Electrolyte formulation preparation method, storage medium and processor

Abstract

This invention provides an electrolyte formulation generation method, a storage medium, and a processor, belonging to the field of secondary battery technology. The electrolyte formulation generation method includes: determining target physical property parameters of the electrolyte and at least one battery performance indicator based on the application requirements of the battery; selecting formulations that meet the target physical property parameters from the feasible domain of electrolyte components according to an electrolyte composition prediction model, and using these as a group of candidate electrolyte formulations; estimating the performance indicator estimate of each of the candidate electrolyte formulations according to a performance indicator estimation model, and calculating the weighted sum of the estimated performance indicators for each of the at least one battery performance indicator; and selecting the electrolyte formulation whose weighted sum meets a preset standard as the recommended electrolyte formulation. The electrolyte formulation generation method of this application can comprehensively consider cost, physical property factors, and battery performance to recommend the optimal battery electrolyte formulation, achieving high-efficiency and low-cost electrolyte design.

Description

Technical Field

[0001] This invention relates to the field of secondary battery technology, and more specifically to a method for generating an electrolyte formulation, a storage medium, and a processor. Background Technology

[0002] Secondary batteries, represented by lithium-ion batteries, are used in many fields such as mobile phones, digital devices, electric vehicles, electric bicycles, power tools, and energy storage. The requirements for their energy density, cycle performance, and safety performance are becoming increasingly stringent. The electrolyte plays a crucial role in the battery, performing functions such as ion transport, electronic insulation, and forming a solid electrolyte film at the positive and negative electrode interfaces. The electrolyte in a secondary battery is generally prepared by mixing solvents, lithium salts, and necessary additives in specific proportions under certain conditions.

[0003] Currently, most methods for finding the optimal electrolyte formulation are based on human experience, continuously optimizing existing formulations. Due to the complexity of electrolyte components, a lot of work and a long time period are often required to design an electrolyte that meets actual needs.

[0004] In recent years, several methods have emerged that recommend optimal battery electrolyte formulations using mathematical statistics and machine learning algorithms. However, these methods have significant limitations in optimizing electrolyte components and often fail to address multiple battery performance requirements. For example, an electrolyte component optimization method based on a mass triangle model (CN102931436A) is only applicable to electrolyte formulations containing three components and only considers the optimization of electrolyte ionic conductivity. Conductivity primarily affects the battery's rate performance and low-temperature performance. Therefore, this method has significant limitations in optimizing electrolyte components and cannot meet the complex performance requirements of batteries. Summary of the Invention

[0005] The purpose of this invention is to provide a method that can recommend the optimal battery electrolyte formulation by comprehensively considering physical properties and multiple battery performance factors, thereby achieving high-efficiency and low-cost electrolyte design.

[0006] To achieve the above objectives, embodiments of the present invention provide a method for generating an electrolyte formulation, comprising:

[0007] Based on the application requirements of the battery, determine the target physical property parameters of the electrolyte and at least one battery performance indicator.

[0008] Based on the electrolyte composition prediction model, formulations that meet the target physical property parameters are selected from the feasible domain of electrolyte components and used as a group of candidate electrolyte formulations.

[0009] The performance index of each candidate electrolyte formulation is estimated based on the performance index estimation model. The weighted sum of the estimated performance index values ​​of each formulation in at least one battery performance index is calculated. The electrolyte formulation whose weighted sum meets the preset standard is the recommended electrolyte formulation.

[0010] Preferably, the electrolyte composition prediction model is obtained through the following steps:

[0011] Step 1: Determine the first group of electrolyte formulations based on the target physical property parameters;

[0012] Step 2: Prepare electrolyte solutions according to each of the electrolyte formulations in the first group, and test the physical properties of each electrolyte; and

[0013] Step 3: Based on the correspondence between each electrolyte formulation and the corresponding measured physical property parameters, establish an electrolyte composition prediction model through data fitting method.

[0014] Furthermore, the electrolyte formulation generation method also verifies the electrolyte component prediction model through the following steps:

[0015] Step 4: Based on the generation method of the first set of electrolyte formulations, obtain the second set of electrolyte formulations, prepare the corresponding electrolyte solutions, and test the physical property parameters of each electrolyte to verify the accuracy of the electrolyte composition prediction model; and

[0016] Step 5: If the test results show that the electrolyte composition prediction model exceeds the preset error range, then determine a supplementary electrolyte formula according to the generation method of the first set of electrolyte formulas, and repeat steps 2-4 based on the supplementary electrolyte formulas until the electrolyte composition prediction model meets the preset error range.

[0017] Alternatively, the data fitting method may be selected from the following to achieve the best fitting effect: least squares, gradient descent, iterative least squares, and Ley-Marx algorithm.

