Construction method of TiH binary system machine learning potential

Abstract

The invention discloses a construction method of a TiH binary system machine learning potential. The construction method comprises the following steps: constructing a reference electronic structure data set by adopting first principle calculation software VASP; training and testing an artificial neural network potential-ANN potential model on the reference data set by using the artificial neural network model; vASP and ANN potential obtained through training are used for calculating and comparing a Dimer curve, a lattice constant, an elastic modulus and the like for performance testing; and performing targeted improvement on data with large comparison error between the ANN model and the DFT result in model test and performance test results by increasing related configuration weights until an ANN potential model conforming to the precision is obtained. The machine learning potential model constructed by the invention can better fit the interatomic interaction potential energy surface of the TiH binary system, solves the problems of small time-space scale and low precision of the traditional empirical potential in the calculation of the first principle, and provides a technical basis for the high-precision and large-size simulation and research of the TiH binary system.

Description

Technical Field

[0001] The invention relates to the field of TiH alloys and proposes a method for constructing a machine learning potential of a TiH binary system. Background Art

[0002] Titanium is an excellent hydrogen storage metal with good compatibility with hydrogen. The low hydrogen solubility limit makes α-Ti very easy to form hydrides in a hydrogen-rich environment, and its hydrogen absorption density can reach 9.2×10 22 atom / cm 3 , and the absorption of hydrogen is reversible. Therefore, titanium hydride is widely used in hydrogen storage, preparation materials of solid deuterium or tritium targets for neutron generators, and hot hydrogen processing of titanium alloys. Therefore, understanding the structure and physicochemical properties of titanium hydride in the TiH binary system is of great significance for basic research and engineering application research.

[0003] At present, the research on the physical and chemical phenomena of TiH binary system mainly focuses on the influence of surface oxide and Ti matrix integrity on H adsorption, migration and hydrogenation, as well as the influence of hydride formation on embrittlement and local corrosion resistance. However, due to the limitations of existing characterization and modeling techniques, there are still many knowledge gaps and challenges. For example, although DFT simulation based on density functional theory has very high computational accuracy, its time complexity is O(N 3 ) That is, the amount of calculation is proportional to the cube of the number of atoms in the structure. Therefore, the time and space scales of the system that can be studied are relatively small. Complex defects in the system, such as dislocation networks, stacking layers, and interfaces / phase boundaries, are usually difficult to solve using DFT alone. Classical molecular dynamics simulation is an alternative method that can study larger time and space scales. However, the calculation results of atomic potential energy and forces using traditional empirical potentials often have large deviations. There has been a phenomenon in which the molecular dynamics simulation based on traditional empirical potentials is inconsistent with the ab initio calculation results for the diffusion of H in titanium hydrides. The lack of a reliable interatomic interaction potential in the TiH binary system affects the prediction of hydrogenation phenomena.

[0004] The molecular dynamics simulation using machine learning potential has high precision close to that of first-principles calculations and high efficiency of traditional semi-empirical potentials. The machine learning algorithm can better fit the potential energy surface of atomic interactions in the TiH binary system, solve the problems of small time and space scales in first-principles calculations and low precision of traditional empirical potentials, and provide a technical basis and theoretical guidance for high-precision large-scale simulation and experimental development of the TiH binary system. Summary of the invention

[0005] In view of the problem that the current TiH binary system lacks molecular simulation tools that take both accuracy and scale into consideration, the present invention provides a method for constructing a machine learning potential for the TiH binary system. A reference electronic structure data set for the TiH binary system is constructed based on first-principles calculation software. An artificial neural network is trained on the reference data set by using the AENET software package to obtain an ANN potential model and perform model testing. VASP and the trained ANN potential are used to perform performance testing on dimer curves, EV curves, RDF curves, lattice constants, elastic moduli and defect formation energy. Targeted improvements are made to poorly performing performances until a machine learning potential model for the TiH binary system that meets accuracy requirements is obtained.

[0006] To solve the technical problem, the present invention adopts the following technical solution:

[0007] A method for constructing a machine learning potential for a TiH binary system is performed in the following steps:

[0008] Step 1: Use the first-principles software VASP to construct a basic reference electronic structure data set including dimer structure, stable crystal configuration of Ti and common TiH compounds, random structure, and AIMD trajectory extraction structure. The specific method is as follows:

[0009] 11) Dimer structure: Create a supercell according to the VASP structure file POSCAR format, fix the position of one atom, and move another atom in one direction until the entire Dimer curve potential well is covered.

