CN114550844A - Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy - Google Patents
Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy Download PDFInfo
- Publication number
- CN114550844A CN114550844A CN202210106825.6A CN202210106825A CN114550844A CN 114550844 A CN114550844 A CN 114550844A CN 202210106825 A CN202210106825 A CN 202210106825A CN 114550844 A CN114550844 A CN 114550844A
- Authority
- CN
- China
- Prior art keywords
- machine learning
- potential energy
- potential
- energy
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000010801 machine learning Methods 0.000 title claims abstract description 191
- 238000000034 method Methods 0.000 title claims abstract description 87
- 238000005381 potential energy Methods 0.000 title claims abstract description 69
- 230000033116 oxidation-reduction process Effects 0.000 title claims abstract description 59
- 238000005182 potential energy surface Methods 0.000 claims abstract description 169
- 238000012549 training Methods 0.000 claims abstract description 83
- 238000004364 calculation method Methods 0.000 claims abstract description 65
- 239000000376 reactant Substances 0.000 claims abstract description 64
- 238000010168 coupling process Methods 0.000 claims abstract description 61
- 238000005859 coupling reaction Methods 0.000 claims abstract description 61
- 230000008878 coupling Effects 0.000 claims abstract description 59
- 238000005070 sampling Methods 0.000 claims abstract description 54
- 238000006243 chemical reaction Methods 0.000 claims abstract description 27
- 230000006870 function Effects 0.000 claims abstract description 25
- 238000000329 molecular dynamics simulation Methods 0.000 claims description 32
- 230000010354 integration Effects 0.000 claims description 18
- 238000004088 simulation Methods 0.000 claims description 17
- 230000005595 deprotonation Effects 0.000 claims description 14
- 238000010537 deprotonation reaction Methods 0.000 claims description 14
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 11
- 238000012937 correction Methods 0.000 claims description 11
- 230000008859 change Effects 0.000 claims description 6
- 239000002253 acid Substances 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 claims description 5
- 230000001133 acceleration Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 2
- 230000008569 process Effects 0.000 description 10
- 239000000463 material Substances 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 3
- 230000002787 reinforcement Effects 0.000 description 3
- 239000000243 solution Substances 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000006356 dehydrogenation reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 239000007864 aqueous solution Substances 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000006757 chemical reactions by type Methods 0.000 description 1
- 239000013626 chemical specie Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000005624 perturbation theories Effects 0.000 description 1
- 230000005610 quantum mechanics Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000027756 respiratory electron transport chain Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/30—Prediction of properties of chemical compounds, compositions or mixtures
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C10/00—Computational theoretical chemistry, i.e. ICT specially adapted for theoretical aspects of quantum chemistry, molecular mechanics, molecular dynamics or the like
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/70—Machine learning, data mining or chemometrics
Landscapes
- Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Chemical & Material Sciences (AREA)
- Crystallography & Structural Chemistry (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Organic Low-Molecular-Weight Compounds And Preparation Thereof (AREA)
Abstract
The present disclosure provides a method for accelerating acidity constant and oxidation-reduction potential calculation based on machine learning potential energy, comprising: extracting partial structures from the pre-equilibrium trajectory, and performing first principle calculation on the potential energy surface of the reactant and the potential energy surface of the product to obtain potential energy and atomic force corresponding to the partial structures and using the potential energy and the atomic force as an initial training data set; obtaining a reactant machine learning potential energy surface and a product machine learning potential energy surface after machine learning potential energy function training, and coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface to obtain a sampling potential energy surface; updating a training data set and a machine learning potential energy surface; coupling the machine learning potential energy surface of the reactant and the machine learning potential energy surface of the product, and calculating accumulated average vertical energy differences corresponding to different coupling parameters and corresponding reaction free energy; the free energy of the reaction is converted into an acidity constant and an oxidation-reduction potential. The present disclosure also provides an apparatus, an electronic device, and a readable storage medium for accelerating an acidity constant and an oxidation-reduction potential calculation based on machine learning potential energy.
Description
Technical Field
The disclosure relates to the technical field of computational chemistry and physics, in particular to a computing method, device and electronic equipment and a readable storage medium for computing acidity constants and oxidation-reduction potentials based on machine learning potential energy accelerated full-atom molecular dynamics.
Background
The acidity constant and redox potential are fundamental physicochemical properties of the chemical species of interest, and have irreplaceable roles in understanding and predicting the chemical properties of molecules and materials. For example, the description of the acidity constant and the redox potential on the proton electron transfer provides a thought for the research of the reaction mechanism thereof. By calculating these properties, it is helpful to better understand the mechanism of the reaction and even to guide the design of new materials. Jun Cheng and Michelel Sprik can accurately calculate the acidity constants and redox potentials of various molecules/ions in the aqueous solution by free energy perturbation (free energy perturbation) and thermodynamic integration (thermo-kinetic integration) methods and using the full atom first principle molecular dynamics simulation. However, since this method is based on the first principle, the calculation is expensive, and the application in a large scale is still limited at present.
For accelerating first-principles molecular dynamics calculations, a currently common approach is to fit first-principles potential energy surfaces using machine learning. The method is realized by expanding the local environment of the atoms into high-dimensional vectors and establishing a single mapping of the vectors to the atomic energy and the atomic force. The method of machine learning potential surface fitting can well repeat the trajectory of first-nature principle molecular dynamics, but the method of calculating free energy by using the method still needs to be developed and perfected. This also greatly limits the use of machine learning potentials to accelerate the calculation of acidity constants and redox potentials.
Disclosure of Invention
In order to solve at least one of the above technical problems, the present disclosure provides a method, an apparatus, an electronic device, and a readable storage medium for accelerating acidity constant and oxidation-reduction potential calculation based on machine learning potential.