[0018] Optionally, the method for determining the feasible domain of the electrolyte components includes: determining the material composition of the electrolyte, and the upper and lower limits of the mass fraction of each material in the material composition, based on the target physical property parameters, and the sum of the mass fractions of each material in the material composition is 1.

[0019] Furthermore, the method for determining the first group of electrolyte formulations includes: generating the first group of electrolyte formulations from the feasible domain of electrolyte components by means of enumeration or at least one experimental design method.

[0020] Optionally, the experimental design method may include at least one of the following: classical screening design, response surface methodology, Taguchi design, orthogonal hypersaturation design, filling design, deterministic screening design, and mixture design.

[0021] Preferably, the training of the performance metric estimation model includes:

[0022] The electrolyte formulations in the training dataset are constructed into an electrolyte formulation matrix x according to the order of their material composition.

[0023] For each performance metric in the battery performance metrics, construct a performance metric value output vector B. α (x), where B α The i-th component B of (x) α (x i ) represents the formulation of the i-th electrolyte. i Battery performance values ​​under performance index α;

[0024] Based on x and B α (x) Construct the log-marginal likelihood function logP(B) corresponding to one of the battery performance indicators α. α (x)), and initialize the hyperparameter set ξ of the performance index estimation model;

[0025] Apply gradient descent to ξ to maximize the logarithmic marginal likelihood function logP(B) α (x)) is used to iteratively update the hyperparameter ξ value, and the final log marginal likelihood function logP(B) is obtained. α (x) serves as the prior model for the performance index estimation model corresponding to performance index α.

[0026] The training dataset consists of a set of candidate electrolyte formulations and validated electrolyte formulations, along with corresponding battery performance index values ​​for each electrolyte formulation.

[0027] Furthermore, the log-marginal likelihood function is calculated using the following formula:

[0028]

[0029] In the formula, I is the identity matrix; σ 2 For B α The variance of the noise in (x); n is the number of electrolyte formulation samples in the training dataset; P(B α (x) represents the measured cycle performance index value B under the hyperparameter set ξ. α The probability of (x) occurring;

[0030] μ(x)=[μ(x1),μ(x2),...,μ(x n )], where μ(x i ) = ax i +b, where a and b are hyperparameters in the hyperparameter set ξ;

[0031] k(x i ,x j ) represents the formulation of the i-th electrolyte. i With the j-th electrolyte formulation x j The covariance function, σ f Let l be a hyperparameter in the hyperparameter set ξ, and σ f Let be the standard deviation of the Gaussian kernel function, and l be the characteristic length scale of the Gaussian kernel function.

[0032] Preferably, the steps for obtaining the performance index estimate include:

[0033] Obtain one electrolyte formulation x* from a set of candidate electrolyte formulations, and calculate the posterior distribution formula for that electrolyte formulation;

[0034] Based on the posterior distribution formula, the conditional probability distribution formula for the battery performance index value at performance index α is determined; and

[0035] Based on the conditional probability distribution formula, estimate the mean value of the battery performance index x* with respect to performance index α. and variance and the mean This serves as an estimate of the performance index of electrolyte formulation x* in performance index α.

[0036] Preferably, the weighted sum also includes a weighted value of the electrolyte production cost.

[0037] Optionally, the electrolyte formulations that meet the preset criteria are selected as recommended electrolyte formulations, including: sorting the weighted sums from largest to smallest, and selecting the corresponding electrolyte formulations from the front to the back of the resulting sequence as recommended electrolyte formulations.

[0038] Optionally, at least one battery performance indicator includes at least one of the following: cycle performance indicator, DCR indicator, rate performance indicator, storage performance indicator, and safety performance indicator.

[0039] Optionally, the corresponding physical property parameters of the target physical property index include at least one of dielectric constant, viscosity, melting point, boiling point, and ionic conductivity.

[0040] Preferably, the electrolyte formulation comprises at least a solvent, an electrolyte salt, and additives, and the additives include at least one of film-forming additives, flame-retardant additives, overcharge prevention additives, and overcharge protection additives.

[0041] Electrolyte salts include at least one of hexafluorophosphate, difluorosulfonyl imide salts, nitrates, nitrites, fluorides, chlorides, bromides, iodides, difluorophosphate, difluorooxalate borate, dioxalate borate, tetrafluorooxalate borate, difluorosulfonyl imide salts, and bis(trifluoromethanesulfonyl) imide salts; and

[0042] The cations in electrolyte salts include at least one of the following: Li + Na + K+, Mg 2+ Ca 2+ Zn 2+ Al 3+ .

