Method for predicting ion distribution in lithium-sulfur battery

Abstract

The present invention provides a method for predicting ion distribution in a lithium-sulfur battery. In one embodiment, the method for predicting ion distribution in a lithium-sulfur battery may comprise: a first step of forming ion distribution image data in the battery by inputting a boundary value of the battery and an initial value of the battery into a data formation unit that includes a battery internal ion distribution prediction model based on numerical analysis; and a second step of inputting an initial battery state and the ion distribution image data obtained in the first step into an image generation unit so as to output an ion distribution prediction image through an LSTM-based artificial intelligence block.

Description

Method for predicting ion distribution in lithium-sulfur batteries

[0001] The present invention relates to a method for predicting ion distribution within a lithium-sulfur battery, and relates to a battery system that forms ion distribution image data at micro-scale and micro-scale within a battery through a numerical analysis-based internal ion distribution prediction model, and predicts ion distribution within a battery at meso-scale using the same through an LSTM-based artificial intelligence block.

[0002]

[0003] Lithium-sulfur batteries have a high theoretical energy density of approximately 2600 Wh / kg, which is approximately 7 times that of current lithium-ion batteries (570 Wh / kg). Furthermore, sulfur, which is used as a cathode material, is abundant and inexpensive, which has the advantage of lowering the manufacturing cost of batteries, and thus has garnered much attention. Recently, with the growing interest in low-cost, high-capacity secondary batteries for improving the driving range of electric vehicles, lithium-sulfur batteries are being re-evaluated as next-generation secondary batteries. Accordingly, active research is being conducted on electrodes and electrolytes for lithium-sulfur batteries. Research on electrodes and electrolytes for lithium-sulfur batteries is based on understanding the behavior of ions inside the battery.

[0004] However, when predicting the behavior of ions within a battery using a conventional molecular dynamics model, there was a problem in that prediction was possible at the molecular level (micro-scale) but not at the actual battery size (meso-scale).

[0005] In addition, it was impossible to predict the ion behavior of the SEI (Solid Electrolyte Interphase; a thin film that forms on the surface of the negative electrode material when the battery is first charged after manufacturing) layer within the battery, and there was a problem in that it was difficult to predict the ion behavior within the battery during repeated charge and discharge cycles.

[0006]

[0007] One object of the present invention is to provide a method for predicting ion distribution in a lithium-sulfur battery, which can predict ion behavior within the battery at a meso scale.

[0008] In addition, one object of the present invention is to provide a method for predicting ion distribution in a lithium-sulfur battery, which is capable of predicting ion behavior in an SEI layer within the battery.

[0009] In addition, one object of the present invention is to provide a method for predicting ion distribution in a lithium-sulfur battery, which is capable of predicting ion behavior within the battery during repeated charge and discharge cycles.

[0010] The tasks of the present invention are not limited to the tasks mentioned above, and other tasks not mentioned will be clearly understood by those skilled in the art from the description below.

[0011]

[0012] As a technical means for achieving the above-mentioned technical task, the present invention provides a method for predicting ion distribution in a lithium sulfur battery.

[0013] Specific details of other embodiments for solving the problem are included in the description and drawings of the invention.

[0014] In one embodiment, the method may include a first step of forming ion distribution image data within a battery by inputting boundary values ​​of the battery and initial values ​​of the battery into a data forming unit including a battery internal ion distribution prediction model based on numerical analysis; and a second step of inputting the initial battery state and the ion distribution image data acquired in the first step into an image generating unit and outputting an ion distribution prediction image through an LSTM-based artificial intelligence block.

[0015] In one embodiment, the first step may include a step 1-1 of forming micro-scale intra-battery ion distribution image data by inputting the initial values ​​into a molecular dynamics simulation model; and a step 1-2 of inputting the micro-scale intra-battery ion distribution image data obtained in the step 1-1 into a lattice Boltzmann model to output meso-scale intra-battery ion distribution image data.

