the empirical criterion method, the linear fitting method and the traditional machine learning method. It should be noted that the scheme in this embodiment is applicable to various types of hydrogen storage alloys. In order to facilitate the description of the scheme, this embodiment mainly takes the hydrogen storage alloy as a Ti-based hydrogen storage alloy as an example for description.

[0070] This embodiment provides a method for predicting the hydrogen storage performance of a hydrogen storage alloy, comprising the following steps:

[0071] Step 210: Acquire a sample database, wherein the sample database includes characteristic parameters of various hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure;

[0072] In this embodiment, the sample database is used to train the model. The sample database includes multiple groups of sample data, and each sample data includes characteristic parameters of a hydrogen storage alloy and the corresponding hydrogen storage capacity of the hydrogen storage alloy. For example: the sample database includes characteristic parameters of titanium-aluminum alloys and the corresponding hydrogen storage capacity is 1.76, and the characteristic parameters of TiMn alloys and the corresponding hydrogen storage capacity is 1.70. The hydrogen storage capacity corresponding to the above-mentioned various hydrogen storage alloys can also be the hydrogen storage capacity corresponding to various hydrogen storage alloys at different temperatures. For example: the hydrogen storage capacity of the titanium-aluminum alloy is 1.76 when the temperature is 10 degrees, and the hydrogen storage capacity of the titanium-aluminum alloy is 1.70 when the temperature is 20 degrees. The various hydrogen storage alloys have different compositions but the same crystal structure. For example, the crystal structure is C 14 Crystalline phase or C 15 Crystalline structure.

[0073] Among them, the characteristic parameters may include: the average value of atomic radius, the variance of atomic radius, the average value of ionic radius, the variance of ionic radius, the average value of lattice constant a, the variance of lattice constant a, the average value of lattice constant b, the variance of lattice constant b, the average value of lattice constant c, the variance of lattice constant c, the average value of elastic modulus, the variance of elastic modulus, the average value of electron concentration, the variance of electron concentration, the average value of the number of valence electrons, the variance of the number of valence electrons, the average value of valence electron density, the variance of valence electron density and atomic size factor, etc. The lattice constant a is the lattice constant in the a-axis direction of the unit cell, the lattice constant b is the lattice constant in the b-axis direction of the unit cell, and the lattice constant c is the lattice constant in the c-axis direction of the unit cell.

[0074] It should be noted that the above-mentioned characteristic parameters may also include any number of the above-mentioned characteristics, or add characteristics on the basis of the above-mentioned characteristic parameters, which are not limited in this embodiment. For example, the characteristic parameters include: the average value of atomic radius, the variance of atomic radius, the average value of ionic radius, the variance of ionic radius, the average value of lattice constant a, the variance of lattice constant a, the average value of lattice constant b, the variance of lattice constant b, the average value of lattice constant c, the variance of lattice constant c, the average value of elastic modulus, the variance of elastic modulus, the average value of electron concentration, the variance of electron concentration, the average value of the number of valence electrons, the variance of the number of valence electrons, the average value of valence electron density, and the variance of valence electron density.

[0075] By setting the above parameters as characteristic parameters, the hydrogen storage alloy can be better characterized, which helps to train a more accurate model.

[0076] In some embodiments, the characteristic parameter includes temperature, and the hydrogen storage capacity corresponding to the various hydrogen storage alloys is the hydrogen storage capacity corresponding to the various hydrogen storage alloys at different temperatures.

[0077] In this embodiment, the characteristic parameters may include the above parameters and the temperature parameter. Accordingly, the hydrogen storage capacity is the hydrogen storage capacity at different temperatures. 0.8 Zr 0.2 The hydrogen storage capacity of Mn1Cr1 at a temperature of 10 degrees is 1.76, Ti 0.9 Zr 0.2 The hydrogen storage capacity of the component Mn1Cr1 is 1.80 at a temperature of 10 degrees, and the hydrogen storage capacity of the component TiCr2 is 2.0 at a temperature of 20 degrees.

[0078] By introducing temperature as a characteristic parameter, when making a prediction, the temperature to be predicted is input at the same time, and the prediction result at that temperature can be output, thereby realizing the prediction of hydrogen storage performance at different temperatures and meeting the current prediction requirements for hydrogen storage performance at different temperatures.

[0079] In some embodiments, obtaining a sample database comprises the following steps:

[0080] First, a variety of hydrogen storage alloys and the corresponding hydrogen storage capacities of the various hydrogen storage alloys are obtained;

[0081] In this embodiment, the hydrogen storage capacity corresponding to the above-mentioned various hydrogen storage alloys may be the hydrogen storage capacity at room temperature, or the hydrogen storage capacity corresponding to various hydrogen storage alloys at different temperatures. For example, taking the hydrogen storage alloy as a Ti-based hydrogen storage alloy as an example, the various hydrogen storage alloys may include TiMn series, TiZrMn series, titanium aluminum series, TiMnCr series, TiZrMnCr series, TiZrMnCrCu series, TiZrMnCrCuFe series, etc., for example, Ti 0.8 Zr 0.2 Mn1Cr1, TiMn2, etc. When the hydrogen storage capacity corresponding to the above-mentioned various hydrogen storage alloys is the hydrogen storage capacity corresponding to various hydrogen storage alloys at different temperatures, it can be to collect m kinds of Ti-based alloy components and their corresponding hydrogen storage capacities at different temperatures. The data sources of the hydrogen storage capacity, hydrogen storage temperature and alloy composition of the above-mentioned hydrogen storage alloys include existing technical literature, experimental results or tool manuals, etc.