[0043] On the other hand, the present invention provides a machine-readable storage medium storing instructions for causing a machine to execute the electrolyte formulation generation method of the present application.

[0044] In another aspect, the present invention provides a processor for running a program, wherein the program is run to execute the electrolyte formulation generation method of the present application.

[0045] The above technical solution realizes an electrolyte formulation generation method based on hierarchical statistical learning. First, a set of candidate electrolyte formulations is initially screened through the target physical property parameters of the electrolyte and the corresponding electrolyte component prediction model. Then, the performance index estimation model estimates the performance index of each formulation in at least one battery performance index. Finally, the optimal electrolyte formulation in the set of candidate electrolyte formulations is recommended by the weighted sum of the estimated values ​​of each formulation under each performance index. The weighted sum can also further include the cost weighting value of the electrolyte, so as to comprehensively consider cost, physical property factors and battery performance to recommend the optimal battery electrolyte formulation, and realize the high-efficiency and low-cost design of electrolyte. Attached Figure Description

[0046] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0047] Figure 1 This is a flowchart of an embodiment of the electrolyte formulation generation method of this application;

[0048] Figure 2 This is a flowchart of the preliminary screening stage in another embodiment of the electrolyte formulation generation method of this application; and

[0049] Figure 3 yes Figure 2In this embodiment, a flowchart is provided for selecting a recommended formulation from a set of candidate electrolyte formulations obtained in the initial screening stage. Detailed Implementation

[0050] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0051] This invention provides a method for generating an electrolyte formulation. The electrolyte generated by this method can be a lithium-ion battery electrolyte or an electrolyte for other power batteries. The implementation process is as follows: Figure 1 As shown, it includes:

[0052] Step 1: Based on the application requirements of the battery, determine the target physical property parameters of the electrolyte and at least one battery performance indicator;

[0053] Step 2: Based on the electrolyte composition prediction model, select formulations that meet the target physical property parameters from the feasible domain of electrolyte components, and use them as a group of candidate electrolyte formulations.

[0054] Step 3: Estimate the performance index of each candidate electrolyte formulation based on the performance index estimation model, calculate the weighted sum of the estimated performance index values ​​of each formulation in at least one battery performance index, and select the electrolyte formulation whose weighted sum meets the preset standard as the recommended electrolyte formulation.

[0055] It should be noted that steps one and two are the initial screening stages of electrolyte formulation, which can obtain a set of candidate formulations that meet the target physical property parameters. Step three is to further select the optimal formulation that meets at least one battery performance index from the set of candidate formulations obtained from the initial screening through a performance index estimation model.

[0056] In step one, the target physical property parameters of the electrolyte and the battery performance parameters are set according to the application requirements of the battery to be designed. The target physical property parameters are used for the initial screening of the electrolyte formulation, and the battery performance parameters are used for further screening based on the initial screening.

[0057] The initial screening process can, for example, use the electrolyte composition prediction model in step two to initially screen out a set of candidate electrolyte formulations that meet the target physical property parameters from a large number of randomly generated electrolyte formulations. This set of candidate electrolyte formulations obtained in this way does not rely on the existing experience of the technical personnel, nor on the existing formulation base. Therefore, in general, a set of candidate electrolyte formulations that are outside the range expected by the technical personnel can be obtained in the initial screening stage.

[0058] In step two, the method for determining the feasible domain of electrolyte components includes: determining the material composition of the electrolyte, the upper limit and lower limit of the mass fraction of each material in the material composition, based on the target physical property parameters, and the sum of the mass fractions of each material in the material composition is 1.

[0059] The further screening process based on the initial screening corresponds to step three above. The performance index estimate value of each of the candidate electrolyte formulations in a set of candidate electrolyte formulations is estimated by the performance index estimation model, and the weighted sum of the performance index estimates is obtained. This allows for a comprehensive evaluation of each candidate formulation in a set of candidate electrolyte formulations in at least one battery performance index. Therefore, the best formulation that meets each of the at least one battery performance index can be further screened from the set of candidate electrolyte formulations obtained in the initial screening stage and recommended.

[0060] It should be noted that the above-mentioned optimal formulation can be multiple formulations that rank highly in comprehensive evaluation. For example, if the estimated values ​​of performance indicators are not significantly different, or if electrolytes are prepared according to the above-mentioned multiple optimal formulations and the measured parameters are not significantly different or each has its own advantages and disadvantages, then multiple recommended formulations with better performance can be selected from a set of candidate electrolyte formulations.

[0061] Furthermore, based on the above-mentioned multiple recommended formulas, other factors such as production costs and environmental pollution can be considered to select the final recommended formula.

[0062] In some implementations, at least one battery performance indicator includes at least one of the following: cycle performance indicator, DCR indicator, rate performance indicator, storage performance indicator, and safety performance indicator.

[0063] In some implementations, the corresponding physical property parameters of the target physical property index include at least one of dielectric constant, viscosity, melting point, boiling point, and ionic conductivity.

[0064] In some embodiments, the electrolyte formulation comprises at least a solvent, an electrolyte salt, and additives, wherein the additives include at least one of film-forming additives, flame-retardant additives, overcharge protection additives, and overcharge protection additives.

[0065] Electrolyte salts include at least one of hexafluorophosphate, difluorosulfonyl imide salt, nitrate, nitrite, fluoride, chloride, bromide, iodide, difluorophosphate, difluorooxalate borate, dioxalate borate, tetrafluorooxalate borate, difluorosulfonyl imide salt, and bis(trifluoromethanesulfonyl) imide salt.

[0066] The cations in electrolyte salts include at least one of the following: Li + Na + K+, Mg2+ Ca 2+ Zn 2+ Al 3+ .

[0067] The electrolyte formulation generation method of this embodiment has the following technical advantages compared with the prior art:

[0068] (1) By using an electrolyte composition prediction model, a set of candidate electrolyte formulations that meet the target physical property parameters are obtained in the initial screening stage. On the one hand, this set of candidate electrolyte formulations does not rely on the existing experience of technical personnel, nor on the existing formulation basis, and can obtain electrolyte candidate formulations that are outside the expected range of technical personnel.

[0069] On the other hand, by using an electrolyte composition prediction model to conduct preliminary screening among a large number of randomly generated electrolyte formulations, we can avoid aimlessly searching through a large number of electrolyte formulations, thereby avoiding the manpower and material costs of preparing a large number of electrolytes and testing electrolyte parameters. Furthermore, preliminary screening from a large number of random formulations can avoid missing potentially excellent formulations.

[0070] (2) During the re-screening process using the performance index estimation model, formulas that simultaneously meet both physical property parameters and battery performance indicators can be found among the candidate formulas obtained from the initial screening, avoiding the manpower and material costs of large-scale electrolyte preparation and electrolyte parameter testing. At the same time, it can also help technicians further discover the correspondence between electrolyte composition and battery performance indicators.

[0071] Based on this embodiment, the implementation effect of the initial screening stage can be optimized through the following preferred implementation methods, the process of which is as follows: Figure 2 As shown.

[0072] In some implementations, a reliable electrolyte composition prediction model is obtained through the following steps:

[0073] Step 1: Determine the first group of electrolyte formulations based on the target physical property parameters;

[0074] Step 2: Prepare electrolyte solutions according to each of the electrolyte formulations in the first group, and test the physical properties of each electrolyte; and

[0075] Step 3: Based on the correspondence between each electrolyte formulation and the corresponding measured physical property parameters, establish an electrolyte composition prediction model through data fitting method.

[0076] Among them, the data fitting method is selected from the following to achieve the best fitting effect: least squares method, gradient descent method, iterative least squares, and Ley-Marx algorithm.

[0077] In some implementations, the correctness of the obtained electrolyte composition prediction model can be further verified and optimized through the following steps:

[0078] Step 4: Based on the generation method of the first set of electrolyte formulations, obtain the second set of electrolyte formulations, prepare the corresponding electrolyte solutions, and test the physical property parameters of each electrolyte to verify the accuracy of the electrolyte composition prediction model; and

[0079] Step 5: If the test results show that the electrolyte composition prediction model exceeds the preset error range, then determine a supplementary electrolyte formula according to the generation method of the first set of electrolyte formulas, and repeat steps 2-4 based on the supplementary electrolyte formulas until the electrolyte composition prediction model meets the preset error range.

[0080] The first electrolyte formulation can be determined through the following steps:

[0081] A first set of electrolyte formulations is generated from the feasible domain of electrolyte components using an enumeration method or at least one experimental design method. The experimental design method may be at least one of the following: classical screening design, response surface methodology, Taguchi design, orthogonal hypersaturation design, filling design, deterministic screening design, or mixing design.

[0082] To further illustrate the process of the initial screening stage in the electrolyte formulation generation method of this application, the following describes the actual process of obtaining a set of candidate electrolyte formulations in an implemented case.

[0083] 1. The battery application is for low-temperature fast charging scenarios. The selected solvents include EC (ethylene carbonate), PC (propylene carbonate), EMC (ethyl methyl carbonate), DMC (dimethyl carbonate), and DEC (diethyl carbonate). The most commonly used lithium salt is lithium hexafluorophosphate (LiPF6), with a lithium salt concentration of 1 mol·L⁻¹. -1 ;

[0084] 2. Based on the battery application scenario, the mass fraction ranges for EC, PC, EMC, DEC, and DMC are set to 0.1-0.3, 0-0.1, 0-0.7, 0-0.3, and 0-0.3, respectively. The mass fractions of each solvent satisfy the following relationship:

[0085] EC+PC+DMC+DEC+EMC=1 (1)

[0086] 3. The experimental design method was a mixture design, ensuring that the sum of the proportions of the specified variables equaled 1. Table 1 below shows a set of electrolyte formulations (for the solvent, the lithium salt was selected in the previous steps as having a concentration of 1 mol·L⁻¹). -1 Lithium hexafluorophosphate).

[0087] Table 1. Solvent Composition in Mixing Design

[0088]

[0089]

[0090] 4. After preparing the electrolyte in the order shown in Table 1, measure the conductivity at -10℃.

[0091] 5. The electrolyte composition prediction model is established by selecting the method with the smallest fitting error among the following fitting methods: least squares, gradient descent, iterative least squares, and Ley-Marx algorithm. In this embodiment, the least squares method has a smaller root mean square error, making the model effective. The formula is as follows:

[0092]

[0093] Where m represents the number of samples and n represents the parameter dimension. In this example, the parameter dimension includes the main effects of each solvent component and the pairwise interactions between solvents. Vectorizing the above equation yields:

[0094] Xβ=y (3)

[0095]

[0096] To find the best estimate of β The problem can be transformed into the following using the least squares method:

[0097] S(β)=||Xβ-y|| (5)

[0098] By differentiating S(β) to find the maximum and minimum values, we can obtain:

[0099] X T Xβ=X T y (6)

[0100] If X T If X is not singular, then β has a unique solution:

[0101]

[0102] 6. Several predicted electrolyte formulations were randomly selected and prepared into electrolyte solutions for mathematical model accuracy verification. Experimental errors were all within ±0.11 ms / cm, and the predicted sample conductivity order was consistent with the experimental results, ensuring the validity of the simulation results. The results are shown in Table 2 below:

[0103] Table 2 Simulation Results: Predicted Values ​​and Measured Values

[0104]

[0105] 7. Based on battery application requirements, the performance indicators are set as follows: conductivity at -10℃ >6ms / cm;

[0106] 8. 10,000 electrolyte formulations were randomly and uniformly generated. The conductivity at -10℃ was predicted using an electrolyte composition prediction model. Electrolyte formulations that meet the requirements were then selected.

[0107] exist Figure 1 Based on the illustrated embodiments, the implementation effect of the re-screening stage can be optimized through the following preferred implementation methods, the process of which is as follows: Figure 3 As shown.

[0108] In some implementations, the performance metric estimation model is trained using a Gaussian process regression method, the process of which includes:

[0109] The electrolyte formulations in the training dataset are constructed into an electrolyte formulation matrix x according to the order of their material composition.

[0110] For each performance metric in the battery performance metrics, construct a performance metric value output vector B. α (x), where B α The i-th component B of (x) α (x i ) represents the formulation of the i-th electrolyte. i Battery performance values ​​under performance index α;

[0111] Based on x and B α (x) Construct the log-marginal likelihood function logP(B) corresponding to one of the battery performance indicators α. α (x)), and initialize the hyperparameter set ξ of the performance index estimation model;

[0112] Apply gradient descent to ξ to maximize the logarithmic marginal likelihood function logP(B) α (x)) is used to iteratively update the hyperparameter ξ value, and the final log marginal likelihood function logP(B) is obtained. α (x) serves as the prior model for the performance index estimation model corresponding to performance index α.

[0113] The training dataset consists of a set of candidate electrolyte formulations and validated electrolyte formulations, along with the corresponding battery performance metrics for each formulation. It should be noted that the training dataset should select electrolyte formulations that have been experimentally verified to meet the application requirements of the designed battery, and at least satisfy the corresponding target physical property parameters or performance metrics.

[0114] Furthermore, the log-marginal likelihood function is calculated using the following formula:

[0115]

[0116] In the formula, I is the identity matrix; σ 2 For B α The variance of the noise in (x); n is the number of electrolyte formulation samples in the training dataset; P(B α (x) represents the measured cycle performance index value B under the hyperparameter set ξ. α The probability of (x) occurring;

[0117] μ(x)=[μ(x1),μ(x2),...,μ(x n )], where μ(x i ) = ax i +b, where a and b are hyperparameters in the hyperparameter set ξ;

[0118] k(x i ,x j ) represents the formulation of the i-th electrolyte. i With the j-th electrolyte formulation x j The covariance function, σ f Let l be a hyperparameter in the hyperparameter set ξ, and σ f Let be the standard deviation of the Gaussian kernel function, and l be the characteristic length scale of the Gaussian kernel function.

[0119] In some implementations, the steps for obtaining the performance index estimate include:

[0120] Obtain one electrolyte formulation x* from a set of candidate electrolyte formulations, and calculate the posterior distribution formula for that electrolyte formulation;

[0121] Based on the posterior distribution formula, the conditional probability distribution formula for the battery performance index value at performance index α is determined; and

[0122] Based on the conditional probability distribution formula, estimate the mean value of the battery performance index x* with respect to performance index α. and variance and the mean This serves as an estimate of the performance index of electrolyte formulation x* in performance index α.

[0123] In some implementations, the weighted sum also includes a weighted value of the electrolyte production cost. The weighted sums of the candidate formulations are sorted from largest to smallest, and the corresponding electrolyte formulations are selected from the front to the back of the resulting sequence as the recommended electrolyte formulations.

[0124] To further illustrate the process of the re-screening stage in the electrolyte formulation generation method of this application, the following describes the actual process of obtaining a recommended formulation from a set of candidate electrolyte formulations in an implemented case.

[0125] Step 1, Data Preprocessing: The electrolyte formulation is processed into a matrix form according to the order of solvent, electrolyte salt, and additives;

[0126] Step 2: Input the electrolyte formulation matrix and battery performance index values ​​D = {x, B(x)}. The battery performance index values ​​include any item from cycle performance testing, DCR testing, rate performance testing, storage performance testing, and safety testing. Since the prediction methods for each battery performance index value and its confidence interval are the same, subsequent steps will use the prediction of cycle performance test index values ​​and their confidence intervals as an example. Subsequent steps will use a new dataset D = {x, B(x)}. α (x)}, where x represents the input electrolyte formulation vector matrix, B α (x) represents the output vector of the cyclic performance test index values ​​in the Gaussian model;

[0127] Step 3: Train the model for the cyclic performance test index based on Gaussian process regression (GPR). The calculation process is as follows:

[0128] 3.1) The cyclic performance test index values ​​include the determined cyclic performance test index values ​​and random noise, and the calculation formula is as follows:

[0129] B α (x i )=f α (x i )+ε i (8)

[0130] Where, i∈[1,n], i∈Z, x i For any input electrolyte formulation, ε i The noise is Gaussian and independent, and follows a pattern with a mean of 0 and a variance of σ. 2 The Gaussian distribution, i.e., ε i ~N(0,σ 2 f α (x i ) indicates that x i The relevant definite cyclic performance test index values ​​are determined by a mean function μ(x) i ) and a covariance function k(x) i ,x j Determine the function f. α (x i The process is a Gaussian process, expressed as follows:

[0131] fα (x i )~N[(μ(x i ),k(x i ,x j (9)

[0132] 3.2) In practical applications, defining a Gaussian process requires specifying the mean function and covariance function terms. Here, the Gaussian kernel function is used as the covariance kernel function. In this study, the expressions for the mean function and covariance function are as follows:

[0133] μ(x i ) = ax i +b (10)

[0134]

[0135] Where a and b represent the hyperparameters of the Gaussian model, σ f The standard deviation of the Gaussian kernel function is represented by l, and the characteristic length scale of the kernel function is represented by l.

[0136] 3.3) By the properties of the joint Gaussian distribution, we can obtain that for the sample set {x, B} α For (x)}, B α (x) also follows a Gaussian distribution, as shown below:

[0137] B α (x)~N[(μ(x),k(x,x)+σ 2 I] (12)

[0138] Where I is the identity matrix, σ 2 Let μ(x) be the variance term caused by noise, k(x,x) be the mean function, and k(x,x) be the covariance term. The expression for the mean function μ(x) is as follows:

[0139] μ(x)=[μ(x1),μ(x2),...,μ(x n (13)

[0140] The covariance matrix k(x,x) of the input electrolyte formulation sample set is calculated using the following formula:

[0141]

[0142] 3.4) Let the hyperparameter set Initialize the hyperparameters. Then calculate the log-marginal likelihood function, using the following formula:

[0143]

[0144] Where n is the sample size, P(B) α(x) represents the measured cycle performance index value B under the hyperparameter set ξ. α The probability of (x) occurring is maximized by performing gradient descent on ξ to update the hyperparameter ξ value iteratively.

[0145] Step 4: Input the target input, which is the electrolyte formulation to be designed obtained from the preliminary screening stage of physical property parameters. * As can be seen from the explanation in step 3.1, B α (x * () represents the cycle performance test index value of the electrolyte to be designed.

[0146] B α (x * )=f α (x * )+ε (16)

[0147] Where f α (x * The distribution follows a Gaussian distribution, and its mean function is μ(x). * ) = ax * +b, where the values ​​of a and b have been calculated in step 3.4, k(x * ,x * The Gaussian kernel function is used for calculation.

[0148] According to the properties of the Gaussian distribution, [B] α (x),f α (x * The joint distribution of ] is still a Gaussian distribution, and the corresponding prior expression for the joint Gaussian distribution is as follows:

[0149]

[0150] Where, k(x,x) * The training set inputs electrolyte formulation x and the target inputs electrolyte formulation x. * The covariance matrix, k(x) * k(x,x) is a subset of k(x,x) * The transpose of k(x, x) * Taking the calculation of ) as an example, the calculation formula is as follows:

[0151]

[0152] Where, k(x) * ,x i All of them are calculated from the derived hyperparameters of equation (11);

[0153] Through Bayesian inference calculations, f can be determined. α (x * The posterior distribution of ) is:

[0154] P(f α (x * B α (x))~N[(μ * ,(σ * ) 2 (19)

[0155] in:

[0156] μ * =k(x * ,x)[k(x,x)+σ 2 I] -1 [B α [(x)-μ(x)]+μ(x) * (20)

[0157] (σ * ) 2 =k(x * ,x * )-k(x * ,x)[k(x,x)+σ 2 I] -1 k(x,x * ) (twenty one)

[0158] In the formula, μ * Let (σ*) be the mean of the predicted values. 2 This represents the variance of the predicted values.

[0159] Given the target input electrolyte formulation x * The average predicted value μ of the cycle performance test index of the target input electrolyte formulation can be estimated based on the constructed GPR-based cycle performance test index value model. * and variance (σ) * ) 2 +σ 2 Similarly, other battery performance indicators can be estimated based on steps 1 through 4.

[0160] Step 5: Calculate the cost of each electrolyte formulation;

[0161] Step 6: Based on the requirements, the results of battery performance indicators and costs are weighted and summed to rank the electrolyte formulations and provide recommended electrolyte formulations.

[0162] This invention provides a storage medium storing a program that, when executed by a processor, implements the electrolyte formulation generation method of this invention.

[0163] This invention provides a processor for running a program, wherein the program executes the electrolyte formulation generation method of this invention.

[0164] This invention provides a device including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the electrolyte formulation generation method of this invention. The device described herein can be a server, PC, PAD, mobile phone, etc.

[0165] This application also provides a computer program product that, when executed on a data processing device, is adapted to execute a program that initializes the electrolyte formulation generation method steps of the present invention.

[0166] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0167] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0168] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0169] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0170] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0171] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0172] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0173] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0174] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for preparing an electrolyte formulation, comprising: Based on the application requirements of the battery, determine the target physical property parameters of the electrolyte and at least one battery performance indicator. Based on the electrolyte composition prediction model, formulations that meet the target physical property parameters are selected from the feasible domain of electrolyte components and used as a group of candidate electrolyte formulations. The performance index of each candidate electrolyte formulation is estimated using a performance index estimation model. A weighted sum of the estimated performance indexes for each of the at least one battery performance index is then calculated. The electrolyte formulation whose weighted sum meets a preset standard is then selected as the recommended electrolyte formulation. The electrolyte composition prediction model is obtained through the following steps: Step 1: Determine the first electrolyte formulation based on the target physical property parameters; Step 2: Prepare electrolyte solutions according to each of the electrolyte formulations in the first group, and test the physical properties of each electrolyte. and Step 3: Based on the correspondence between each electrolyte formulation and the corresponding measured physical property parameters, a prediction model for the electrolyte composition is established using a data fitting method. The method for determining the feasible domain of the electrolyte components includes: determining the material composition of the electrolyte, and the upper and lower limits of the mass fraction of each material in the material composition, based on the target physical property parameters, wherein the sum of the mass fractions of each material in the material composition is 1; and The method for determining the formulation of the first group of electrolytes includes: The first set of electrolyte formulations is generated from the feasible domain of electrolyte components using an enumeration method or at least one experimental design method. The training of the performance metric estimation model includes: The electrolyte formulations in the training dataset are constructed into an electrolyte formulation matrix x according to the order of their material composition. For each of the battery performance metrics, construct a performance metric value output vector. ,in, The i Each component For the first i Electrolyte formulation In performance metrics The battery performance indicators are as follows; Based on the x and the Construct a performance index corresponding to one of the battery performance indicators. log-marginal likelihood function And initialize the hyperparameter set ξ of the performance index estimation model; Gradient descent is applied to ξ to maximize the log-marginal likelihood function. This process is used to iteratively update the hyperparameter ξ value, resulting in the final log-marginal likelihood function. As corresponding to the performance indicators The prior model of the performance index estimation model, The training dataset consists of: a set of candidate electrolyte formulations and validated electrolyte formulations, along with the corresponding battery performance index values ​​for each electrolyte formulation. The steps for obtaining the performance index estimate include: Obtain one electrolyte formulation x* from the set of candidate electrolyte formulations, and calculate the posterior distribution formula of the electrolyte formulation; Based on the aforementioned posterior distribution formula, the performance index is determined. The conditional probability distribution formula for the battery performance index values; and Based on the conditional probability distribution formula, estimate the x* in the performance index. The average value of battery performance indicators and variance and the mean As the electrolyte formulation x* in the performance indicators Estimated performance metrics.

2. The method for generating the electrolyte formulation according to claim 1, characterized in that, The electrolyte composition prediction model was further validated through the following steps: Step 4: Based on the generation method of the first group of electrolyte formulations, obtain the second group of electrolyte formulations, prepare the corresponding electrolyte solutions, and test the physical property parameters of each electrolyte to verify the accuracy of the electrolyte composition prediction model. and Step 5: If the test results show that the electrolyte composition prediction model exceeds the preset error range, then a supplementary electrolyte formula is determined according to the generation method of the first set of electrolyte formulas, and steps 2-4 are repeated based on the supplementary electrolyte formulas until the electrolyte composition prediction model meets the preset error range.

3. The method for generating the electrolyte formulation according to claim 1, characterized in that, The data fitting method is selected from the following to achieve the best fitting effect: least squares method, gradient descent method, iterative least squares method, and Ley-Marx algorithm.

4. The method for generating the electrolyte formulation according to claim 1, characterized in that, The experimental design method includes at least one of the following: classical screening design, response surface methodology, Taguchi design, orthogonal hypersaturation design, filling design, deterministic screening design, and mixture design.

5. The method for generating the electrolyte formulation according to claim 1, characterized in that, The log-marginal likelihood function is calculated using the following formula: ， In the formula, It is the identity matrix; for The variance of the noise in the training dataset; n is the number of samples of electrolyte formulations in the training dataset; To measure the actual cyclic performance test index values ​​under the hyperparameter set ξ The probability of occurrence; ,in, a and b are hyperparameters in the hyperparameter set ξ; , For the i-th electrolyte formulation With the j-th electrolyte formulation The covariance function, and Let be the hyperparameters in the hyperparameter set ξ, and Let be the standard deviation of the Gaussian kernel function. is the characteristic length scale of the Gaussian kernel function.

6. The method for generating the electrolyte formulation according to claim 1, characterized in that, The weighted sum also includes a weighted value for the production cost of the electrolyte.

7. The method for generating the electrolyte formulation according to claim 1, characterized in that, The step of using the weighted sum and corresponding electrolyte formulations that meet preset standards as recommended electrolyte formulations includes: The weighted sums are sorted from largest to smallest, and the corresponding electrolyte formulations are selected from the front to the back of the resulting sequence as recommended electrolyte formulations.

8. The method for generating an electrolyte formulation according to claim 1, characterized in that, The at least one battery performance indicator includes at least one of the following: cycle performance indicator, DCR indicator, rate performance indicator, storage performance indicator, and safety performance indicator.

9. The method for generating an electrolyte formulation according to claim 1, characterized in that, The corresponding physical property parameters of the target physical property index include at least one of dielectric constant, viscosity, melting point, boiling point, and ionic conductivity.

10. The method for generating an electrolyte formulation according to claim 1, characterized in that, The electrolyte formulation comprises at least a solvent, an electrolyte salt, and additives, and the additives include at least one of film-forming additives, flame-retardant additives, overcharge prevention additives, and overcharge protection additives. The electrolyte salt comprises at least one of hexafluorophosphate, difluorosulfonyl imide salt, nitrate, nitrite, fluoride, chloride, bromide, iodide, difluorophosphate, difluorooxalate borate, dioxalate borate, tetrafluorooxalate borate, difluorosulfonyl imide salt, and ditrifluoromethanesulfonyl imide salt; and The cations in the electrolyte salt include at least one of the following: Li + Na + K+, Mg 2+ Ca 2+ Zn 2+ Al 3+ .

11. A machine-readable storage medium storing instructions for causing a machine to perform: the electrolyte formulation generation method as described in any one of claims 1-10.

12. A processor, characterized in that, For running a program, wherein the program is run to execute: the electrolyte formulation generation method as described in any one of claims 1-10.

Images