[0010] 12) Stable crystal configurations of Ti and common TiH compounds: Construct HCP, BCC, FCC, Diamond, SC, A12, A13, A15 structures of Ti and initial configurations of Ti2H, TiH, TiH2 according to the VASP structure file POSCAR format; use VASP for high-precision structural optimization; on this basis, apply multiple strains to the corresponding directions of the crystal according to the deformation matrix of different crystal systems to form a large number of perturbation configurations of the stable structure of the crystal system.

[0011] 13) Random structure: The ab initio random structure search program AIRSS was used to perform structural searches for various possible symmetries and arrangements under different TiH component ratios to generate a large number of random structures. VASP was used to perform coarse-precision optimization on these random structures, and structures were extracted from the optimization trajectory at regular intervals.

[0012] 14) AIMD trajectory structure extraction: Constant temperature and constant volume AIMD simulations are performed for H2 at different temperatures and pressures, Ti, TiH, TiH2 perfect crystals at different temperatures and compositions, and different types of defect structures, and a structure is extracted from the MD simulation trajectory at a certain number of steps.

[0013] Step 2: Use the AENET software package to train and test the ANN (artificial neural network) potential model on the basic reference electronic structure data set. The specific method is as follows:

[0014] 21) Use the generate.x tool in the AENET software package to compile the structures and energies in the reference electronic structure dataset into a set in a specific format.

[0015] 22) Debug the parameters in the model, including the cutoff radius, the shortest distance between atoms, the order of the descriptor basis function expansion, the size of the neural network, and the learning rate, and set the number of iterations and the proportion of the test set; use the train.x tool to divide the data set into a training set and a test set, and train the ANN potential model on the training set.

[0016] 23) Use predict.x to test the ANN potential model based on the reference electronic structure data (E DFT ) and ANN potential prediction results (E ANN ), where

[0017] When the RMSE and MAE of the ANN in both the training and test sets converged to below 10 meV, the model was considered to have DFT accuracy.

[0018] Step 3, using VASP and the trained ANN potential model to calculate and compare the dimer curve, EV curve, RDF curve, equilibrium lattice constant, elastic modulus and defect formation energy as a performance test;

[0019] 31) For structures with different distances between two atoms, VASP and ANN potentials were used to perform single-point calculations, extract energy-distance information, and obtain Dimer curves with the horizontal coordinate of VASP and ANN being the distance between the two atoms and the vertical coordinate being the total energy of the system.

[0020] 32) The stable crystal configuration of Ti and common TiH compounds was optimized. The volume of the optimized structure was expanded and compressed by changing the normalization coefficient in POSC AR with a continuous fixed step length. The single-point calculations of structures with different volumes were performed using VASP and ANN potential, respectively. The volume of each atom was used as As the abscissa, the energy volume curve - EV curve is obtained.

[0021] 33) Constant temperature and constant volume AIMD and ANN-MD simulations were performed on the body-centered cubic Ti supercell respectively. The supercell was heated from room temperature to a temperature higher than the melting point of Ti to melt the structure. The last frame structure was taken as the initial structure and a constant temperature simulation was performed at this temperature. The distance between each particle in the system and other particles was calculated according to the position information of atoms in the trajectory file XDATCAR generated by AIMD and the trajectory file dump generated by ANN-MD. A suitable distance range and step size were selected. For each distance interval, the number of particle pairs in this interval was counted. The count result was divided by the total number of particle pairs and the volume of the distance interval to obtain the value of the radial distribution function g(r), and the RDF curve with the horizontal axis being the distance and the vertical axis being the g(r) was obtained.

[0022] 34) The close-packed hexagonal phase structure of Ti was exported from the Material Studio software structure library as a .cif file format, and then converted into the POSCAR format using the VESTA software. The structure was then optimized with high precision using the VASP software to obtain the equilibrium lattice constant.

[0023] 35) According to the deformation matrix of different crystal systems, quasi-continuous strain is applied to the corresponding direction of the crystal structure after high-precision optimization. VASP and LAMMPS are used to perform single-point calculations on the structure under different strains. Nonlinear curve fitting is performed based on the relationship between strain and energy to obtain the crystal elastic constants, which are then brought into the bulk modulus calculation formula to obtain the bulk modulus B.