[0082] Then, first-principles calculations were performed on various hydrogen storage alloys to obtain the elemental crystal parameters of various hydrogen storage alloys;

[0083] In the present embodiment, the first principle refers to the calculation and analysis based on the basic principles of physics, without relying on empirical parameters or analogies, and directly starting from the basic theories such as quantum mechanics and classical mechanics. The electronic structure, mechanical properties and other properties of the material can be predicted by first principle calculations without relying on the fitting of experimental data. The first principle calculations include lattice and structure optimization, self-consistent field calculations, energy-volume fitting and other calculations. By performing first principle calculations on various hydrogen storage alloys respectively, the element crystal parameters of various hydrogen storage alloys can be obtained. The above-mentioned first principle calculations can be achieved using existing methods, which will not be repeated here. The calculated element crystal parameters include: atomic radius, ionic radius, lattice constant in the a-axis direction of the unit cell, lattice constant in the b-axis direction of the unit cell, lattice constant electronegativity in the c-axis direction of the unit cell, bulk modulus, electron concentration, valence electron number, valence electron density and other parameters. The above-mentioned calculation process can be calculated using density functional theory (DFT), and the specific calculation process belongs to the prior art, which will not be repeated here.

[0084] Taking Ti-based hydrogen storage alloy as an example, based on the collected m kinds of Ti-based alloy components, the atomic radius, ionic radius, lattice constant in the a-axis direction of the unit cell, lattice constant in the b-axis direction of the unit cell, lattice constant electronegativity in the c-axis direction of the unit cell, bulk modulus, electron concentration, valence electron number, valence electron density and other parameters corresponding to the element crystal are collected through first-principles calculation and other means. The crystal structure is as follows Figure 2 As shown, Figure 2The microscopic crystal structure diagrams of different metals according to the embodiments of the present application are schematically shown. The basic parameters corresponding to the constituent elements of the Ti-based hydrogen storage alloy, that is, the element crystal parameters are shown in Table 1 below.

[0085] Table 1: Basic parameters corresponding to each component element of Ti-based hydrogen storage alloy

[0086]

[0087]

[0088] Then, based on the element crystal parameters of the various hydrogen storage alloys, characteristic parameters of the various hydrogen storage alloys are calculated according to preset characteristic parameter calculation formulas;

[0089] In the present embodiment, the above-mentioned preset characteristic parameter calculation formulas include a calculation formula for the average value of atomic radius, a calculation formula for the variance of atomic radius, a calculation formula for the average value of ionic radius, a calculation formula for the variance of ionic radius, a calculation formula for the average value of lattice constant a, a calculation formula for the variance of lattice constant a, a calculation formula for the average value of lattice constant b, a calculation formula for the variance of lattice constant b, a calculation formula for the average value of lattice constant c, a calculation formula for the variance of lattice constant c, a calculation formula for the average value of elastic modulus, a calculation formula for the variance of elastic modulus, a calculation formula for the average value of electron concentration, a calculation formula for the variance of electron concentration, a calculation formula for the average value of the number of valence electrons, a calculation formula for the variance of the number of valence electrons, a calculation formula for the average value of valence electron density, a calculation formula for the variance of valence electron density, an atomic size factor calculation formula, etc. The specific calculation formulas are as follows:

[0090]

[0091]

[0092]

[0093] Among them, C i represents the atomic percentage of the i-th metal in the alloy, n is the number of alloy components, R i represents the atomic radius of the ith metal component, r i represents the ionic radius of the i-th metal component, a i represents the electronegativity of the lattice constant along the a-axis of the unit cell of the i-th component metal, b i represents the electronegativity of the lattice constant in the direction of the b-axis of the unit cell of the i-th component metal, ci represents the electronegativity of the lattice constant in the direction of the c-axis of the unit cell of the i-th component metal, G i represents the bulk modulus of the ith component metal, e / a i represents the electron concentration of the ith metal component, X irepresents the number of valence electrons of the ith metal component, dn i Represents the valence electron density parameter of the i-th component metal. mean is the average atomic radius, R var is the variance of the atomic radius, r mean is the average ionic radius, r var is the variance of the ionic radius, a mean is the average value of the lattice constant a, a var is the variance of the lattice constant a, b mean is the average value of the lattice constant b, b var is the variance of the lattice constant b, c mean is the average value of the lattice constant c, c var is the variance of the lattice constant c, G mean is the average value of elastic modulus, G var is the variance of the elastic modulus, e / a mean is the average electron concentration, e / a var is the variance of the electron concentration, X mean is the average number of valence electrons, X var is the variance of the number of valence electrons, dn mean is the average value of the valence electron density, dn var is the variance of the valence electron density, and δ is the atomic size factor. The above-mentioned component metals are the elements in the hydrogen storage alloy.

[0094] Finally, a sample database is obtained based on the characteristic parameters of the various hydrogen storage alloys and the hydrogen storage capacities corresponding to the various hydrogen storage alloys.

[0095] In this embodiment, after the characteristic parameters of various hydrogen storage alloys are calculated, a sample database is constructed with the corresponding hydrogen storage capacity, that is, each sample data includes the characteristic parameters of a hydrogen storage alloy and the corresponding hydrogen storage capacity.