[0016] In one embodiment, the step 1-1 may include: a) a step of receiving the initial value and calculating electronegativity through a voltage propagation simulation model; b) a step of receiving the electronegativity calculated in the step a) and calculating partial charges through a partial charge calculation model; c) a step of receiving the partial charges calculated in the step b) and calculating interatomic interaction energy through a molecular interaction calculation model; and d) a step of calculating molecular momentum by receiving the interatomic interaction energy calculated in the step c) and forming ion distribution image data and charge distribution data in a battery during micro-scale charge and discharge through a molecular momentum calculation model.

[0017] In one embodiment, the initial values ​​may include initial ion distribution and charge / discharge voltage within the battery.

[0018] In one embodiment, the voltage propagation simulation model may be an EChemDID model that calculates the electronegativity using the following formula.

[0019]

[0020]

[0021]

[0022] ( is the electronegativity of the atom to be found, is the initial electronegativity of each atom, is the electrochemical potential added to atom i, is the electrochemical potential added to atom i over time, is the electrochemical potential added to atom j over time, k is the effective diffusivity, is the relaxation rate, is a switching function, R ij is the distance between atoms i and j, is the local weight function, W i is the total metal coordination number, is the electronegativity of the atom to be obtained over time)

[0023] In one embodiment, the partial charge calculation model may be a Charge equlibration (ACKS2) model that calculates the partial charge q using the following formula.

[0024]

[0025] (e is the molecular electric circuit, J is the blocked Coulomb interaction, q i is the charge of the i atom, q j is the charge of the j atom, X i is the electronegativity of the i atom)

[0026] In one embodiment, the intermolecular interaction model is expressed by the following equation: system It may be a reactive force field model that calculates .

[0027]

[0028]

[0029]

[0030]

[0031]

[0032]

[0033] (D ijEnergy well depth, α ij is the potential energy between atoms i and j, r ij Interatomic distance, r vdw Equilibrium bond distance, C is the Coulomb constant, q i , q j Charges of atoms i and j, respectively, r ij is the interatomic distance, is the shielding parameter, BO ij Coupling order, De, P be,1 , P over , λ over , λ angle , k a , k b is an empirical parameter, Δ i is the degree to which atoms are bonded in excess of the normal case, BO a is the combined order a, BO b is the bond order b, Φ is the interatomic bond angle, Φ0 is the equilibrium angle)

[0034] In one embodiment, the above molecular momentum calculation model can calculate the molecular momentum using the Newtonian equations of motion below.

[0035]

[0036]

[0037]

[0038] (F is for force, is the total energy change, m is mass, a is acceleration, v is velocity, x is position)

[0039] In one embodiment, the steps 1-2 may be performed by repeatedly predicting the propagation process and collision process of the potential, ions, and electrolyte within the battery by inputting ion distribution image data and charge distribution data during micro-scale charge / discharge, and using the lattice Boltzmann model, thereby predicting the ion distribution at a specific point in time.

[0040] In one embodiment, the prediction of the transfer process of the above potential can be made by calculating the charge density probability distribution using the following formula.

[0041]

[0042] (h i is the discretized potential distribution function in direction i at grid cell x, where i is the discretized direction in the grid, and c i is the unit velocity vector for the discretized direction, is the non-equilibrium charge probability density function)

[0043] In one embodiment, the propagation process of the ion can be predicted by calculating the ion density probability distribution using the following formula.

[0044]

[0045] (g i is the ion distribution function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the non-equilibrium ion probability density function)

[0046] In one embodiment, the propagation process of the above flow rate can be predicted by calculating the fluid density probability distribution of the electrolyte using the following formula.

[0047]

[0048] (f i is the fluid probability density function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the non-equilibrium velocity probability density function)

[0049] In one embodiment, the collision process prediction may be to predict the change in the distribution of charges, ions, and electrolytes after the collision within the battery using the following formula.

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056] (τ h is the potential collision relaxation time, is the equilibrium potential probability density function, is the time rate of change for the phase, S is the potential response value, is the time staff, W i is the weight coefficient, i0 is the exchange current density, R is the gas constant, T is the temperature, L h is the reaction coefficient, is the ion collision relaxation time, is the equilibrium ion probability density function, r k is an electrochemical reaction term, C k is the ion quantity at grid x, τ is the collision relaxation time, is the equilibrium fluid probability density function, G i is an external force in the discretized direction, is the density of the fluid, ω i is the weighting factor in the discretized direction, u is the velocity of the fluid, c is the unit sound velocity, and G is the dissipative resistance value.