[0024] For example, the deformation matrix of the hexagonal crystal system is:

[0025]

[0026] Its crystal elastic constant is C 11 , C 12 , C 13 , C 33 and C 44 , the calculation formula of its bulk modulus B is:

[0027]

[0028] 36) The structure after high-precision optimization in step 34) is expanded by VESTA software to obtain a supercell, and a single vacancy, a single tetrahedral interstitial H atom, and a single octahedral interstitial H atom structure are established on the basis of the supercell, and then various permutations and combinations are considered to construct double vacancies, double tetrahedral interstitial H atoms, and double octahedral interstitial H atom structures, and the distances between point defects are sorted to obtain the first neighbor, second neighbor, and third neighbor configurations of double vacancies, double tetrahedral interstitial H atoms, and double octahedral interstitial H atoms. VASP and LAMMPS are used to perform single-point calculations on the first to third neighbor supercell structures of perfect crystals, single defects, and double defects, respectively. The energy of the primitive cell Ti and H2 molecules is calculated separately, and the energy is divided by the number of atoms to obtain the chemical potential of Ti and H2; according to the defect formation energy calculation formula E f =E d -E p ±nμ, where E d is the total energy of the defective system, E p is the total energy of a perfect crystal, μ is the chemical potential of the missing atom, and n is the number of defective atoms. The first to third neighbor structure formation energies of single and double defects are obtained and compared with the calculation results of VASP and LAMMPS.

[0029] Step 4: For the data with large errors between the ANN potential model and the DFT results (i.e., VASP calculation results) in the model test and performance test results, targeted improvements are made by increasing the relevant configuration weights until an ANN potential model that meets the accuracy is obtained. The specific method is:

[0030] 41) Set the error limit of VASP and ANN calculation results to δ = |E DFT -E ANN |≤30meV, for data exceeding the error limit, extract and check the configuration, perform optimization or constant temperature and constant volume AIMD and ANN-md simulation, and extract the optimization or simulation trajectory into the data set; for performance with large errors in performance testing, perform targeted constant temperature and constant volume AIMD and ANN-md simulations, and extract the simulation trajectory into the data set;

[0031] 42) Repeat steps 2, 3 and 41) and continue iterating until the error drops to an acceptable range, thereby obtaining an ANN potential model that meets the accuracy requirements.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. The machine learning potential model trained by the present invention can describe the high-dimensional atomic interaction potential energy surface of the TiH binary system very accurately based on the reference data of electronic structure calculation with the calculation accuracy close to DFT;

[0034] 2. Molecular dynamics simulation based on ANN potential model is very efficient, many orders of magnitude faster than DFT calculation and comparable to traditional empirical potential, which enables it to simulate larger time and space scales;

[0035] 3. The nonlinear functional form of the ANN potential is very general and flexible, and does not require any prior knowledge of the physical functional form of the atomic interactions in the TiH system;

[0036] 4. The ANN potential model of the present invention provides a TiH binary system simulation tool that takes both accuracy and scale into consideration. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of common crystal configurations and titanium hydrides of Ti in the present invention, (a) HCP phase structure of Ti, (b) BC C phase structure of Ti, (c) FCC structure of Ti, (d) SC structure of Ti, (e) Diamand structure of Ti, (f) A12 structure of Ti, (g) A13 structure of Ti, (h) A15 structure of Ti, (i) TiH2, (j) TiH, (k) Ti2H;

[0038] Figure 2 It is a figure showing the fitting results of the ANN model of the present invention to the DFT data in the reference data set;

[0039] Figure 3 It is a comparison diagram between the ANN model and DFT calculation of the dimer curve of the present invention;

[0040] Figure 4 1 is a comparison diagram of the EV curve of the present invention between the ANN model and the DFT calculation, (a) is the EV curve of a common crystal configuration of Ti, and (b) is the EV curve of a common titanium hydride;

[0041] Figure 5 It is a comparison diagram between the ANN model and DFT calculation of the radial distribution function of the present invention;

[0042] Figure 6 It is a schematic diagram of the iterative convergence of the loss function in the ANN model training process of the present invention;

[0043] Figure 7 It is a schematic diagram of the complete data set of the present invention, with a total of 35081 structures. DETAILED DESCRIPTION

[0044] In order to more clearly illustrate the technical solution of the present invention, the preferred embodiment of the present invention will be described in detail below in conjunction with the accompanying drawings used. The described embodiment is only one of the preferred embodiments of the present invention, rather than all 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.