[0096] Through first-principles calculations, characteristic quantities at the microscopic level can be collected, which together with the hydrogen storage capacity constitute sample data, making the sample database more granular and facilitating the training of more accurate prediction models.

[0097] Step 220: Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models;

[0098] In this embodiment, the above-mentioned multiple machine learning models can be pre-set, and the machine learning algorithms include linear regression LR, linear regression variant LASSO, gradient boosting XGBoost, random forest RF, support vector machine SVM, Stacking, etc.

[0099] For each machine learning model, a sample database can be used for model training, and the model parameters can be continuously optimized to obtain the corresponding initial prediction model. The parameters of each machine learning model are different. They can be pre-set according to the actual situation, and then continuously trained and optimized. For example, the parameters of LR include {regularization type penalty, regularization strength C, optimization algorithm solver, maximum number of iterations max_iter}, the parameters of LASSO include {regularization strength alpha, maximum number of iterations max_iter}, the parameters of XGBoost include {maximum depth of the tree max_depth, learning rate learning_rate, number of trees n_estimators, sampling ratio subsample, column sampling ratio colsample_bytree, minimum loss reduction gamma, regularization strength alpha}, and the parameters of RF include {maximum depth of the tree max_depth, learning rate learning_rate, number of trees n_estimators, sampling ratio subsample, column sampling ratio colsample_bytree, minimum loss reduction gamma, regularization strength alpha}. _depth, the number of trees n_estimators, the minimum number of samples required for internal node re-division min_samples_split, the minimum number of samples for leaf nodes min_samples_leaf, the maximum number of features for each tree max_features, whether there is a sampling bootstrap with replacement}, the parameters of SVM include {regularization parameter C, kernel function type kernel, kernel function coefficient gamma, polynomial kernel function degree, independent term coef0 in kernel function}, and the parameters of Stacking include {base learner estimators, meta learner meta_learner}. It should be noted that the above machine learning model training process belongs to the prior art and will not be repeated here.

[0100] In some embodiments, the training of multiple preset machine learning models based on the sample database to obtain multiple initial prediction models includes:

[0101] First, data in the sample database is cleaned to obtain a new sample database;

[0102] In this embodiment, the data cleaning is to remove some irrelevant data, which can be specifically achieved by calculating the correlation between variables and then removing the data with low correlation. Taking the hydrogen storage alloy as a Ti-based hydrogen storage alloy as an example, the correlation of variables in the sample database is calculated, such as Figure 3 As shown, Figure 3 The variable correlation analysis diagram in the database according to the embodiment of the present application is schematically shown. The above correlation calculation can be implemented by using the existing correlation calculation method, which will not be repeated here. The above data cleaning can also be performed by using methods such as outlier processing and missing value processing, which are not limited in this embodiment.

[0103] Then, based on the new sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models.

[0104] By cleaning the sample database, irrelevant data in the sample database can be removed, which helps to train a more accurate prediction model. By performing correlation analysis and cleaning operations on the machine learning feature database, the accuracy of subsequent model predictions can be improved.

[0105] In some embodiments, the step of training a plurality of preset machine learning models based on the sample database to obtain a plurality of initial prediction models comprises the following steps:

[0106] Step A1: Divide the sample database into k subsets to obtain k sub-sample sets;

[0107] In this embodiment, the above division may be a random division into k mutually exclusive subsets to obtain k sub-sample sets. The value of k is an integer greater than or equal to 2. Each sub-sample set includes at least one sample, and each sample includes characteristic parameters of the hydrogen storage alloy and the hydrogen storage capacity corresponding to various hydrogen storage alloys.

[0108] Step A2: selecting k-1 sub-sample sets from the k sub-sample sets as training sets and the remaining sub-sample sets as validation sets, training each machine learning model based on the training sets, and testing the trained machine learning model using the validation sets;

[0109] In this embodiment, one sub-sample set is selected as a validation set, and the remaining sub-sample sets are selected as training sets, and then each machine learning model is trained separately. The training process includes using the training set to train the model, and then using the validation set to test the trained machine learning model.

[0110] Step A3: Repeat steps A1-A2 until every sample in the sample database is traversed as a validation set to obtain multiple initial prediction models.

[0111] In this embodiment, steps A1-A2 are repeated until every sample in the sample database is traversed as a validation set, that is, repeated k times, and each sub-sample set is used as a validation set to participate in the training of each machine learning model. More training times help to obtain a more accurate initial prediction model.

[0112] Step 230: determining an optimal initial prediction model from the multiple initial prediction models;

[0113] Wherein, determining the optimal initial prediction model from the multiple initial prediction models includes:

[0114] First, each initial prediction model is used to predict the verification set to obtain the prediction results corresponding to each initial prediction model;

[0115] In this embodiment, the sample data in the validation set includes characteristic parameters of the hydrogen storage alloy, and the characteristic parameters of the hydrogen storage alloy are respectively input into each initial prediction model to obtain a prediction result. Since the number of samples in the validation set may be multiple, the prediction result obtained is also the prediction result of multiple samples.

[0116] Then, based on the prediction results and actual results corresponding to the various initial prediction models, an optimal initial prediction model is determined.