[0057] In one embodiment, the inputs of the lattice Boltzmann model may further include the initial flow rate of the electrolyte and the pressure inside the battery.

[0058] In one embodiment, in the second step, an expected risk value of the battery and an expected life value of the battery may be additionally output.

[0059] In one embodiment, the boundary values ​​may include the type of the negative electrode of the battery, the type of the positive electrode of the battery, the initial ion distribution of the battery, and the terminal voltage of the battery.

[0060] In one embodiment, the initial battery state may include the voltage of the battery.

[0061]

[0062] According to the solution to the problem of the present invention described above, according to one embodiment of the present invention, it is possible to predict the behavior of ions within a battery at the meso scale.

[0063] Additionally, according to one embodiment of the present invention, the ion behavior of the SEI layer within the battery can be predicted.

[0064] Additionally, according to one embodiment of the present invention, it is possible to predict the behavior of ions within a battery during repeated charge and discharge cycles.

[0065] The effects of the present invention are not limited to the effects described above, and effects not mentioned can be clearly understood by a person skilled in the art to which the present invention pertains from this specification and the attached drawings.

[0066]

[0067] Figure 1 illustrates a method for predicting ion distribution in a lithium sulfur battery according to one embodiment of the present invention.

[0068] Figure 2 illustrates a first step performed according to one embodiment of the present invention.

[0069] Figure 3 illustrates the execution of step 1-1 according to one embodiment of the present invention.

[0070] Figure 4 illustrates the execution of steps 1-2 according to one embodiment of the present invention.

[0071] Figure 5 illustrates a second step performed according to one embodiment of the present invention.

[0072]

[0073] A most preferred embodiment according to the present invention comprises: a first step of forming ion distribution image data within a battery by inputting a boundary value of a battery and an initial value of the battery into a data forming unit including a battery internal ion distribution prediction model based on numerical analysis; and a second step of inputting an initial battery state and the ion distribution image data acquired in the first step into an image generating unit and outputting an ion distribution prediction image through an LSTM-based artificial intelligence block.

[0074]

[0075] Below, with reference to the attached drawings, embodiments of the present invention are described in detail to facilitate easy implementation by those skilled in the art. However, the present invention can be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, irrelevant parts have been omitted for clarity, and similar reference numerals have been used throughout the specification to indicate similar elements.

[0076] Throughout this specification, when a part is said to be "connected" to another part, this includes not only cases where it is "directly connected" but also cases where it is "electrically connected" with another element in between.

[0077] Throughout this specification, when it is said that an element is “on” another element, this includes not only cases where the element is in contact with the other element, but also cases where another element exists between the two elements.

[0078] Throughout this specification, when a part is said to "include" a certain component, this does not mean that it excludes other components, but rather that it may include other components, unless otherwise specifically stated. The terms "about," "substantially," and the like, as used throughout this specification, are used to mean at or near the numerical value when manufacturing and material tolerances inherent to the meanings stated, and are used to prevent unscrupulous infringers from unfairly exploiting disclosures that state precise or absolute values ​​to aid understanding of this specification. The terms "step of doing" or "step of" as used throughout this specification do not mean "step for."

[0079] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings and the following description. However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Like reference numbers designate like elements throughout the specification.

[0080] Hereinafter, a method for predicting ion distribution in a lithium sulfur battery according to one embodiment of the present invention will be described.

[0081] Figure 1 illustrates a method for predicting ion distribution in a lithium sulfur battery according to one embodiment of the present invention. Referring to Figure 1, the method for predicting ion distribution in a lithium sulfur battery according to the present invention includes a first step and a second step.

[0082] In one example, in the first step, image data of the ion distribution within the battery is formed through a data formation unit. In one example, the data formation unit includes a numerical analysis-based prediction model for the ion distribution within the battery. The prediction model for the ion distribution within the battery includes a molecular dynamics simulation model and a lattice Boltzmann model. The boundary values ​​and initial values ​​of the battery are input to the data formation unit.