[0045] Example 1

[0046] Step 1: Use the first-principles software VASP to construct a basic reference electronic structure data set including dimer structure, stable crystal configuration of Ti and common TiH compounds, random structure, and AIMD trajectory extraction structure. The specific method is as follows:

[0047] 11) Dimer structure: Create a VASP structure file in POSCAR format The supercell fixes the position of an atom and moves in a certain direction at regular intervals (such as ) and move another atom until the entire dimer curve potential well is covered, resulting in a total of 121 dimer structures.

[0048] 12) Stable crystal configurations of Ti and common TiH compounds: According to the VASP structure file POSCAR format, the HCP, BCC, FCC, Diamond, SC, A12, A13, A15 structures of Ti and the initial configurations of Ti2H, TiH, TiH2 are constructed. The stable crystal configurations of some Ti and common TiH compounds are as follows: Figure 1 As shown; VASP is used for high-precision structural optimization; on this basis, multiple strains are applied to the corresponding directions of the crystal according to the deformation matrix of different crystal systems to form a large number of perturbation configurations of the stable structure of the crystal system.

[0049] 13) Random structure: The ab initio random structure search program AIRSS was used to perform structural searches for various possible symmetries and arrangements of TiH with composition ratios of 32:0, 32:1, 16:1, 8:1, 4:1, 2:1, 1:1, and 1:2, generating a large number of random structures. VASP was used to perform coarse-precision optimization on these random structures, extracting one structure every 20 steps from the latter part of the optimization trajectory.

[0050] 14) AIMD trajectory extraction structure: for H2 at 0-1Gpa ​​and 1-2GPa at 200K, 300K, and 500K, HCP phase of Ti, TiH, TiH2 and other perfect crystals at 300K, 600K, 900K, 1200K, 1500K, 1800K, 2100K, and 2400K, c HConstant temperature and constant volume AIMD simulations were performed on HCP phases of Ti with compositions of =0%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, and 40%, and one structure was extracted from the MD simulation trajectory every 20 steps.

[0051] Step 2: Use the AENET software package to train and test the ANN potential (artificial neural network potential) model on the basic reference electronic structure data set. The specific method is as follows:

[0052] 21) Use the generate.x tool of the AENET software package to compile the reference electronic structure data set into a collection of specific formats.

[0053] 22) After debugging the model parameters, the cutoff radius is set to The shortest distance between atoms is The expansion orders of the descriptor basis function are n=5 and l=6, the neural network has two hidden layers: 40×20, the learning rate is 0.001, the number of iterations is set to 20,000 steps, and the test set ratio is 10%. The train.x tool is used to divide the data set into a training set and a test set, and the ANN potential model is trained on the training set.

[0054] 23) Use predict.x to test the ANN potential model. The ANN prediction of the structure formation energy in the entire data set is compared with the DFT results. Figure 2 As shown, the closer the data points are to the y=x line, the more accurate the ANN's prediction of the data is; the more data points are close to the y=x line, the better the ANN fits the configuration space of the reference data set.

[0055] Step 3: Use VASP and the trained ANN potential model to calculate and compare the dimer curve, EV curve, RDF curve, equilibrium lattice constant, elastic modulus and defect formation energy as a performance test. The specific method is as follows:

[0056] 31) For the dimer structures with different distances between two atoms, VASP and trained ANN potential were used to perform single-point calculations, extract energy-distance information, and obtain dimer curves with the horizontal coordinate of VASP and ANN as the distance between the two atoms and the vertical coordinate as the total energy of the system. The results are as follows Figure 3 As shown, the ANN calculation curve is basically consistent with the DFT calculation curve, indicating that ANN has DFT accuracy on the Dimer curve.

[0057] 32) The volume expansion and compression of the stable crystal configuration of Ti and common TiH compounds are carried out. The single-point calculations are performed using VASP and ANN potential for structures with different volumes. As the horizontal axis, the energy volume curve-EV curve is obtained, and the result is as follows Figure 4 As shown, the ANN calculation curve and the DFT calculation curve basically coincide with each other, indicating that ANN has DFT accuracy on the EV curve.