[0117] In this embodiment, the above-mentioned actual results refer to the hydrogen storage capacity in the sample data in the validation set. After obtaining the predicted results and the actual results, the optimal initial prediction model is determined according to the regression performance evaluation index between the predicted results and the actual results of each initial prediction model. The above-mentioned regression performance evaluation index can be set according to actual needs. For example, the error and correlation between the actual results and the predicted results are calculated, so as to select the initial prediction model with small error and large correlation as the optimal initial prediction model. Taking multiple machine learning models including linear regression LR, linear regression variant LASSO, gradient boosting XGBoost, random forest RF, support vector machine SVM, and Stacking as examples, the accuracy of the machine learning models can be evaluated separately, specifically, the performance can be evaluated separately from evaluation indicators such as mean square error (MSE), mean absolute error (MAE), determination coefficient (R2), and Pearson correlation coefficient (PC) PC, such as Figure 4 As shown, Figure 4 The schematic diagram shows the accuracy evaluation diagram of different initial prediction models according to the embodiment of the present application, and finally selects the optimal initial prediction model. Figure 4 It can be seen that the initial prediction model trained by the machine learning model XGBoost is the optimal initial prediction model.

[0118] Step 240: Based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model;

[0119] In this embodiment, the Ada ensemble model is a powerful machine learning algorithm that improves prediction performance by combining multiple weak learners into a strong learner. The Ada ensemble model includes the Ada_XGBoost model, the Ada_SVM model, etc., which is equivalent to further improving the performance of the Ada ensemble model based on the optimal initial prediction model. For example, the Ada_SVM model can further improve the performance based on the SVM model, and the Ada_XGBoost model can further improve the performance based on the XGBoost model.

[0120] The above establishment process includes: taking the optimal initial prediction model as a weak learner. Initializing the Ada integrated model: setting the parameters of the Ada integrated model, such as the number of iterations, learning rate, etc. The number of iterations determines the number of weak learners trained, and the learning rate controls the contribution of each weak learner. Iterative training process: In each iteration, a weak learner is trained according to the current sample weight distribution. The error rate of the weak learner is calculated, and its weight is calculated based on the error rate. The weak learner with a lower error rate has a higher weight. Update the sample weights so that the weights of the samples that are misclassified increase and the weights of the samples that are correctly classified decrease. In this way, subsequent weak learners will pay more attention to samples that are difficult to classify. Combine weak learners: Combine multiple weak learners according to their weights to obtain the final Ada integrated model. Train the integrated model: The sample database can be divided into a training set and a validation set, and the Ada integrated model is trained using the training set. The output of the optimal initial prediction model is used as one of the input features of the weak learner and trained together with the original features. Validation set adjustment: During the training process, use the validation set to monitor the performance of the model and adjust the parameters and hyperparameters of the model based on the results of the validation set. For example, if the performance on the validation set no longer improves, you can stop training early to prevent overfitting.

[0121] When judging whether the training is complete, the performance can be evaluated by using the mean square error (MSE), mean absolute error (MAE), determination coefficient (R2) and Pearson correlation coefficient (PC) PC and other evaluation indicators. Taking the optimal initial prediction model obtained by the XGBoost model as an example, please refer to Figure 5-Figure 8 , Figure 5 The MSE diagram of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown. Figure 6 The MAE diagram of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown. Figure 7 A PC diagram of the accuracy index of the ad a_XGB integrated prediction model according to an embodiment of the present application is schematically shown. Figure 8The R2 graph of the accuracy index of the ada_XGB integrated prediction model according to the embodiment of the present application is schematically shown. When the MSE and MAE values ​​are particularly low, and the R2 and PC values ​​are particularly high, the training is judged to be complete, and a performance prediction model can be obtained. At this time, the optimal prediction model is ada_XGBoost, with MSE reaching 0.06, MAE reaching 0.15, R2 reaching 0.75, and PC reaching 0.88. The optimal parameter combination in the model is {n_estimators=45, max_depth=4, learning_rate=0.5, random_state=0, C=1.0, kernel='rbf', gamma='scale'}.

[0122] Step 250: Based on the performance prediction model, perform hydrogen storage prediction on the hydrogen storage alloy to be predicted, and obtain a hydrogen storage prediction result.

[0123] In the present embodiment, the above-mentioned hydrogen storage prediction for the hydrogen storage alloy to be predicted can first obtain the characteristic parameters of the hydrogen storage alloy to be predicted. The above-mentioned characteristic parameters are the same as the characteristic parameters in the above-mentioned step 210, and will not be repeated here. The characteristic parameters of the hydrogen storage alloy to be predicted are input into the performance prediction model, and the hydrogen storage prediction result is output. It should be noted that the characteristic parameters of the hydrogen storage alloy to be predicted can be obtained by performing first-principles calculations on the hydrogen storage alloy to be predicted to obtain the element crystal parameters of the hydrogen storage alloy to be predicted, and then based on the element crystal parameters of the hydrogen storage alloy to be predicted, the characteristic parameters of the hydrogen storage alloy to be predicted are calculated according to the preset characteristic parameter calculation formula. In specific implementation, the hydrogen storage alloy to be predicted can be input, and then the program automatically calculates the characteristic parameters of the hydrogen storage alloy to be predicted according to the above-mentioned steps, and then the characteristic parameters of the hydrogen storage alloy to be predicted are input into the performance prediction model, and the hydrogen storage prediction result is output. During the prediction, if the hydrogen storage alloy to be predicted is C 14 Crystal phase, you can use the performance prediction model trained by the sample database of the same structure to predict, or you can use the performance prediction model trained by the sample database of similar structure to predict. Among them, the performance prediction model trained by the sample database of the same structure is the most accurate. Fig.10 , Fig.10 The Ti according to the embodiment of the present application is schematically shown. 0.75 Zr 0.35 Mn 2-y-z Cr y Ni z Cu i Schematic diagram of hydrogen storage prediction results.