[0083] In one example, the initial values ​​may include the initial ion distribution within the battery and the charge / discharge voltage. In one example, the boundary values ​​may include the type of the negative electrode of the battery, the type of the positive electrode of the battery, the initial ion distribution of the battery, and the voltage across the battery.

[0084] In one example, Step 1 includes Steps 1-1 and 1-2. In Step 1-1, microscale ion distribution image data is generated using a molecular dynamics simulation model. Then, in Step 1-2, based on this, a lattice Boltzmann model is used to generate mesoscale ion distribution image data using the microscale ion distribution image data obtained in Step 1-1.

[0085] In one example, in the second step, the initial battery state and the ion distribution image data acquired in the first step are input to the image generation unit to output an ion distribution prediction image. In one example, the image generation unit learns the ion distribution image data acquired in the first step through an LSTM-based artificial intelligence block to generate meso-scale ion distribution image data, which is the actual size of the battery.

[0086] In one example, the initial battery state may be the battery voltage. In another example, the battery voltage input to the LSTM artificial intelligence block in the second stage may be input in text format.

[0087] In one example, in the second stage, the initial battery state and the ion distribution image data acquired in the first stage are input into the image generation unit, so that in addition to the ion distribution prediction image, the expected risk value of the battery and the expected life value of the battery can be additionally output.

[0088]

[0089] FIG. 2 illustrates a state in which step 1 is performed according to one embodiment of the present invention, FIG. 3 illustrates a state in which step 1-1 is performed according to one embodiment of the present invention, and FIG. 4 illustrates a state in which step 1-2 is performed according to one embodiment of the present invention.

[0090] Referring to Fig. 2, the first stage includes steps 1-1 and 1-2.

[0091] In one example, step 1-1 sequentially performs steps a), b), c), and d) using a molecular dynamics simulation model. Step a) is performed using a voltage propagation simulation model, step b) is performed using a charge calculation model, step c) is performed using an intermolecular interaction calculation model, and step d) is performed using a molecular momentum calculation model.

[0092]

[0093] Steps a) to d) are described in detail with reference to Fig. 3.

[0094] In step a), initial values ​​and boundary values ​​are input, and electronegativity is calculated through a voltage propagation simulation model. As described above, the initial values ​​may include the initial ion distribution and charge / discharge voltage within the battery, and the boundary values ​​may include the type of the negative electrode of the battery, the type of the positive electrode of the battery, the initial ion distribution of the battery, and the voltage at both ends of the battery.

[0095] In one example, the voltage propagation simulation model is the EChemDID model, which calculates electronegativity using the following formula.

[0096] Formula 1)

[0097] Formula 2)

[0098] Formula 3)

[0099] Here, is the electronegativity of the atom to be found, is the initial electronegativity of each atom, is the electrochemical potential added to atom i, is the electrochemical potential added to atom i over time, is the electrochemical potential added to atom j over time, k is the effective diffusivity, is the relaxation rate, is a switching function, R ij is the distance between atoms i and j, is the local weight function, W i is the total metal coordination number, is the electronegativity of the atom to be obtained over time.

[0100] The initial electronegativity of an atom is given by Equation 1 External voltage to The electronegativity of the atom to be calculated by adding / 2 . The voltage propagation at the electrode is simulated using Equation 2. Then, the change in electronegativity of the atom due to the propagated voltage is calculated using Equation 3.

[0101] In step b), the electronegativity calculated in step a) is input and the partial charge is calculated using the partial charge calculation model.

[0102] In one example, the partial charge calculation model is the Charge equlibration (ACKS2) model, which calculates the partial charge q using the following formula.

[0103] Formula 4)

[0104] Here, e is the molecular electric circuit, J is the blocked Coulomb interaction, and q i is the charge of the i atom, q j is the charge of the j atom, X i is the electronegativity of the i atom.

[0105] Electronegativity X changed through formula 4 i Calculate the partial charge q by reflecting .