According to one aspect of the present disclosure, there is provided a method of accelerating acidity constant and oxidation-reduction potential calculation based on machine learning potential energy, comprising:
obtaining a pre-equilibrium trajectory through molecular dynamics simulation or Monte Carlo simulation;
extracting partial structures from the pre-balanced track, respectively carrying out first principle calculation on a reactant potential energy surface and a product potential energy surface to obtain potential energy and atomic force corresponding to the partial structures, and taking the potential energy and the atomic force as an initial training data set for training a machine learning potential energy function;
through the initial training data set, after machine learning potential energy function training, obtaining a reactant machine learning potential energy surface and a product machine learning potential energy surface, coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface through a free energy perturbation method, and obtaining a sampling potential energy surface after coupling;
updating the training data set and the machine learning potential energy surface based on the sampling potential energy surface to obtain an updated training data set and an updated machine learning potential energy surface;
based on the updated data set and the updated machine learning potential energy surface, coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface by a free energy perturbation calculation method, calculating accumulated average vertical energy differences corresponding to different coupling parameters, and calculating corresponding reaction free energy by a thermodynamic integration method;
converting the reaction free energy into an acidity constant and an oxidation-reduction potential.
According to the calculation method of the potential energy accelerated acidity constant and the oxidation-reduction potential based on the machine learning, the training data set and the machine learning potential energy surface are updated based on the sampling potential energy surface, and the updated training data set and the updated machine learning potential energy surface are obtained, wherein the calculation method comprises the following steps:
sampling on the sampling potential energy surface to obtain sampling data;
performing molecular dynamics simulation based on the sampling data to obtain a simulation track;
selecting a partial structure on the simulation track, respectively carrying out first principle calculation on a potential energy surface of a reactant and a potential energy surface of a product to obtain potential energy and atomic force corresponding to the partial structure, and taking the potential energy and the atomic force as a training data set to serve as an updated training data set;
through the updated training data set, after machine learning potential energy function training, obtaining an updated reactant machine learning potential energy surface and an updated product machine learning potential energy surface, coupling the reactant potential energy surface and the product potential energy surface by using a free energy perturbation method, and obtaining a sampling potential energy surface after coupling;
wherein the updated machine learning potential surface comprises the updated reactant machine learning potential surface and an updated product machine learning potential surface.
According to at least one embodiment of the present disclosure, a method for calculating potential-accelerated acidity constant and oxidation-reduction potential based on machine learning is provided, which performs molecular dynamics simulation based on the sampled data to obtain a simulation trajectory, including:
and extrapolating the current sampling potential energy surface along the direction from small to large of the coupling parameter to carry out molecular dynamics simulation.
According to the method for calculating the accelerated acidity constant and the oxidation-reduction potential based on the machine learning potential energy, the training data set and the machine learning potential energy surface are updated based on the sampling potential energy surface, the updated training data set and the updated machine learning potential energy surface are obtained, and the method is repeated for multiple times.
The method for calculating the potential energy accelerated acidity constant and the oxidation-reduction potential based on machine learning according to at least one embodiment of the present disclosure comprises the free energy perturbation and the thermodynamic integration, and the calculation formula is as follows:
Eη＝(1-η)Eini+ηFfin，η∈[0，1]，
wherein E isηRepresenting sampled potential energy surfaces constructed by free energy perturbations, EiniRepresenting the machine learning potential surface of the reactant, EfinRepresenting the machine learning potential surface of the product, eta represents the coupling parameter delta A representing the free energy change required,<ΔE>ηrepresents EηSampling the vertical energy difference on the potential surface, and calculating the same structureStatistical averaging of the energy differences over the potential energy surfaces of the product and reactant.
According to the calculation method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy, the calculation method for converting reaction free energy into acidity constant and oxidation-reduction potential is as follows:
wherein, DeltadpA represents the deprotonation free energy of the corresponding conjugate acid,
represents H3O+Deprotonation free energy of (k)BTln[coΛ3]Representing a translational entropy correction term, ΔoxA represents the redox free energy of the corresponding molecule,
indicating the standard Gibbs free energy of formation, Delta, of gas protonsZPE represents the zero energy correction term for protons.
The method for calculating the acidity constant and the oxidation-reduction potential based on the machine learning potential energy acceleration according to at least one embodiment of the present disclosure, and the method for iteratively updating the free energy perturbation are applicable to the type of the machine learning potential energy surface, and comprise the following steps: deep Potential, BPNN, or GAP.
According to still another aspect of the present disclosure, there is provided an apparatus for calculating an accelerated acidity constant and an oxidation-reduction potential based on machine learning potential, comprising:
the pre-equilibrium track acquisition module is used for acquiring a pre-equilibrium track through molecular dynamics simulation or Monte Carlo simulation;
the initial training data set acquisition module extracts a partial structure from the pre-balanced track, respectively performs first principle calculation on a reactant potential energy surface and a product potential energy surface to obtain potential energy and atomic force corresponding to the partial structure, and takes the potential energy and the atomic force as an initial training data set for training a machine learning potential energy function;
the sampling potential surface coupling module is used for obtaining a reactant machine learning potential surface and a product machine learning potential surface after the initial training data set is trained through a machine learning potential function, coupling the reactant machine learning potential surface and the product machine learning potential surface through a free energy perturbation method, and obtaining a sampling potential surface after coupling;
the training data set and machine learning potential energy surface updating module is used for updating the training data set and the machine learning potential energy surface based on the sampling potential energy surface to obtain an updated training data set and an updated machine learning potential energy surface;
the reaction free energy acquisition module is used for coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface by a free energy perturbation calculation method based on the updated data set and the updated machine learning potential energy surface, calculating accumulated average vertical energy differences corresponding to different coupling parameters, and calculating corresponding reaction free energy by a thermodynamic integration method;
and the acidity constant and oxidation-reduction potential acquisition module is used for converting the free energy of the reaction into an acidity constant and an oxidation-reduction potential.