[0124] It should be noted that the characteristic parameters of the hydrogen storage alloy in the sample database do not include temperature variables, and when the hydrogen storage capacity corresponding to the hydrogen storage alloy is the hydrogen storage capacity corresponding to the hydrogen storage alloy at different temperatures, the obtained hydrogen storage prediction result may be a hydrogen storage capacity prediction result including all temperatures. The characteristic parameters of the hydrogen storage alloy in the sample database include temperature variables, and when the hydrogen storage capacity corresponding to the hydrogen storage alloy is the hydrogen storage capacity corresponding to the hydrogen storage alloy at different temperatures, the obtained hydrogen storage prediction result is a hydrogen storage capacity prediction result under the corresponding temperature system.

[0125] For example, taking Ti-based hydrogen storage alloy as an example, the performance prediction model can be used to predict the alloy in the collected sample database, and the predicted value and the actual value are as follows: Fig. 9 As shown, Fig. 9 A schematic diagram showing a comparison of the predicted value and the actual value of the ada_XGB integrated prediction model according to an embodiment of the present application is shown. Fig. 9 It can be seen that the relative error is as low as within 15%, indicating that the performance prediction model can achieve high-precision prediction of the hydrogen storage capacity of hydrogen storage alloys.

[0126] In the above implementation process, a sample database is obtained, the sample database includes characteristic parameters of various hydrogen storage alloys and the corresponding hydrogen storage capacity of various hydrogen storage alloys, and the various hydrogen storage alloys have different compositions but the same crystal structure; based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models; the optimal initial prediction model is determined from the multiple initial prediction models; based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model; based on the performance prediction model, hydrogen storage prediction is performed on the predicted hydrogen storage alloy to obtain a hydrogen storage prediction result. By selecting hydrogen storage alloys with different compositions and the same crystal structure as the sample database, the optimal initial prediction model is selected from multiple initial prediction models trained by multiple machine learning models through cross-validation of the machine learning model, and the Ada integrated model is further used for parameter iterative screening, so that the rapid and accurate prediction of the hydrogen storage capacity of the hydrogen storage alloy can be achieved, and the accuracy of the prediction model prediction is improved. Compared with the empirical criterion method, linear fitting method and traditional machine learning method in the prior art, the method is more efficient and accurate, significantly reducing the cost of experimental screening and shortening the experimental cycle. The sample database contains hydrogen storage alloys of different compositions with the same crystal structure, which can eliminate the influence of crystal structure changes caused by external factors of alloy composition on the formation of hydrogen storage performance, further improving the prediction accuracy.

[0127] In some embodiments, the following steps are also included:

[0128] Firstly, the multi-element hydrogen storage alloy to be predicted is obtained;

[0129] In this embodiment, the multi-element hydrogen storage alloy to be predicted refers to determining the alloy composition, that is, which elements are included, but not determining the alloy ratio.

[0130] Then, the alloy components in the multi-component hydrogen storage alloy to be predicted are divided according to preset intervals to obtain a plurality of alloy ratio combinations;

[0131] In this embodiment, the composition ratio of the multi-element hydrogen storage alloy is divided into x% intervals to generate multiple alloy ratio combinations. For example, if TiMn2-xCrx is divided into 5% ratio, TiMn2, TiMn 1.9 Cr, etc., wherein x is set to an interval of 2x0.05. The above preset intervals can be set according to specific circumstances.

[0132] Then, based on the performance prediction model, hydrogen storage prediction is performed on each alloy ratio combination to obtain hydrogen storage prediction results of multiple alloy ratio combinations;

[0133] In this embodiment, the characteristic parameters corresponding to each alloy component ratio combination are automatically calculated by the program, and these calculated characteristic parameters are substituted into the previously trained performance prediction model to obtain the hydrogen storage prediction result under each alloy ratio. The above process of hydrogen storage prediction is the same as the step in step 250, and will not be repeated here. It should be noted that the required hydrogen storage temperature can also be determined first, so as to predict the hydrogen storage prediction result under the required hydrogen storage temperature.

[0134] Finally, based on the hydrogen storage prediction results of the plurality of alloy ratio combinations, the alloy composition with the best hydrogen storage performance is obtained.

[0135] In this embodiment, according to the hydrogen storage prediction results of multiple alloy ratio combinations, the alloy composition with the best hydrogen storage performance corresponding to the predicted multi-element hydrogen storage alloy can be determined, so that which alloy composition ratio has the best hydrogen storage density can be determined. In the subsequent experimental preparation, metal elements are added according to this alloy composition ratio. For example, if the alloy composition with the best hydrogen storage performance corresponding to the predicted multi-element hydrogen storage alloy is determined to be TiMn2, only Ti and Mn need to be added in the experimental preparation, and the amount of material is 1:2. If the alloy composition with the best hydrogen storage performance corresponding to the predicted multi-element hydrogen storage alloy is determined to be TiMn2, only Ti and Mn need to be added in the experimental preparation, and the amount of material is 1:2. 0.9 Zr 0.2 MnCr, to prepare the metal of this component, add Ti, Zr, Mn, and Cr metal powders in a ratio of 0.9 to 0.2 to 1 to 1 according to the amount of substance.