[0106] In step c), the partial charges calculated in step b are input and the interatomic interaction energy is calculated through the intermolecular interaction calculation model.

[0107] In one example, the intermolecular interaction model is expressed by the following equation, where E is the intermolecular interaction energy system It is a reactive force field model that calculates .

[0108] Formula 5)

[0109] Formula 6)

[0110] Formula 7)

[0111] Formula 8)

[0112] Formula 9)

[0113] Formula 10)

[0114]

[0115] Here, E system is the intermolecular interaction energy, E vdw is the interatomic van der Waals energy, E coulomb is the interatomic Coulomb energy, E bond is the intraatomic binding energy, E over is the energy of the superbonding state of the molecule, E angle is the intramolecular bond angle energy.

[0116] Also, D ij Energy well depth, α ij is the potential energy between atoms i and j, r ij Interatomic distance, r vdw Equilibrium bond distance, C is the Coulomb constant, q i , q j Charges of atoms i and j, respectively, r ij is the interatomic distance, is the shielding parameter, BO ijCoupling order, De, P be,1 , P over , λ over , λ angle , k a , k b is an empirical parameter, Δ i is the degree to which atoms are bonded in excess of the normal case, BO a is the combined order a, BO b is the bond order b, Φ is the interatomic bond angle, and Φ0 is the equilibrium angle.

[0117] E of Equation 5 through Equation 6 to Equation 10 system Equations 6 to 10 are equations for simulating energy according to specific conditions, E vdw represents the energy according to the interatomic distance, E coulomb represents the energy according to the distance and charge between atoms, and E bond - represents the energy according to the bond order between atoms, and E over represents the energy when an atom forms more bonds than at equilibrium, and E angle represents the energy according to the angle between the atomic bonds. Adding all the energies together yields the intermolecular energy.)

[0118] And, in step d), the interatomic interaction energy calculated in step c) is input to form ion distribution image data and charge distribution data within the battery during micro-scale charge / discharge through a molecular momentum calculation model.

[0119] In one example, the molecular momentum calculation model can calculate molecular momentum using Newton's equations of motion below.

[0120] Formula 11)

[0121] Formula 12)

[0122] Formula 13)

[0123] Here, F is the force applied to the molecule, is the total energy change of the molecule, m is the mass of the molecule, a is the acceleration of the molecule, v is the velocity of the molecule, and x is the position of the molecule.

[0124] Equation 11 is the formula for calculating the force applied to the atoms in the simulation, which is the product of the mass of the atoms and the acceleration. Equation 12 is the formula for calculating the force applied to the atoms in the simulation, which is the acceleration multiplied by the mass of the atoms. ) can be derived by multiplying the current velocity by half the acceleration * microseconds. Equation 13 can be derived by multiplying the current velocity by half the acceleration * microseconds. Equation 13 can be derived by multiplying the current position + velocity * microseconds + half the acceleration * microseconds squared.

[0125]

[0126] Referring to Fig. 4, steps 1 and 2 utilize a lattice Boltzmann model. The inputs of the lattice Boltzmann model are ion distribution image data and charge distribution data within the battery during micro-scale charge and discharge, and the output is the ion distribution at a specific point in time. The lattice Boltzmann model repeatedly predicts the propagation process and collision process of the potential, ions, and electrolyte within the battery to predict the ion distribution at a specific point in time.

[0127] In one example, the inputs to the lattice Boltzmann model may further include the initial flow rate of the electrolyte and the pressure inside the battery.

[0128] In one example, in the second step, an expected risk value for the battery and an expected life value for the battery may be additionally output.

[0129]

[0130] In one example, the charge propagation process is predicted by calculating the charge density probability distribution using the formula below.

[0131] Formula 14)

[0132] Here, h iis the discretized potential distribution function in direction i at grid cell x, where i is the discretized direction in the grid, and c i is the unit velocity vector for the discretized direction, is the non-equilibrium charge probability density function.

[0133] The probability distribution of charge density within the battery is calculated using Equation 14.

[0134]

[0135] In one example, the prediction of the ion propagation process calculates the ion density probability distribution using the formula below.