According to at least one embodiment of the present disclosure, the apparatus for calculating accelerated acidity constant and oxidation-reduction potential based on machine learning potential energy, based on the sampled potential energy surface, updating the training data set and the machine learning potential energy surface, and obtaining an updated training data set and an updated machine learning potential energy surface, includes:
sampling on the sampling potential energy surface to obtain sampling data;
performing molecular dynamics simulation based on the sampling data to obtain a simulation track;
selecting a partial structure on the simulation track, respectively carrying out first principle calculation on a potential energy surface of a reactant and a potential energy surface of a product to obtain potential energy and atomic force corresponding to the partial structure, and taking the potential energy and the atomic force as a training data set to serve as an updated training data set;
through the updated training data set, after machine learning potential energy function training, obtaining an updated reactant machine learning potential energy surface and an updated product machine learning potential energy surface, coupling the reactant potential energy surface and the product potential energy surface by using a free energy perturbation method, and obtaining a sampling potential energy surface after coupling;
wherein the updated machine learning potential surface comprises the updated reactant machine learning potential surface and an updated product machine learning potential surface.
According to at least one embodiment of the present disclosure, a device for calculating potential-accelerated acidity constant and oxidation-reduction potential based on machine learning is configured to perform molecular dynamics simulation based on the sampled data to obtain a simulation trajectory, including:
and extrapolating the current sampling potential energy surface along the direction from small to large of the coupling parameter to carry out molecular dynamics simulation.
According to the device for calculating the acidity constant and the oxidation-reduction potential based on the machine learning potential energy, the training data set and the machine learning potential energy surface are updated based on the sampling potential energy surface, the updated training data set and the updated machine learning potential energy surface are obtained, and the process is repeated for multiple times.
The device for calculating the potential energy accelerated acidity constant and the oxidation-reduction potential based on the machine learning according to at least one embodiment of the present disclosure, the method for calculating the coupling of the free energy perturbation comprises the free energy perturbation and the thermodynamic integral, and the calculation formula is as follows:
En＝(1-η)Eini+ηEfin，η∈[0，1]，
wherein E isηRepresenting sampled potential energy surfaces constructed by free energy perturbations, EiniRepresenting the machine learning potential surface of the reactant, EfinRepresenting the machine learning potential surface of the product, eta represents the coupling parameter delta A representing the free energy change required,<ΔE>ηrepresents EηThe vertical energy difference on the potential energy surface is sampled and obtained by calculating the statistical average of the energy difference of the same structure on the potential energy surfaces of the product and the reactant.
According to the device for calculating accelerated acidity constant and oxidation-reduction potential based on machine learning potential energy of at least one embodiment of the present disclosure, the calculation method for converting reaction free energy into acidity constant and oxidation-reduction potential is as follows:
wherein, DeltadpA represents the deprotonation free energy of the corresponding conjugate acid,
represents H3O+Deprotonation free energy of (k)BTln[coΛ3]Representing a translational entropy correction term, ΔoxA represents the redox free energy of the corresponding molecule,
representing the standard Gibbs free energy of formation, Delta, of gas protonsZPE represents the zero energy correction term for protons.
The device for calculating the acidity constant and the oxidation-reduction potential based on the machine learning potential energy acceleration according to at least one embodiment of the present disclosure, and the method for iteratively updating the free energy perturbation, are applicable to the type of the machine learning potential energy surface, and comprise: deep Potential, BPNN, or GAP.
According to yet another aspect of the present disclosure, there is provided an electronic device including:
a memory storing execution instructions;
a processor executing execution instructions stored by the memory to cause the processor to perform any of the methods described above.
According to yet another aspect of the present disclosure, there is provided a readable storage medium having stored therein execution instructions for implementing any of the above methods when executed by a processor.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the disclosure and together with the description serve to explain the principles of the disclosure.
Fig. 1 is a flow chart diagram of a method for accelerated acidity constant and oxidation-reduction potential calculation based on machine learning potential energy according to one embodiment of the present disclosure.
FIG. 2 is a schematic flow diagram of a method of obtaining an updated training data set and an updated machine learning potential surface by updating the training data set and the machine learning potential surface according to one embodiment of the present disclosure.
FIG. 3 is a schematic flow diagram of a method of coupling reactant and product potential energy surfaces, updating a data set, accelerating acidity constants, and redox potential calculations, according to one embodiment of the present disclosure.
FIG. 4 is a graph represented by H according to one embodiment of the present disclosure3O+，H2S，HCl，H2O and HS-The deprotonation free energy and acidity constant calculation results are shown schematically.
FIG. 5 is a solution of Cl, OH, HS, O according to one embodiment of the present disclosure2And CO2And the calculation result of the free energy of reduction and the redox potential of the corresponding redox couple is shown schematically.
Fig. 6 is a schematic block diagram of an apparatus for machine learning potential-based accelerated acidity constant and oxidation-reduction potential calculation using a hardware implementation of a processing system according to an embodiment of the present disclosure.
Description of the reference numerals
1000 potential energy accelerated acidity constant and oxidation-reduction potential calculating device based on machine learning
1002 pre-equilibrium trajectory acquisition module
1004 initial training data set acquisition module
1006 sampling potential energy surface coupling module
1008 training data set and machine learning potential energy surface updating module
1010 reaction free energy acquisition module
1012 acidity constant and oxidation-reduction potential acquisition module
1100 bus
1200 processor
1300 memory
1400 and other circuits.
Detailed Description
The present disclosure will be described in further detail with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are for purposes of illustration only and are not to be construed as limitations of the present disclosure. It should be further noted that, for the convenience of description, only the portions relevant to the present disclosure are shown in the drawings.
It should be noted that the embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict. Technical solutions of the present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Unless otherwise indicated, the illustrated exemplary embodiments/examples are to be understood as providing exemplary features of various details of some ways in which the technical concepts of the present disclosure may be practiced. Accordingly, unless otherwise indicated, features of the various embodiments may be additionally combined, separated, interchanged, and/or rearranged without departing from the technical concept of the present disclosure.