[0058] 33) The constant temperature and constant volume AIMD and ANN-MD simulations were performed on the body-centered cubic Ti supercell. The supercell was heated from room temperature to 2500K to melt the structure. The last frame structure was used as the initial structure and constant temperature simulation was performed at 2500K. According to the atomic position information in the XDATCAR (AIMD) and dump (ANN-MD) files, g(r) was solved to obtain the RDF curve with the horizontal axis as the distance and the vertical axis as the radial distribution function. The results are shown in Figure 5 As shown, ANN basically reproduces the radial distribution function curve of Ti obtained by DFT calculation. The position of each peak, peak width, and the height of the second and third nearest neighbor peaks are well predicted, indicating that ANN has DFT accuracy in the radial distribution function.

[0059] 34) The HCP phase structure of Ti was exported from the Material Studio software structure library as a .cif file format, and then converted to POSCAR format using VESTA software. The structure was optimized with high precision using VASP software to obtain the equilibrium lattice constant and c / a of HCP phase Ti. The results are shown in Table 1. The lattice constant and c / a predicted by ANN are very close to those by DFT, and the c / a ratio is consistent with the experimental value.

[0060] 35) According to the deformation matrix of the hexagonal crystal system, quasi-continuous strain is applied to the corresponding direction of the crystal structure after high-precision optimization. VASP and LAMMPS are used to perform single-point calculations on the structure under different strains. Nonlinear curve fitting is performed based on the relationship between strain and energy to obtain the crystal elastic constant C. 11 , C 12 , C 13 , C 33 and C 44 , the bulk modulus B is obtained according to the bulk modulus calculation formula. The results are shown in Table 1. 13 , C 44 The difference between the experimental value and the DFT calculation result is less than 10 GPa, C 11 , C 12 , C 33 , C 66 The difference is less than 15 GPa, and B is less than 2 GPa. The overall prediction results are close to the experimental values ​​and DFT results.

[0061] 36) The structure optimized with high precision in 34) was expanded by VESTA software to obtain a supercell. On the basis of the supercell, the first neighbor, second neighbor, and third neighbor configurations of single vacancy, single tetrahedral interstitial H atom, single octahedral interstitial H atom, double vacancy, double tetrahedral interstitial H atom, and double octahedral interstitial H atom were established. VASP and LAM MPS were used to perform single-point calculations on the first to third neighbor supercell structures of perfect crystal, single defect, and double defect, respectively. The first to third neighbor structure formation energies of single defect and double defect were obtained according to the formation energy calculation formula. The results are shown in Table 2. The ANN predicted single vacancy formation energy is close to the DFT result; the ANN predicted value of the single OIS-H formation energy is basically consistent with the DFT calculation result; the OIS formation energy is lower than the TIS interstitial formation energy, which is consistent with the DFT and theoretical research results; the ANN predicted values ​​of double vacancies, double TIS-H, and double OIS-H are basically consistent with the DFT calculation results.

[0062] Step 4: For the data with large errors between the ANN potential model and the DFT results in the model test and performance test results, targeted improvements are made by increasing the relevant configuration weights until an ANN potential model that meets the accuracy is obtained. The specific method is:

[0063] 41) For data exceeding the error limit, extract and check the configuration, perform optimization or constant temperature and constant volume AIMD and ANN-MD simulations, and extract the optimization or simulation trajectory into the data set; for performance with large errors in performance testing, perform targeted constant temperature and constant volume AIMD and ANN-MD simulations, and extract the simulation trajectory into the data set.

[0064] 42) Repeat steps 2, 3, and 41) and iterate until the error is reduced to an acceptable range. The MAE and RMSE of the ANN model in the training set and test set are: train-MAE = 0.00361 eV / atom, trian-RMSE = 0.00582 eV / atom, test-MAE = 0.00416 eV / atom, test-RMSE = 0.00824 eV / atom, all of which converge to less than 10 meV. Figure 6 As shown, the model is considered to have DFT accuracy at this time, and the complete data set is as follows Figure 7 shown.