[0136] Formula 15)

[0137] Here, g i is the ion distribution function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the nonequilibrium ion probability density function.

[0138] The density probability distribution of ions in the battery is calculated using Equation 15.

[0139]

[0140] In one example, the propagation process of the flow velocity is predicted by calculating the flow velocity probability distribution using the formula below.

[0141] Formula 16)

[0142] Here, f i is the fluid probability density function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the non-equilibrium velocity probability density function.

[0143] The density probability distribution of the electrolyte flow rate within the battery is calculated using Equation 16.

[0144] What we want to obtain through Equations 14 to 16 is the probability distribution of the charge, ion, and flow rate within the battery at the next point in time. The values ​​for the probability distribution obtained through Equations 14 to 16 are used to predict the collision process described below.

[0145]

[0146] In one example, collision process prediction predicts the post-collision distribution changes of charges, ions, and electrolytes within a battery using the following formula.

[0147] Formula 17)

[0148] Formula 18)

[0149] Formula 19)

[0150] Formula 20)

[0151] Formula 21)

[0152] Formula 22)

[0153] Here, τ h is the potential collision relaxation time, is the equilibrium potential probability density function, is the time rate of change for the phase, S is the potential response value, is the time staff, W i is the weight coefficient, i0 is the exchange current density, R is the gas constant, T is the temperature, L h is the reaction coefficient, is the ion collision relaxation time, is the equilibrium ion probability density function, r k is an electrochemical reaction term, C k is the ion quantity at grid x, τ is the collision relaxation time, is the equilibrium fluid probability density function, G i is an external force in the discretized direction, is the density of the fluid, ω iis the weighting coefficient in the discretized direction, u is the velocity of the fluid, c is the unit speed of sound, and G is the dissipative resistance value.

[0154]

[0155] Equation 17 is a formula for obtaining the non-equilibrium charge probability density function mentioned in Equation 14, which describes the collision process of charge values. The difference between the equilibrium potential probability density function and the potential probability density function at the original point in time is expressed as a value divided by the potential collision relaxation time, and the non-equilibrium charge probability density function is calculated by adding the potential reaction value to this.

[0156] The equation for calculating this potential reaction value is Equation 18, which is expressed as a function of the time change rate for the phase because it describes the part where the phase change that occurs due to the electrochemical reaction inside the battery is reflected in the potential collision process, and is defined as a relationship between the potential collision relaxation time, exchange current density, gas constant, temperature, and reaction coefficient.

[0157] Equation 19 is a formula for obtaining the non-equilibrium ion probability density function mentioned in Equation 15, and describes the collision process of ion values. It represents the difference between the equilibrium ion probability density function and the ion probability density function at the original point in time, divided by the ion collision relaxation time, and the non-equilibrium ion probability density function is calculated by adding the electrochemical reaction term to this. Equation 20 is a formula for calculating the electrochemical reaction term, which means the part where the phase change caused by the electrochemical reaction, similar to the potential reaction value, is reflected in the ion collision process. Therefore, it is expressed as a function of the time change rate for the phase and the original ion distribution value.

[0158] Equation 21 is a formula for obtaining the non-equilibrium fluid probability density function mentioned in Equation 16, and describes the collision process of the fluid density value. It is expressed as the difference between the equilibrium fluid probability density function and the fluid probability density function at the original point in time, divided by the collision relaxation time, and the non-equilibrium fluid probability density function is calculated by adding the external force in the discrete direction. Equation 22 is a formula for obtaining the external force in the discrete direction included in the non-equilibrium fluid probability density function, and the external force applied here is the dissipation resistance, which is the resistance value that occurs at the boundary between the fluid and the solid during the electrochemical reaction. Therefore, this external force is defined as a relationship between the dissipation resistance value, fluid density, and fluid velocity.

[0159]

[0160] As described above, the values ​​for the probability distribution obtained through Equations 14 to 16 are used to predict the collision process. What we want to obtain through Equations 17 to 22 is to predict the change in the distribution of the flux-ion density0 charge density probability after the collision. The values ​​for the ion distribution at a specific point in time obtained through Equations 17 to 22 are then used to obtain the probability distribution of the charge, ion, and flux at the next point in time within the battery through Equations 14 to 16.