The use of cross-hatching and/or shading in the drawings is generally used to clarify the boundaries between adjacent components. As such, unless otherwise noted, the presence or absence of cross-hatching or shading does not convey or indicate any preference or requirement for a particular material, material property, size, proportion, commonality between the illustrated components and/or any other characteristic, attribute, property, etc., of a component. Further, in the drawings, the size and relative sizes of components may be exaggerated for clarity and/or descriptive purposes. While example embodiments may be practiced differently, the specific process sequence may be performed in a different order than that described. For example, two processes described consecutively may be performed substantially simultaneously or in reverse order to that described. In addition, like reference numerals denote like parts.
When an element is referred to as being "on" or "on," "connected to" or "coupled to" another element, it can be directly on, connected or coupled to the other element or intervening elements may be present. However, when an element is referred to as being "directly on," "directly connected to" or "directly coupled to" another element, there are no intervening elements present. For purposes of this disclosure, the term "connected" may refer to physically, electrically, etc., and may or may not have intermediate components.
The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, when the terms "comprises" and/or "comprising" and variations thereof are used in this specification, the presence of stated features, integers, steps, operations, elements, components and/or groups thereof are stated but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof. It is also noted that, as used herein, the terms "substantially," "about," and other similar terms are used as approximate terms and not as degree terms, and as such, are used to interpret inherent deviations in measured values, calculated values, and/or provided values that would be recognized by one of ordinary skill in the art.
Fig. 1 is a method S100 for calculating accelerated acidity constant and oxidation-reduction potential based on machine learning potential energy according to an embodiment of the present disclosure, including the following steps.
In step S102, a pre-equilibrium trajectory is obtained by molecular dynamics simulation (MD) or monte carlo simulation (MC).
In step S104, a partial structure is extracted from the pre-balanced trajectory obtained in step S102, a first principle calculation is performed on the reactant potential energy surface and the product potential energy surface, respectively, to obtain potential energy and atomic force corresponding to the partial structure, and the potential energy and the atomic force are used as an initial training data set for training a machine learning potential energy function. Through steps S102 and S104, an initial structure is obtained through molecular dynamics or monte carlo simulation, and the way of calculating the interaction force between atoms includes a classical force field and a first principle calculation method.
In step S106, a reactant machine learning potential surface E is obtained after training by the machine learning potential function through the initial training data set obtained in step S104iniAnd product machine learning potential surface EfinMachine learning potential surface E by coupling reactants by free energy perturbationiniAnd product machine learning potential plane EfinObtaining a sampling potential energy surface E after couplingη。
In step S108, the training data set and the machine learning potential energy surface are updated based on the sampled potential energy surface, and an updated training data set and an updated machine learning potential energy surface are obtained. Fig. 2 shows a specific embodiment of step S108, and as shown in fig. 2, S108 includes the following steps.
In step S1082, sampling is performed on the sampling potential energy surface, and sampling data is obtained.
In step S1084, a molecular dynamics simulation is performed based on the sampled data, and a simulation trajectory is obtained. In a specific implementation, in step S1082, the current sampling potential surface is extrapolated from a direction along the coupling parameter η from a small direction to a large direction, and then, in step S1084, a molecular dynamics simulation is performed.
In step S1086, a partial structure is selected from the simulation trajectory, the potential energy surface of the reactant and the potential energy surface of the product are calculated according to the first principle, so as to obtain potential energy and atomic force corresponding to the partial structure, and the potential energy and the atomic force are used as a training data set to serve as an updated training data set.
In step S1088, after the machine learning potential energy function training is performed through the updated training data set, an updated reactant machine learning potential energy surface and an updated product machine learning potential energy surface are obtained, the reactant machine learning potential energy surface and the product machine learning potential energy surface are coupled by using a free energy perturbation method, and the sampling potential energy surface is obtained after the coupling. The free energy perturbation coupling calculation method comprises free energy perturbation and thermodynamic integration, and the calculation formula is as follows:
Eη＝(1-η)Eini+ηEfin，η∈[0，1]，
wherein E isηRepresenting sampled potential energy surfaces constructed by free energy perturbations, EiniRepresenting the machine learning potential surface of the reactant, EfinRepresenting the machine learning potential energy surface of the product, eta representing a coupling parameter, the value of the coupling parameter eta being a plurality of values between 0 and 1, Delta A representing the free energy variation to be obtained,<ΔE>ηrepresents EηThe vertical energy difference on the potential energy surface is sampled and obtained by calculating the statistical average of the energy difference of the same structure on the potential energy surfaces of the product and the reactant. A method for iteratively updating free energy perturbation is applicable to the types of machine learning potential energy surfaces, and comprises the following steps: deep Potential, BPNN or GAP, the method of iteratively updating the data set includes randomly choosing a simulated structure, or obtaining through concurrent learning and reinforcement learning. The updated machine learning potential surfaces include an updated reactant machine learning potential surface and an updated product machine learning potential surface.
It should be noted that step S108 may be repeated multiple times by this embodiment to obtain an updated training data set and an updated machine learning potential surface.
In step S110, based on the updated data set and the updated machine learning potential surface, the reactant machine learning potential surface and the product machine learning potential surface are coupled by the free energy perturbation calculation method, and the cumulative average vertical energy difference corresponding to different coupling parameters η is calculated<ΔE>ηAnd calculating the corresponding free energy of reaction by a thermodynamic integration method.
In step S112, the reaction free energy is converted into an acidity constant and an oxidation-reduction potential. The calculation method for converting the reaction free energy into the acidity constant and the oxidation-reduction potential is as follows:
wherein, DeltadpA represents the deprotonation free energy of the corresponding conjugate acid,
represents H3O+Deprotonation free energy of (k)BTln[coΛ3]Representing a translational entropy correction term, ΔoxA represents the redox free energy of the corresponding molecule,
representing the standard Gibbs free energy of formation, Delta, of gas protonsZPE represents the zero energy correction term for protons.