[0065] Table 1: ANN, experimental value, and DFT comparison table of lattice constant, c / a, elastic modulus, and bulk modulus of Ti HCP phase

[0066]

[0067] Note: ANN, Exp, and DFT represent the ANN potential model calculation results, experimental values, and VASP calculation results, respectively. DFT (This work) refers to the DFT calculation results in this invention, and DFT without (This work) refers to the DFT calculation results of others in the references.

[0068] References:

[0069] [1]Barrett C,Massalski T.Structure of Metals,CrystallographicMethods,Principles and Data[J].Pergamon Press,1966.

[0070] [2]E,S,Fisher,et al.Single-Crystal Elastic Moduli and the hcp→bccTransformation in Ti,Zr,and Hf[J].Physical Review,1964.DOI:10.1103 / physrev.135.a482.

[0071] [3]Gibson JS, Srinivasan SG, Baskes MI, et al.A multi-state modifiedembedded atom potential method for titanium[J].ModellingSimul.mater.sci.eng,2017,25(1):015010.DOI:10.1088 / 1361-651X / 25 / 1 / 015010.

[0072] Table 2: Comparison of ANN and DFT calculation results of vacancy and interstitial H atom defect formation energy in the HCP phase of Ti of the present invention

[0073]

[0074] Note: TIS-H represents hydrogen atoms in tetrahedral gaps; OIS-H represents hydrogen atoms in octahedral gaps; V represents vacancy defects; 1NN represents first nearest neighbor. f TIS-H,1NN For example, it represents the defect formation energy of hydrogen atoms in the tetrahedral gaps of the Ti-HCP phase that are first neighbors in distance.

[0075] The above description is only an exemplary embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for constructing a machine learning potential for a TiH binary system, characterized in that: Follow these steps: Step 1: Use the first-principles software VASP to construct a basic reference electronic structure data set including dimer structure, stable crystal configurations of Ti and common TiH compounds, random structures, and AIMD trajectory extracted structures; Step 2: Using the AENET software package to train and test the ANN potential model on the basic reference electronic structure data set; Step 3, using VASP and the trained ANN potential model to calculate and compare the dimer curve, EV curve, RDF curve, equilibrium lattice constant, elastic modulus and defect formation energy as a performance test; Step 4: For data with large errors between the ANN potential model and the DFT results in the model test and performance test results, targeted improvements are made by increasing the relevant configuration weights until an ANN potential model that meets the accuracy requirements is obtained.

2. The method for constructing the TiH binary system machine learning potential according to claim 1, characterized in that: In step 1, the method of constructing the basic reference electronic structure dataset using VASP is as follows: 11) Dimer structure: Create a supercell according to the VASP structure file POSCAR format, fix the position of one atom, and move another atom in one direction until the entire Dimer curve potential well is covered; 12) Stable crystal configurations of Ti and common TiH compounds: construct the HCP, BCC, FCC, Diamond, SC, A12, A13, A15 structures of Ti and the initial configurations of Ti2H, TiH, and TiH2 according to the VASP structure file POSCAR format; Use VASP for high-precision structural optimization; on this basis, apply multiple strains to the corresponding directions of the crystal according to the deformation matrix of different crystal systems to form a large number of perturbation configurations of the stable structure of the crystal system; 13) Random structure: Use the ab initio random structure search program AIRSS to search for various possible symmetries and arrangements under different TiH component ratios, generate a large number of random structures, use VASP to perform coarse-precision optimization on these random structures, and extract structures from the optimization trajectory at regular intervals; 14) AIMD trajectory structure extraction: Constant temperature and constant volume AIMD simulations are performed for H2 at different temperatures and pressures, Ti, TiH, Ti H2 perfect crystals at different temperatures and compositions, and different types of defect structures, and a structure is extracted from the MD simulation trajectory at regular intervals.

3. The method for constructing the TiH binary system machine learning potential according to claim 1, characterized in that: In step 2, the method of using the AENET software package to train and test the ANN potential model is as follows: 21) Use the generate.x tool in the AENET software package to compile the structures and energies in the reference electronic structure dataset into a set in a specific format; 22) Debug the parameters in the model, including the cutoff radius, the shortest distance between atoms, the order of the descriptor basis function expansion, the size of the neural network, and the learning rate, and set the number of iterations and the weight of the test set; use the train.x tool to divide the data set into a training set and a test set, and train the ANN potential model on the training set; 23) Use predict.x to test the ANN potential model based on the reference electronic structure data E DFT Compared with the ANN potential prediction results E ANN The mean absolute error MAE and root mean square error RM SE between When the RMSE and MAE of the ANN in both the training and test sets converged to below 10 meV, the model was considered to have DFT accuracy.

4. The method for constructing the TiH binary system machine learning potential according to claim 1, characterized in that: In step 3, the method of comparing the calculation results of VASP and ANN potential as a performance test is as follows: 31) For structures with different distances between two atoms, VASP and ANN potentials were used to perform single-point calculations, extract energy-distance information, and obtain Dimer curves with the horizontal coordinate of VASP and ANN being the distance between the two atoms and the vertical coordinate being the total energy of the system; 32) The stable crystal configuration of Ti and common TiH compounds was optimized. The volume of the optimized structure was expanded and compressed by changing the normalization coefficient in POSCAR with a continuous fixed step length. The single-point calculations of structures with different volumes were performed using VASP and ANN potential, respectively. The volume of each atom was calculated as As the abscissa, we get the energy volume curve, i.e., the EV curve; 33) Perform constant temperature and constant volume AIMD and ANN-MD simulations on the supercell of body-centered cubic Ti, respectively. Heat the supercell from room temperature to a temperature higher than the melting point of Ti to melt the structure. Take the last frame structure as the initial structure and perform constant temperature simulation at this temperature. Calculate the distance between each particle in the system and other particles based on the position information of atoms in the trajectory file XDATCAR generated by AIMD and the trajectory file dump generated by ANN-MD. Select a suitable distance range and step size. For each distance interval, count the number of particle pairs in this interval. Divide the counting result by the total number of particle pairs and the volume of the distance interval to obtain the value of the radial distribution function g(r), and obtain an RDF curve with the horizontal coordinate being the distance and the vertical coordinate being g(r). 34) The close-packed hexagonal phase structure of Ti was exported from the Material Studio software structure library as a .cif file format, and then converted into a POSCAR format using VESTA software. The structure was then optimized with high precision using VASP software to obtain the equilibrium lattice constant; 35) According to the deformation matrix of different crystal systems, quasi-continuous strain is applied to the corresponding direction of the crystal structure after high-precision optimization. VASP and LAMMPS are used to perform single-point calculations on the structure under different strains. Nonlinear curve fitting is performed based on the relationship between strain and energy to obtain the crystal elastic constants, which are then brought into the bulk modulus calculation formula to obtain the bulk modulus. 36) The structure after high-precision optimization in step 34) is expanded by VESTA software to obtain a supercell, and a single vacancy, a single tetrahedral interstitial H atom, and a single octahedral interstitial H atom structure are established on the basis of the supercell, and then various arrangements and combinations are considered to construct double vacancies, double tetrahedral interstitial H atoms, and double octahedral interstitial H atom structures, and the distances between point defects are sorted to obtain the first neighbor, second neighbor, and third neighbor configurations of double vacancies, double tetrahedral interstitial H atoms, and double octahedral interstitial H atoms. VASP and LAMMPS are used to perform single-point calculations on the first to third neighbor supercell structures of perfect crystals, single defects, and double defects, respectively; the energies of the primitive cell Ti and H2 molecules are calculated respectively, and the energy is divided by the number of atoms to obtain the chemical potential of Ti and H2; the defect formation energy is calculated according to the formula E f =E d -E p ±nμ, where E d is the total energy of the defective system, E p is the total energy of the perfect crystal, μ is the chemical potential of the defective atom, and n is the number of defective atoms. The first to third neighbor structure formation energies of single defects and double defects are obtained, and the calculation results of VASP and LAMMPS are compared.

5. The method for constructing the TiH binary system machine learning potential according to claim 1, characterized in that: In step 4, the method for targeted improvement of the data with large errors between the ANN potential model and the DFT results in the model test and performance test results is as follows: 41) For the data with RMSE>30meV in the model test, extract and view the configuration, perform optimization or constant temperature and constant volume AI MD and ANN-MD simulations, and extract the optimization or simulation trajectory into the data set; For performance with large errors in performance testing, targeted constant temperature and constant volume AIMD and ANN-MD simulations are performed, and the simulation trajectories are extracted and put into the data set; 42) Repeat steps 2, 3 and 41) until the error drops to an acceptable range, and an ANN potential model that meets the accuracy requirements is obtained.

Images