[0136] By obtaining a multi-element hydrogen storage alloy to be predicted; dividing the alloy components in the multi-element hydrogen storage alloy to be predicted according to preset intervals to obtain multiple alloy ratio combinations; based on the performance prediction model, hydrogen storage prediction is performed on each alloy ratio combination to obtain hydrogen storage prediction results of multiple alloy ratio combinations; based on the hydrogen storage prediction results of the multiple alloy ratio combinations, the alloy components with the best hydrogen storage performance are obtained. The alloy components with the best hydrogen storage density can be screened out in high throughput, and the hydrogen storage density corresponding to all element ratios of the alloy and the element ratio under the maximum hydrogen storage density can be output by only inputting the elements of the alloy without inputting the specific element ratio, thereby facilitating rapid experimental preparation and saving costs.

[0137] In some embodiments, the following steps are also included:

[0138] First, obtaining the experimental hydrogen storage results corresponding to the alloy composition with the best hydrogen storage performance;

[0139] In this embodiment, after obtaining the alloy composition with the best hydrogen storage performance, experimental preparation can be carried out, and the hydrogen storage performance can be experimentally measured to obtain the experimental hydrogen storage results.

[0140] Then, based on the experimental hydrogen storage results, the hydrogen storage prediction results corresponding to the alloy composition with the best hydrogen storage performance are verified to obtain verification results.

[0141] In this embodiment, the above verification may be to compare the difference between the experimental hydrogen storage result and the hydrogen storage prediction result, thereby verifying whether the hydrogen storage prediction result is certain, so as to obtain a verification result.

[0142] Based on the experimental hydrogen storage results of the alloy composition with the best hydrogen storage performance, the hydrogen storage prediction results corresponding to the alloy composition with the best hydrogen storage performance can be verified, thereby determining the accuracy of the hydrogen storage prediction results. It is possible to quickly and accurately predict high-capacity alloys at different temperatures, and then experimentally verify the alloy powder with high hydrogen storage performance, significantly reducing the cost of experimental screening and shortening the experimental cycle, providing an excellent way to rationally design base alloy powders with high hydrogen storage performance at different temperatures.

[0143] Please see Fig.11 , Fig.11The design flow chart of Ti-based hydrogen storage alloy according to the embodiment of the present application is schematically shown; first, density functional theory (DFT) is used to calculate the information such as the lattice constants a, b, c and bulk modulus of the elements, and then characteristic parameters are calculated based on this information as characteristic quantities, and then a performance prediction model constructed by a machine learning algorithm is used for high-throughput screening to determine the alloy composition with the best hydrogen storage performance, and then experimental verification is performed based on the alloy composition with the best hydrogen storage performance to complete the design of the entire Ti-based hydrogen storage alloy. Rapid and accurate prediction of high-capacity Ti-based alloys at different temperatures is achieved, and then experimental verification of Ti-based alloy powders with high hydrogen storage performance is performed, which significantly reduces the cost of experimental screening and shortens the experimental cycle, providing an excellent way to rationally design Ti-based alloy powders with high hydrogen storage performance at different temperatures.

[0144] Figure 1 FIG. 1 is a flow chart of a method for predicting the hydrogen storage performance of a hydrogen storage alloy in one embodiment. It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0145] Please see Fig.12 , Fig.12 The schematic diagram of the structure of a hydrogen storage performance prediction device for a hydrogen storage alloy according to an embodiment of the present application is schematically shown. This embodiment provides a hydrogen storage performance prediction device for a hydrogen storage alloy, including an acquisition module 410, a training module 420, a determination module 430, an establishment module 440 and a prediction module 450, wherein:

[0146] An acquisition module 410 is used to acquire a sample database, wherein the sample database includes characteristic parameters of various hydrogen storage alloys and hydrogen storage capacities corresponding to various hydrogen storage alloys;

[0147] A training module 420 is used to train multiple preset machine learning models based on the sample database to obtain multiple initial prediction models;

[0148] A determination module 430, configured to determine an optimal initial prediction model from the multiple initial prediction models;

[0149] Establishing module 440, used to establish an Ada integrated model corresponding to the optimal initial prediction model based on the sample database to obtain a performance prediction model;

[0150] The prediction module 450 is used to perform hydrogen storage prediction on the hydrogen storage alloy to be predicted based on the performance prediction model to obtain a hydrogen storage prediction result.

[0151] The hydrogen storage performance prediction device of the hydrogen storage alloy includes a processor and a memory. The above-mentioned acquisition module 410, training module 420, determination module 430, establishment module 440 and prediction module 450 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0152] The processor contains a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the prediction precision and accuracy can be improved by adjusting the kernel parameters.

[0153] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0154] An embodiment of the present invention provides a machine-readable storage medium on which a program is stored. When the program is executed by a processor, the method for predicting the hydrogen storage performance of the hydrogen storage alloy is implemented.

[0155] An embodiment of the present invention provides a processor, which is used to run a program, wherein the method for predicting the hydrogen storage performance of the hydrogen storage alloy is executed when the program is run.