FIG. 3 is a schematic flow diagram of a method of coupling reactant and product potential energy surfaces, updating a data set, accelerating acidity constants, and redox potential calculations, according to one embodiment of the present disclosure. Firstly, a pre-balanced track is obtained through the calculation of a classical force field or a first principle, and a partial structure in a marked track is calculated through the first principle to be used as an initial data set. On the basis of the above, machine learning potential surfaces of reactants and products are constructed. And constructing a molecular dynamics sampling potential energy surface by a free energy perturbation method, performing molecular dynamics or Monte Carlo simulation, and updating a data set. After a good data set is obtained, a potential function is constructed which is finally used for calculating the free energy. And calculating the free energy change by using the final potential function, and converting the free energy change into an acidity constant or an oxidation-reduction potential by using a calculation formula.
FIG. 4 is a graph represented by H according to one embodiment of the present disclosure3O+，H2S，HCl，H2O and HS-The deprotonation free energy and acidity constant calculation results are shown schematically. In this embodiment, the potential energy surface is described using a BLYP functional and grime D3 correction. The root mean square error of the vertical energy difference is in the order of 1.0E-2 eV. The error of atomic force under different coupling parameters does not change much with the simulation time. By increasing the time scale and the integration point, a more accurate thermodynamic integration value can be obtained, and the statistical error of free energy calculation is eliminated. As shown in fig. 4. (a) H2And (4) errors of vertical energy difference under different coupling parameters in the S dehydrogenation process. (b) The root mean square error of the atomic force under different coupling parameters in the HCl dehydrogenation process evolves with the simulation time. (c) H obtained by machine learning potential energy function and first-nature principle molecular dynamics simulation3O+Comparison of thermodynamic integrals of deprotonation processes. (d) And comparing the acidity constant obtained by the machine learning potential energy function and the first-nature principle molecular dynamics simulation with an experimental value.
FIG. 5 is a solution of Cl, OH, HS, O according to one embodiment of the present disclosure2And CO2And the calculation result of the free energy of reduction and the redox potential of the corresponding redox couple is shown schematically. In this embodiment, the atomic force error for a 1ns trajectory is less than
The error of numerical integration can be reduced by increasing the integration points of Cl and OH. The time scale is increased, and the statistical sampling precision can be improved. The residual error in calculating the redox potential versus experimental values is mainly derived from the error of the density functional approximation. As shown in FIG. 5, (a) thermodynamic product of redox processAnd (5) dividing the curve. (b) And (5) calculating the oxidation-reduction potential. (c) - (f) atomic force error analysis of 1ns trajectory under different coupling parameters in the OH reduction process.
Fig. 6 is a device 1000 for calculating potential-accelerated acidity constant and oxidation-reduction potential based on machine learning according to an embodiment of the present disclosure, including:
a pre-equilibrium trajectory acquisition module 1002 in which a pre-equilibrium trajectory is obtained by molecular dynamics simulation (MD) or monte carlo simulation (MC).
An initial training data set obtaining module 1004, in which partial structures are extracted from the pre-balanced trajectory, potential energy surfaces of the reactant and the product are respectively subjected to first principle calculation to obtain potential energy and atomic force corresponding to the partial structures, and the potential energy and the atomic force are used as an initial training data set for training a machine learning potential energy function. The initial structure is obtained by molecular dynamics or monte carlo simulation through the pre-equilibrium trajectory acquisition module 1002 and the initial training data set acquisition module 1004, and the way of calculating the interatomic interaction force includes a classical force field and a first principle calculation method.
A sampling potential surface coupling module 1006, in which the reactant machine learning potential surface E is obtained after the initial training data set obtained by the initial training data set obtaining module 1004 is trained by the machine learning potential energy functioniniAnd product machine learning potential plane EfinMachine learning potential surface E of reactant coupled by free energy perturbationiniAnd product machine learning potential plane EfinObtaining a sampling potential energy surface E after couplingη。
The training data set and machine learning potential energy surface updating module 1008 updates the training data set and the machine learning potential energy surface based on the sampling potential energy surface to obtain an updated training data set and an updated machine learning potential energy surface.
The training data set and machine learning potential energy surface updating module 1008 includes:
and sampling on the sampling potential energy surface to obtain sampling data.
And performing molecular dynamics simulation based on the sampling data to obtain a simulation track. In a specific implementation, a molecular dynamics simulation is performed after a current sampling potential surface is extrapolated from a small to large direction along the coupling parameter η.
And selecting a partial structure on the simulation track, respectively carrying out first principle calculation on the potential energy surface of the reactant and the potential energy surface of the product to obtain potential energy and atomic force corresponding to the partial structure, and taking the potential energy and the atomic force as a training data set to serve as an updated training data set.
And through the updated training data set, after training through a machine learning potential energy function, obtaining an updated reactant machine learning potential energy surface and an updated product machine learning potential energy surface, coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface by using a free energy perturbation method, and obtaining a sampling potential energy surface after coupling. The free energy perturbation coupling calculation method comprises free energy perturbation and thermodynamic integration, and the calculation formula is as follows:
Eη＝(1-η)Eini+ηEfin，η∈[0，1]，
wherein E isηRepresenting sampled potential energy surfaces constructed by free energy perturbations, EiniRepresenting the machine learning potential surface of the reactant, EfinRepresenting the machine learning potential energy surface of the product, eta representing a coupling parameter, the value of the coupling parameter eta being a plurality of values between 0 and 1, Delta A representing the free energy variation to be obtained,<ΔE>ηrepresents EηThe vertical energy difference on the potential energy surface is sampled and obtained by calculating the statistical average of the energy difference of the same structure on the potential energy surfaces of the product and the reactant. A method for iteratively updating free energy perturbation is applicable to the types of machine learning potential energy surfaces, and comprises the following steps: deep patent, BPNN or GAP, and the method for iteratively updating the data set comprises randomly selecting a simulated structure or obtaining the simulated structure through a concurrent learning and reinforcement learning method. The updated machine learning potential surface includes an updated reactant machine learning potential surface and an updated product machine learningStudy potential energy surface. This embodiment may be repeated multiple times to obtain an updated training data set and an updated machine learning potential surface.