[0161] The lattice Boltzmann model iteratively predicts the propagation of potential, ions, and electrolyte within a battery using Equations 14 to 16, until the conditions are satisfied. Finally, the collision process is predicted using Equations 17 to 22, thereby predicting the ion distribution at a specific point in time. In one example, the ion distribution obtained through the lattice Boltzmann model is in the form of an image.

[0162]

[0163] In the second step, the ion distribution values ​​in the form of images acquired in the first step and the initial battery state are input, and an ion distribution prediction image is output through an LSTM-based artificial intelligence block. Fig. 5 illustrates the execution of the second step according to one embodiment of the present invention. As illustrated in Fig. 5, the artificial intelligence block outputs an ion distribution prediction image through learning by utilizing the input ion distribution values ​​and the initial battery state.

[0164] According to one embodiment of the present invention, there is an advantage in that the behavior of ions within a battery can be predicted at the meso scale beyond the micro scale.

[0165] Additionally, according to one embodiment of the present invention, there is an advantage in that the ion behavior of the SEI layer within the battery can be predicted.

[0166] Additionally, according to one embodiment of the present invention, there is an advantage in that the behavior of ions within a battery can be predicted during repeated charge and discharge cycles.

[0167] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single entity may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined manner.

[0168] The scope of the present invention is indicated by the claims described below rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention.

Claims

1. A first step of forming image data of ion distribution inside a battery by inputting boundary values ​​and initial values ​​of the battery into a data forming section including a prediction model of ion distribution inside a battery based on numerical analysis; and A method for predicting ion distribution in a lithium-sulfur battery, comprising: a second step of inputting the initial battery state and the ion distribution image data acquired in the first step into an image generation unit and outputting an ion distribution prediction image through an LSTM-based artificial intelligence block; 2. In paragraph 1, The above first step is, Step 1-1 of forming micro-scale battery ion distribution image data by inputting the above initial values ​​into the molecular dynamics simulation model; and A method for predicting ion distribution in a lithium-sulfur battery, comprising a step 1-2 of inputting the micro-scale ion distribution image data in the battery obtained in the step 1-1 into the lattice Boltzmann model and outputting the meso-scale ion distribution image data in the battery.

3. In paragraph 2, Step 1-1 above, a) A step of inputting the above initial values ​​and calculating electronegativity through a voltage propagation simulation model; b) a step of calculating a partial charge using a partial charge calculation model by inputting the electronegativity calculated in step a); c) a step of calculating the interatomic interaction energy through an intermolecular interaction calculation model by inputting the partial charge calculated in step b); and d) a molecular momentum calculation step of inputting the interatomic interaction energy calculated in step c) and forming ion distribution image data and charge distribution data within the battery during micro-scale charge and discharge through a molecular momentum calculation model; a method for predicting ion distribution within a lithium-sulfur battery, including the step of calculating the molecular momentum; 4. In paragraph 3, The above initial values ​​are a method for predicting ion distribution in a lithium-sulfur battery including the initial ion distribution in the battery and the charge / discharge voltage.