[0156] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.13As shown. The computer device includes a processor A01, a network interface A02, a display screen A04, an input device A05 and a memory (not shown in the figure) connected through a system bus. Among them, the processor A01 of the computer device is used to provide computing and control capabilities. The memory of the computer device includes an internal memory A03 and a non-volatile storage medium A06. The non-volatile storage medium A06 stores an operating system B01 and a computer program B02. The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 in the non-volatile storage medium A06. The network interface A02 of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor A01, a method for predicting the hydrogen storage performance of a hydrogen storage alloy is implemented. The display screen A04 of the computer device can be a liquid crystal display or an electronic ink display, and the input device A05 of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0157] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In one embodiment, the method and apparatus for predicting the hydrogen storage performance of the hydrogen storage alloy provided in the present application can be implemented in the form of a computer program. The computer program can be used in the following manner: Fig.13 The computer device shown in the figure is run. The memory of the computer device can store various program modules constituting the hydrogen storage performance prediction method device of the hydrogen storage alloy, for example, Fig.12 The acquisition module 410, training module 420, determination module 430, establishment module 440 and prediction module 450 are shown. The computer program composed of various program modules enables the processor to execute the steps of the method for predicting the hydrogen storage performance of the hydrogen storage alloy of each embodiment of the present application described in this specification.

[0159] Fig.13 The computer device shown can be Fig.12 The acquisition module 410 in the hydrogen storage performance prediction method device of the hydrogen storage alloy shown executes step 210. The computer device may execute step 220 through the training module 420. The computer device may execute step 230 through the determination module 430. The computer device may execute step 240 through the establishment module 440. The computer device may execute step 250 through the prediction module 450.

[0160] An embodiment of the present application provides an electronic device, which includes: at least one processor; a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the above-mentioned method for predicting the hydrogen storage performance of the hydrogen storage alloy by executing the instructions stored in the memory, and the processor implements the following steps when executing the instructions:

[0161] Acquire a sample database, wherein the sample database includes characteristic parameters of a plurality of hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure;

[0162] Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models;

[0163] Determining an optimal initial prediction model from the multiple initial prediction models;

[0164] Based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model;

[0165] Based on the performance prediction model, hydrogen storage prediction is performed on the hydrogen storage alloy to be predicted to obtain a hydrogen storage prediction result.

[0166] In one embodiment, obtaining a sample database includes:

[0167] Obtain various hydrogen storage alloys and the corresponding hydrogen storage capacities of various hydrogen storage alloys;

[0168] First-principles calculations were performed on various hydrogen storage alloys to obtain the element crystal parameters of various hydrogen storage alloys;

[0169] Based on the element crystal parameters of the various hydrogen storage alloys, characteristic parameters of the various hydrogen storage alloys are calculated according to preset characteristic parameter calculation formulas;

[0170] Based on the characteristic parameters of the various hydrogen storage alloys and the hydrogen storage capacities corresponding to the various hydrogen storage alloys, a sample database is obtained.

[0171] In one embodiment, the multiple preset machine learning models are trained based on the sample database to obtain multiple initial prediction models, including:

[0172] Cleaning the data in the sample database to obtain a new sample database;

[0173] Based on the new sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models.

[0174] In one embodiment, the characteristic parameter includes temperature, and the hydrogen storage capacities corresponding to the various hydrogen storage alloys are the hydrogen storage capacities corresponding to the various hydrogen storage alloys at different temperatures.

[0175] In one embodiment, the multiple preset machine learning models are trained based on the sample database to obtain multiple initial prediction models, including:

[0176] A1: Divide the sample database into k subsets to obtain k sub-sample sets;

[0177] A2: Select k-1 sub-sample sets from the k sub-sample sets as training sets, and the remaining sub-sample sets as validation sets. Based on the training sets, train each machine learning model respectively, and use the validation sets to test the trained machine learning model.

[0178] A3: Repeat steps A1-A2 until every sample in the sample database is traversed as a validation set to obtain multiple initial prediction models.

[0179] In one embodiment, determining the optimal initial prediction model from the multiple initial prediction models includes:

[0180] Using each initial prediction model to predict the validation set, respectively, to obtain prediction results corresponding to each initial prediction model;

[0181] Based on the prediction results and actual results corresponding to each of the initial prediction models, respectively calculate the regression performance evaluation index of each initial prediction model;

[0182] Based on the regression performance evaluation indicators of the various initial prediction models, an optimal initial prediction model is determined.

[0183] In one embodiment, it further includes:

[0184] Obtaining a multi-element hydrogen storage alloy to be predicted;

[0185] Dividing the alloy components in the multi-component hydrogen storage alloy to be predicted according to preset intervals to obtain a plurality of alloy ratio combinations;

[0186] Based on the performance prediction model, hydrogen storage prediction is performed on each alloy ratio combination to obtain hydrogen storage prediction results of multiple alloy ratio combinations;

[0187] Based on the hydrogen storage prediction results of the plurality of alloy ratio combinations, an alloy composition with optimal hydrogen storage performance is obtained.

[0188] In one embodiment, it further includes:

[0189] Obtaining the experimental hydrogen storage results corresponding to the alloy composition with the best hydrogen storage performance;

[0190] Based on the experimental hydrogen storage results, the hydrogen storage prediction results corresponding to the alloy composition with the optimal hydrogen storage performance are verified to obtain verification results.

[0191] This embodiment provides a vehicle, comprising the above-mentioned hydrogen storage performance prediction device of the hydrogen storage alloy or the above-mentioned electronic device or the above-mentioned machine-readable storage medium.

[0192] In this embodiment, the above-mentioned vehicle can be an engineering vehicle or a passenger car. By arranging a hydrogen storage performance prediction device of a hydrogen storage alloy or the above-mentioned electronic device or the above-mentioned machine-readable storage medium on the vehicle, a rapid and accurate prediction of the hydrogen storage capacity of the hydrogen storage alloy can be achieved, thereby improving the accuracy of the prediction model prediction, thereby facilitating an accurate assessment of the power situation of the vehicle. The sample database has hydrogen storage alloys of different components with the same crystal structure, which can eliminate the influence of crystal structure changes caused by external factors of the alloy composition on the formation of hydrogen storage performance, further improving the prediction accuracy.

[0193] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0194] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0195] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0196] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0197] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0198] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0199] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0200] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0201] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for predicting the hydrogen storage performance of a hydrogen storage alloy, characterized in that: include: Acquire a sample database, wherein the sample database includes characteristic parameters of a plurality of hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure; Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models; Determining an optimal initial prediction model from the multiple initial prediction models; Based on the sample database, an Ada integrated model corresponding to the optimal initial prediction model is established to obtain a performance prediction model; Based on the performance prediction model, hydrogen storage prediction is performed on the hydrogen storage alloy to be predicted to obtain a hydrogen storage prediction result.

2. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 1, characterized in that: The step of obtaining a sample database includes: Obtain various hydrogen storage alloys and the corresponding hydrogen storage capacities of various hydrogen storage alloys; First-principles calculations were performed on various hydrogen storage alloys to obtain the element crystal parameters of various hydrogen storage alloys; Based on the element crystal parameters of the various hydrogen storage alloys, characteristic parameters of the various hydrogen storage alloys are calculated according to preset characteristic parameter calculation formulas; Based on the characteristic parameters of the various hydrogen storage alloys and the hydrogen storage capacities corresponding to the various hydrogen storage alloys, a sample database is obtained.

3. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 2, characterized in that: Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models, including: Cleaning the data in the sample database to obtain a new sample database; Based on the new sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models.

4. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 1, characterized in that: The characteristic parameter includes temperature, and the hydrogen storage capacity corresponding to the various hydrogen storage alloys is the hydrogen storage capacity corresponding to the various hydrogen storage alloys at different temperatures.

5. The method for predicting hydrogen storage performance of hydrogen storage alloy according to claim 1, characterized in that: Based on the sample database, multiple preset machine learning models are trained to obtain multiple initial prediction models, including: A1: Divide the sample database into k subsets to obtain k sub-sample sets; A2: Select k-1 sub-sample sets from the k sub-sample sets as training sets, and the remaining sub-sample sets as validation sets. Based on the training sets, train each machine learning model respectively, and use the validation sets to test the trained machine learning model. A3: Repeat steps A1-A2 until every sample in the sample database is traversed as a validation set to obtain multiple initial prediction models.

6. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 5, characterized in that: The step of determining an optimal initial prediction model from the multiple initial prediction models comprises: Using each initial prediction model to predict the validation set, respectively, to obtain prediction results corresponding to each initial prediction model; Based on the prediction results and actual results corresponding to each of the initial prediction models, respectively calculate the regression performance evaluation index of each initial prediction model; Based on the regression performance evaluation indicators of the various initial prediction models, an optimal initial prediction model is determined.

7. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 1, characterized in that: Also includes: Obtaining a multi-element hydrogen storage alloy to be predicted; Dividing the alloy components in the multi-component hydrogen storage alloy to be predicted according to preset intervals to obtain a plurality of alloy ratio combinations; Based on the performance prediction model, hydrogen storage prediction is performed on each alloy ratio combination to obtain hydrogen storage prediction results of multiple alloy ratio combinations; Based on the hydrogen storage prediction results of the plurality of alloy ratio combinations, an alloy composition with optimal hydrogen storage performance is obtained.

8. The method for predicting hydrogen storage performance of hydrogen storage alloys according to claim 7, characterized in that: Also includes: Obtaining the experimental hydrogen storage results corresponding to the alloy composition with the best hydrogen storage performance; Based on the experimental hydrogen storage results, the hydrogen storage prediction results corresponding to the alloy composition with the optimal hydrogen storage performance are verified to obtain verification results.

9. A device for predicting hydrogen storage performance of a hydrogen storage alloy, characterized in that: include: An acquisition module, used to acquire a sample database, wherein the sample database includes characteristic parameters of a plurality of hydrogen storage alloys and hydrogen storage capacities corresponding to the various hydrogen storage alloys, wherein the various hydrogen storage alloys have different compositions but the same crystal structure; A training module, used to train multiple preset machine learning models based on the sample database to obtain multiple initial prediction models; A determination module, used to determine an optimal initial prediction model from the multiple initial prediction models; An establishment module is used to establish an Ada integrated model corresponding to the optimal initial prediction model based on the sample database to obtain a performance prediction model; The prediction module is used to perform hydrogen storage prediction on the hydrogen storage alloy to be predicted based on the performance prediction model to obtain a hydrogen storage prediction result.

10. An electronic device, characterized in that: The electronic device includes: at least one processor; a memory connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the method for predicting the hydrogen storage performance of the hydrogen storage alloy according to any one of claims 1 to 8 by executing the instructions stored in the memory.

11. A machine-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by a processor, the processor is configured to execute the method for predicting hydrogen storage performance of a hydrogen storage alloy according to any one of claims 1 to 8.

12. A vehicle, characterized in that: A hydrogen storage performance prediction device for a hydrogen storage alloy according to claim 9, an electronic device according to claim 10, or a machine-readable storage medium according to claim 11.

Images