A free energy of reaction obtaining module 1010, wherein for the updated machine learning potential surface based on the updated data set and the updated machine learning potential surface obtained in module 1008, the free energy perturbation calculation method is used to couple the machine learning potential surface of the reactant and the machine learning potential surface of the product, and the cumulative average vertical energy difference corresponding to different coupling parameters η is calculated<ΔE>ηAnd calculating the corresponding free energy of reaction by a thermodynamic integration method.
The acidity constant and oxidation-reduction potential obtaining module 1012 converts the reaction free energy into an acidity constant and an oxidation-reduction potential. The calculation method for converting the reaction free energy into the acidity constant and the oxidation-reduction potential is as follows:
wherein, DeltadpA represents the deprotonation free energy of the corresponding conjugate acid,
represents H3O+Deprotonation free energy of (k)BTln[coΛ3]Representing a translational entropy correction term, ΔoxA represents the redox free energy of the corresponding molecule,
representing the standard Gibbs free energy of formation, Delta, of gas protonsZPE represents the zero energy correction term for protons.
According to yet another aspect of the present disclosure, there is provided an electronic device including:
a memory storing execution instructions;
a processor executing execution instructions stored by the memory to cause the processor to perform the method of any of the above.
According to yet another aspect of the present disclosure, there is provided a readable storage medium having stored therein execution instructions for implementing the method of any one of the above when executed by a processor.
The method utilizes a free energy sampling method and directionally constructs a machine learning potential energy surface, accelerates the calculation of free energy, and obtains an acidity constant and an oxidation-reduction potential. Starting from a pre-equilibrium track or a reasonable initial structure set, a machine learning potential energy surface is constructed by a free energy sampling method, so that the acidity constant and the oxidation-reduction potential are accelerated by using machine learning potential energy. The potential energy surface information of the simulation system comes from quantum mechanics calculation, and the energy and atomic stress of the potential energy surface can be repeated rapidly and accurately by fitting a machine learning potential function. The calculation of the free energy, and thus the acidity constant and the oxidation-reduction potential, is accelerated on the basis of the machine-learned potential energy surface. The method can quickly and accurately calculate the acidity constant and the oxidation-reduction potential, and is suitable for various complex systems. Can be used for calculating the acidity constant and the oxidation-reduction potential of a complex system with high flux, and accelerates the development of materials.
The method is based on the idea of constructing a coupling state potential energy surface by a free energy perturbation theory, the machine learning potential energy surfaces of reactants and products are trained respectively, and the potential energy surface of the reactants is gradually changed to the potential energy surface of the products by constructing sampling potential energy surfaces of different coupling parameters. Meanwhile, iterative learning, concurrent learning or reinforcement learning and other methods are used for iteratively updating the machine learning data set so as to achieve the purpose of accurately predicting potential energy surfaces of reactants and products. And finally, calculating the reaction free energy of the computer system by a free energy perturbation and thermodynamic integration method on the basis of obtaining an accurate machine learning potential energy surface. Where the types of reactions that can be calculated include deprotonation processes and redox processes. And then converting the obtained free energy into corresponding acidity constant and oxidation-reduction potential according to a calculation formula. The present disclosure has the following technical advantages:
1. compared with the traditional method for constructing the machine learning potential surface, the method for constructing the machine learning potential surface uses the free energy perturbation method to assist the construction of the potential surface, so that the machine learning potential surface can be suitable for free energy calculation of different reaction types.
2. Compared with the molecular dynamics based on the first principle, the time scale can be increased to more than 1000 times, and the point taking density of thermodynamic integration is higher, so that the accuracy of numerical calculation is ensured. The time scale is promoted, and meanwhile, the accuracy of the potential energy surface is guaranteed.
3. The scale of the simulation system can be improved by more than 1000 times. The calculation of the acidity constant and the oxidation-reduction potential is accelerated, and the calculation accuracy is advanced to the limit of the functional used.
Fig. 6 is a schematic block diagram of an apparatus for machine learning potential-based accelerated acidity constant and oxidation-reduction potential calculation using a hardware implementation of a processing system according to an embodiment of the present disclosure. The apparatus may include corresponding means for performing each or several of the steps of the flowcharts described above. Thus, each step or several steps in the above-described flow charts may be performed by a respective module, and the apparatus may comprise one or more of these modules. The modules may be one or more hardware modules specifically configured to perform the respective steps, or implemented by a processor configured to perform the respective steps, or stored within a computer-readable medium for implementation by a processor, or by some combination.
The hardware architecture may be implemented with a bus architecture. The bus architecture may include any number of interconnecting buses and bridges depending on the specific application of the hardware and the overall design constraints. The bus 1100 couples various circuits including the one or more processors 1200, the memory 1300, and/or the hardware modules together. The bus 1100 may also connect various other circuits 1400, such as peripherals, voltage regulators, power management circuits, external antennas, and the like.
The bus 1100 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one connection line is shown, but no single bus or type of bus is shown.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and the scope of the preferred embodiments of the present disclosure includes other implementations in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the implementations of the present disclosure. The processor performs the various methods and processes described above. For example, method embodiments in the present disclosure may be implemented as a software program tangibly embodied in a machine-readable medium, such as a memory. In some embodiments, some or all of the software program may be loaded and/or installed via memory and/or a communication interface. When the software program is loaded into memory and executed by a processor, one or more steps of the method described above may be performed. Alternatively, in other embodiments, the processor may be configured to perform one of the methods described above in any other suitable manner (e.g., by means of firmware).
The logic and/or steps represented in the flowcharts or otherwise described herein may be embodied in any readable storage medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions.
For the purposes of this description, a "readable storage medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the readable storage medium include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable read-only memory (CDROM). In addition, the readable storage medium may even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in the memory.
It should be understood that portions of the present disclosure may be implemented in hardware, software, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps of the method implementing the above embodiments may be implemented by hardware that is instructed to implement by a program, which may be stored in a readable storage medium, and when executed, includes one or a combination of the steps of the method embodiments.
In addition, each functional unit in the embodiments of the present disclosure may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a separate product, may also be stored in a readable storage medium. The storage medium may be a read-only memory, a magnetic or optical disk, or the like.