5. In paragraph 3, The above voltage propagation simulation model is, A method for predicting ion distribution in a lithium-sulfur battery, which is an EChemDID model that calculates the electronegativity using the formula below. ( is the electronegativity of the atom you want to find, is the initial electronegativity of each atom, is the electrochemical potential added to atom i, is the electrochemical potential applied to atom i over time, is the electrochemical potential applied to atom j over time, k is the effective diffusivity, is the relaxation rate, is a switching function, R ij is the distance between atoms i and j, is the local weight function, W i is the total metal coordination number, is the electronegativity of the atom to be obtained over time) 6. In paragraph 3, The above partial charge calculation model is, A method for predicting ion distribution in a lithium-sulfur battery, which is a charge equlibration (ACKS2) model that calculates the partial charge q using the formula below. (e is the molecular electric circuit, J is the blocked Coulomb interaction, q i is the charge of the i atom, q j is the charge of the j atom, X i is the electronegativity of the i atom) 7. In paragraph 3, The above molecular interaction model is, The above interatomic interaction energy E is calculated using the formula below. system A method for predicting ion distribution in a lithium-sulfur battery, which is a reactive force field model that calculates . (D ij Energy well depth, α ij is the potential energy between atoms i and j, r ij The distance between atoms, r vdw Equilibrium bond distance, C is the Coulomb constant, q i , q j Charges of atoms i and j, respectively, r ij is the interatomic distance, is the shielding parameter, BO ij Coupling order, De, P be,1 , P over , λ over , λ angle , k a , k b is an empirical parameter, Δ i is the degree to which atoms are bonded in excess of the normal case, BO a is the combined order a, BO b is the bond order b, Φ is the bond angle between atoms, Φ0 is the equilibrium angle) 8. In paragraph 3, The above molecular momentum calculation model is, A method for predicting ion distribution in a lithium-sulfur battery by calculating the molecular momentum using Newton's equations of motion below. (F is force, is the total energy change, m is the mass, a is the acceleration, v is the velocity, x is the position) 9. In paragraph 3, Steps 1-2 above, By inputting ion distribution image data and charge distribution data within the battery during micro-scale charging and discharging, Using the above lattice Boltzmann model, A method for predicting ion distribution in a lithium-sulfur battery, which predicts ion distribution at a specific point in time by repeatedly predicting the propagation process and collision process of potential, ions, and electrolyte in the battery.

10. In paragraph 9, Prediction of the above potential transmission process is, A method for predicting ion distribution in a lithium-sulfur battery, which calculates the charge density probability distribution using the following formula. (h i is the discretized potential distribution function in direction i at grid cell x, where i is the discretized direction in the grid, and c i is the unit velocity vector for the discretized direction, is the non-equilibrium charge probability density function) 11. In paragraph 9, Prediction of the propagation process of the above ions is, A method for predicting ion distribution in a lithium-sulfur battery, which comprises calculating the ion density probability distribution using the following formula. (g i is the ion distribution function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the non-equilibrium ion probability density function) 12. In paragraph 9, Prediction of the propagation process of the above flow rate is, A method for predicting ion distribution in a lithium-sulfur battery, which comprises calculating the probability distribution of the fluid density of the electrolyte using the formula below. (f i is the fluid probability density function, i is the discretized direction, c i is the velocity vector in the discretized direction, is the non-equilibrium velocity probability density function) 13. In paragraph 9, The above collision course prediction is, A method for predicting ion distribution in a lithium-sulfur battery, which predicts changes in the distribution of charges, ions, and electrolytes after collision within the battery using the following formula. (τ h is the potential collision relaxation time, is the equilibrium potential probability density function, is the time rate of change for the phase, S is the potential response value, is the time staff, W i is the weight coefficient, i0 is the exchange current density, R is the gas constant, T is the temperature, L h is the reaction coefficient, is the ion collision relaxation time, is the equilibrium ion probability density function, r k is an electrochemical reaction term, C k is the ion abundance in grid x, τ is the collision relaxation time, is the equilibrium fluid probability density function, G i is an external force in the discretized direction, is the density of the fluid, ω i is the weighting factor in the discretized direction, u is the velocity of the fluid, c is the unit sound velocity, and G is the dissipative resistance value).

14. In paragraph 9, The input to the above lattice Boltzmann model is, A method for predicting ion distribution in a lithium-sulfur battery, further including the initial flow rate of the electrolyte and the pressure inside the battery.

15. In paragraph 1, In the above second step, A method for predicting ion distribution in a lithium-sulfur battery, which additionally outputs an expected risk value of the battery and an expected life value of the battery.

16. In paragraph 1, The above boundary values ​​are, A method for predicting ion distribution in a lithium-sulfur battery, including the type of the battery's cathode, the type of the battery's anode, the initial ion distribution of the battery, and the terminal voltage of the battery.

17. In paragraph 1, The above initial battery status is, A method for predicting ion distribution in a lithium-sulfur battery including the voltage of the battery.

Images