In the description herein, reference to the description of the terms "one embodiment/implementation," "some embodiments/implementations," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment/implementation or example is included in at least one embodiment/implementation or example of the present application. In this specification, the schematic representations of the terms described above are not necessarily the same embodiment/mode or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments/modes or examples. Furthermore, the various embodiments/aspects or examples and features of the various embodiments/aspects or examples described in this specification can be combined and combined by one skilled in the art without conflicting therewith.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
It will be understood by those skilled in the art that the foregoing embodiments are provided merely for clarity of explanation and are not intended to limit the scope of the disclosure. Other variations or modifications may occur to those skilled in the art, based on the foregoing disclosure, and are still within the scope of the present disclosure.
Claims (10)
1. A potential energy accelerated acidity constant and oxidation-reduction potential calculation method based on machine learning is characterized by comprising the following steps:
obtaining a pre-equilibrium trajectory through molecular dynamics simulation or Monte Carlo simulation;
extracting partial structures from the pre-balanced track, respectively carrying out first principle calculation on a reactant potential energy surface and a product potential energy surface to obtain potential energy and atomic force corresponding to the partial structures, and taking the potential energy and the atomic force as an initial training data set for training a machine learning potential energy function;
through the initial training data set, after machine learning potential energy function training, obtaining a reactant machine learning potential energy surface and a product machine learning potential energy surface, coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface through a free energy perturbation method, and obtaining a sampling potential energy surface after coupling;
updating the training data set and the machine learning potential energy surface based on the sampling potential energy surface to obtain an updated training data set and an updated machine learning potential energy surface;
based on the updated data set and the updated machine learning potential energy surface, coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface by a free energy perturbation calculation method, calculating accumulated average vertical energy differences corresponding to different coupling parameters, and calculating corresponding reaction free energy by a thermodynamic integration method; and
converting the reaction free energy into an acidity constant and an oxidation-reduction potential.
2. The method of calculating accelerated acidity constants and oxidation-reduction potentials based on machine-learned potential energies of claim 1, wherein updating the training data set and the machine-learned potential energy surface based on the sampled potential energy surface to obtain an updated training data set and an updated machine-learned potential energy surface comprises:
sampling on the sampling potential energy surface to obtain sampling data;
performing molecular dynamics simulation based on the sampling data to obtain a simulation track;
selecting a partial structure on the simulation track, respectively carrying out first principle calculation on a potential energy surface of a reactant and a potential energy surface of a product to obtain potential energy and atomic force corresponding to the partial structure, and taking the potential energy and the atomic force as a training data set to serve as an updated training data set; and
through the updated training data set, after machine learning potential energy function training, obtaining an updated reactant machine learning potential energy surface and an updated product machine learning potential energy surface, coupling the reactant potential energy surface and the product potential energy surface by using a free energy perturbation method, and obtaining a sampling potential energy surface after coupling;
wherein the updated machine learning potential surface comprises the updated reactant machine learning potential surface and an updated product machine learning potential surface.
3. The method for calculating potential energy accelerated acidity constant and oxidation-reduction potential based on machine learning according to claim 2, wherein molecular dynamics simulation is performed based on the sampled data to obtain a simulation trajectory, comprising:
and extrapolating the current sampling potential energy surface along the direction from small to large of the coupling parameter to carry out molecular dynamics simulation.
4. The method of calculating accelerated acidity constant and oxidation-reduction potential based on machine learning potential energy of claim 1, wherein the training data set and the machine learning potential energy surface are updated based on the sampled potential energy surface, and the updated training data set and the updated machine learning potential energy surface are obtained and repeated for a plurality of times.
5. The method for calculating potential energy accelerated acidity constant and oxidation-reduction potential based on machine learning according to claim 1, wherein the free energy perturbation coupling calculation method comprises the free energy perturbation and thermodynamic integration, and the calculation formula is as follows:
Eη＝(1-η)Eini+ηEfin，η∈[0，1]，
wherein E isηRepresenting sampled potential energy surfaces constructed by free energy perturbations, EiniRepresenting the machine learning potential surface of the reactant, EfinRepresenting the machine learning potential surface of the product, eta represents the coupling parameter delta A representing the free energy change required,<ΔE>ηrepresents EηThe vertical energy difference on the potential energy surface is sampled and obtained by calculating the statistical average of the energy difference of the same structure on the potential energy surfaces of the product and the reactant.
6. The method for accelerating acidity constant and oxidation-reduction potential calculation based on machine learning potential energy of claim 1, wherein the calculation method for converting reaction free energy into acidity constant and oxidation-reduction potential is as follows:
wherein, DeltadpA represents the deprotonation free energy of the corresponding conjugate acid,
represents H3O+Deprotonation free energy of (k)BTln[coΛ3]Representing a translational entropy correction term, ΔoxA represents the redox free energy of the corresponding molecule,
representing the standard Gibbs free energy of formation, Delta, of gas protonsZPE represents the zero energy correction term for protons.
7. The method for calculating the acidity constant and the oxidation-reduction potential based on the machine learning potential energy acceleration according to claim 1, wherein the method for iteratively updating the free energy perturbation is applied to the type of the machine learning potential surface, and comprises the following steps: deep Potential, BPNN, or GAP.
8. An apparatus for calculating an accelerated acidity constant and an oxidation-reduction potential based on machine learning potential energy, comprising:
the pre-equilibrium track acquisition module is used for acquiring a pre-equilibrium track through molecular dynamics simulation or Monte Carlo simulation;
the initial training data set acquisition module extracts a partial structure from the pre-balanced track, respectively performs first principle calculation on a reactant potential energy surface and a product potential energy surface to obtain potential energy and atomic force corresponding to the partial structure, and takes the potential energy and the atomic force as an initial training data set for training a machine learning potential energy function;
the sampling potential surface coupling module is used for obtaining a reactant machine learning potential surface and a product machine learning potential surface after the initial training data set is trained through a machine learning potential function, coupling the reactant machine learning potential surface and the product machine learning potential surface through a free energy perturbation method, and obtaining a sampling potential surface after coupling;
the training data set and machine learning potential energy surface updating module is used for updating the training data set and the machine learning potential energy surface based on the sampling potential energy surface to obtain an updated training data set and an updated machine learning potential energy surface;
the reaction free energy acquisition module is used for coupling the reactant machine learning potential energy surface and the product machine learning potential energy surface through a free energy perturbation calculation method based on the updated data set and the updated machine learning potential energy surface, calculating accumulated average vertical energy differences corresponding to different coupling parameters, and calculating corresponding reaction free energy through a thermodynamic integration method; and
and the acidity constant and oxidation-reduction potential acquisition module is used for converting the free energy of the reaction into an acidity constant and an oxidation-reduction potential.
9. An electronic device, comprising:
a memory storing execution instructions; and
a processor executing execution instructions stored by the memory to cause the processor to perform the method of any of claims 1 to 7.
10. A readable storage medium having stored therein execution instructions, which when executed by a processor, are configured to implement the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210106825.6A CN114550844B (en) | 2022-01-28 | 2022-01-28 | Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210106825.6A CN114550844B (en) | 2022-01-28 | 2022-01-28 | Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114550844A true CN114550844A (en) | 2022-05-27 |
| CN114550844B CN114550844B (en) | 2024-04-30 |
Family
ID=81673649
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210106825.6A Active CN114550844B (en) | 2022-01-28 | 2022-01-28 | Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114550844B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117524347A (en) * | 2023-11-20 | 2024-02-06 | 中南大学 | First principle prediction method for acid radical anion hydration structure accelerated by machine learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2018049608A (en) * | 2016-09-15 | 2018-03-29 | 東レ株式会社 | Calculation method of free energy |
| CN108536998A (en) * | 2018-03-05 | 2018-09-14 | 复旦大学 | A kind of structure of matter suitable for machine learning potential energy surface construction describes method |
| CN109859806A (en) * | 2019-01-17 | 2019-06-07 | 中山大学 | A kind of prediction drug-target bond strength absolute freedom energy perturbation method |
-
2022
- 2022-01-28 CN CN202210106825.6A patent/CN114550844B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2018049608A (en) * | 2016-09-15 | 2018-03-29 | 東レ株式会社 | Calculation method of free energy |
| CN108536998A (en) * | 2018-03-05 | 2018-09-14 | 复旦大学 | A kind of structure of matter suitable for machine learning potential energy surface construction describes method |
| CN109859806A (en) * | 2019-01-17 | 2019-06-07 | 中山大学 | A kind of prediction drug-target bond strength absolute freedom energy perturbation method |
Non-Patent Citations (1)
| Title |
|---|
| 陈林红, 尚仁成: "LaCl分子结构与基态X~1Σ~+势能函数", 物理学报, no. 11, 12 November 2002 (2002-11-12), pages 66 - 70 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117524347A (en) * | 2023-11-20 | 2024-02-06 | 中南大学 | First principle prediction method for acid radical anion hydration structure accelerated by machine learning |
| CN117524347B (en) * | 2023-11-20 | 2024-04-16 | 中南大学 | First principle prediction method for acid radical anion hydration structure accelerated by machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114550844B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | An empirical-data hybrid driven approach for remaining useful life prediction of lithium-ion batteries considering capacity diving | |
| van der Laan et al. | Asymptotic optimality of likelihood-based cross-validation | |
| Cheng et al. | Residual lifetime prediction for lithium-ion battery based on functional principal component analysis and Bayesian approach | |
| Han et al. | A comparative study of charging voltage curve analysis and state of health estimation of lithium-ion batteries in electric vehicle | |
| Bickel et al. | Subsampling methods for genomic inference | |
| Shi et al. | A fast state-of-health estimation method using single linear feature for lithium-ion batteries | |
| Miguel et al. | Electrochemical model and sigma point Kalman filter based online oriented battery model | |
| CN113933713B (en) | Solid-state lithium battery state estimation method and system based on electrochemical model | |
| CN114550844A (en) | Method for accelerating acidity constant and oxidation-reduction potential based on machine learning potential energy | |
| CN111965476A (en) | Low-voltage diagnosis method based on graph convolution neural network | |
| Eckert et al. | Fractional algebraic identification of the distribution of relaxation times of battery cells | |
| CN117301949B (en) | Vehicle, method for identifying abnormal battery cell of vehicle, storage medium and electronic device | |
| Chen et al. | A novel combined estimation method of online full‐parameter identification and adaptive unscented particle filter for Li‐ion batteries SOC based on fractional‐order modeling | |
| Wu et al. | Prediction of remaining useful life of the lithium-ion battery based on improved particle filtering | |
| Li et al. | State-of-health estimation method for fast-charging lithium-ion batteries based on stacking ensemble sparse Gaussian process regression | |
| CN110991741B (en) | Section constraint probability early warning method and system based on deep learning | |
| Xue et al. | Machine learning based swift online capacity prediction of lithium-ion battery through whole cycle life | |
| CN116113838A (en) | Systems and methods for optimizing current excitation to identify electrochemical parameters of a battery based on analytical sensitivity formulas | |
| Franke-Lang et al. | Analysis of electrochemical impedance spectroscopy on zinc-air batteries using the distribution of relaxation times | |
| Hofmann et al. | Physics-informed neural networks for state of health estimation in lithium-ion batteries | |
| Tang et al. | Model parameter identification for lithium-ion batteries using adaptive multi-context cooperatively co-evolutionary parallel differential evolution algorithm | |
| CN111239615A (en) | Method and device for determining parameters of battery model, storage medium and computer equipment | |
| Lopetegi et al. | A New Battery SOC/SOH/eSOH Estimation Method Using a PBM and Interconnected SPKFs: Part I. SOC and Internal Variable Estimation | |
| CN111931325A (en) | Vehicle power battery SOC estimation method fusing ambient temperature | |
| Miyazaki et al. | Bayesian statistics-based analysis of AC impedance spectra |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |