Battery simulation model generation method, electronic device, and computer-readable storage medium

Abstract

Embodiments of the present application provide a battery simulation model generation method, an electronic device and a computer readable storage medium. It relates to the technical field of lithium battery modeling. The method comprises: determining a target mapping mode according to a parameter type of a parameter included in measured data; determining a thermodynamic parameter of a semi-electrode using the target mapping mode; generating an initial model of a target battery based on the thermodynamic parameter; generating an initial population according to initial values of each to-be-determined parameter, performing first iteration processing with the initial population as a starting point, and determining a target individual that minimizes a target function; determining an iteration direction of the target function based on second derivative information; performing second iteration processing based on the iteration direction, and determining target values of each to-be-determined parameter when the target function meets a predetermined termination condition; and setting the initial model based on the target values of each to-be-determined parameter to obtain a target model. The present application solves the technical problem of limited experimental data and unsatisfactory simulation accuracy of the battery model in related technologies.

Description

Technical Field

[0001] This application relates to the field of lithium battery modeling technology, and in particular to a method for generating a battery simulation model, an electronic device, and a computer-readable storage medium. Background Technology

[0002] Lithium-ion batteries are among the most common energy storage and conversion devices, widely used in electric vehicles, energy storage systems, and portable electronic devices. In practical applications, users typically focus on key performance indicators such as charging speed, capacity and range, cycle life, and operational safety, which directly depend on the battery's internal state evolution. To achieve accurate state assessment and optimized control, electrochemical models are widely used because they can reveal the multi-physics coupling mechanisms within the battery. These models can systematically describe key processes such as lithium-ion diffusion in electrode materials, migration in the electrolyte, electrochemical reaction kinetics, and thermal effects, thus holding significant value in battery simulation and management systems.

[0003] However, electrochemical models typically rely on a large number of thermodynamic and kinetic parameters, such as electrode porosity, particle radius, electrode thickness, charge transfer rate constant, and diffusion coefficient. These parameters are usually nested within multiple nonlinear coupled equations, which are numerous and complex, making them difficult to obtain directly or accurately measure experimentally. Under conditions of insufficient information, the acquisition of parameters for electrochemical models is limited. Related technologies often use subjective experience to set parameters that are difficult to obtain, resulting in simulation results that are less accurate than actual battery responses. Summary of the Invention

[0004] This application provides a method for generating a battery simulation model, an electronic device, and a computer-readable storage medium to alleviate or solve the technical problem in related technologies where the simulation accuracy of battery models is not ideal due to limitations in experimental data.

[0005] In a first aspect, embodiments of this application provide a method for generating a battery simulation model, comprising:

[0006] The target mapping method is determined based on the types of parameters included in the measured data. The measured data is the data obtained from actual measurements of the target battery. The target mapping method is used to map the measured data from the full electrode to the half electrode. The full electrode is an electrode state including both positive and negative electrodes, and the half electrode is an electrode state including either the positive electrode or the negative electrode.

[0007] The measured data are processed using the target mapping method to determine the thermodynamic parameters of the half-electrode, including the lithium intercalation boundary value of the half-electrode.

[0008] Based on the thermodynamic parameters, an initial model of the target battery is generated, and the initial model contains various undetermined parameters other than the thermodynamic parameters.

[0009] An initial population is generated based on the initial values ​​of each undetermined parameter. A first iteration is performed starting from the initial population within a first predetermined parameter range to determine the target individual that minimizes the objective function. The initial values ​​are random initialization values ​​within the first predetermined parameter range. The objective function is used to represent the difference between the simulation parameters of the initial model and the measured parameters of the target battery. The target individual is generated based on the intermediate values ​​of each undetermined parameter.

[0010] Based on the second derivative information of the objective function, the iteration direction of the objective function is determined, and the iteration direction is the direction pointing to the minimum value of the objective function;

[0011] Starting with the current gradient value of the objective function, a second iteration is performed based on the iteration direction to determine the target values ​​of each undetermined parameter when the objective function satisfies the predetermined termination condition. The current gradient value is the gradient value of the objective function changing towards the minimum value when each undetermined parameter is an intermediate value.

[0012] The initial model is set based on the target values ​​of each undetermined parameter to obtain the target model.

[0013] Secondly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor implements any of the methods of embodiments of this application when executing the computer program.

[0014] Thirdly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method of any one of the embodiments of this application.

[0015] Based on the battery simulation model method described in the first aspect above, this application has at least the following beneficial effects or advantages: By utilizing limited data under low-information conditions, the thermodynamic parameters involved in the electrochemical model of lithium-ion batteries are individually identified, while the remaining undetermined parameters are comprehensively identified. Substituting the thermodynamic parameters and combining the identified electrochemical model sensitivity analysis results with the measured data from conventional battery charge-discharge experiments, other parameters in the electrochemical model are further evaluated. Based on limited simulation or observation data, efficient identification of key operating condition parameters is achieved. The use of a combination of heuristic algorithms and gradient optimization algorithms to efficiently identify and correct the parameters of the lithium-ion battery electrochemical model effectively reduces the difficulties in the practical application of the electrochemical model and significantly improves the accuracy of simulation results. This not only improves the adaptability and feasibility of the model but also provides a practical technical path for the state assessment and performance prediction of lithium-ion batteries under actual operating conditions.

[0016] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application, it can be implemented according to the contents of the specification. In order to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0017] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the various drawings denote the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings depict only some embodiments according to this application and should not be construed as limiting the scope of this application.

[0018] Figure 1 A flowchart illustrating the method for generating a battery simulation model according to an embodiment of this application is shown;

[0019] Figure 2 A first schematic diagram of a method for generating a battery simulation model according to an embodiment of this application is shown;

[0020] Figure 3 A schematic diagram illustrating the principle of the battery simulation model generation method according to an embodiment of this application is shown;

[0021] Figure 4 An iterative schematic diagram of the battery simulation model generation method according to an embodiment of this application is shown;

[0022] Figure 5 A flowchart illustrating the method for generating a battery simulation model according to an embodiment of this application is shown.

[0023] Figure 6 A sensitivity diagram of the battery simulation model generation method according to an embodiment of this application is shown;

[0024] Figure 7 A second schematic diagram of a method for generating a battery simulation model according to an embodiment of this application is shown;

[0025] Figure 8 A third schematic diagram of the battery simulation model generation method according to an embodiment of this application is shown;

[0026] Figure 9 A fourth schematic diagram of the battery simulation model generation method according to an embodiment of this application is shown;

[0027] Figure 10 A block diagram of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0028] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the concept or scope of this application. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0029] To facilitate understanding of the technical solutions of the embodiments of this application, the relevant technologies of the embodiments of this application are described below. The following relevant technologies are optional solutions and can be combined with the technical solutions of the embodiments of this application in any way, and all of them fall within the protection scope of the embodiments of this application.

[0030] Each parameter has varying degrees of sensitivity to the model output, making effective adjustment through trial and error difficult under limited information conditions. This significantly restricts the efficiency and versatility of electrochemical models in engineering practice, especially in various scenarios where detailed experimental data is commonly lacking. Traditional electrochemical model parameter identification typically relies on numerous complex experiments or high-precision measurement equipment, such as electrochemical impedance spectroscopy (EIS), high-rate charge-discharge testing, or multi-physics field collaborative measurement. This is not only costly but also hinders its widespread application in actual battery cell production and management processes. In contrast, the parameter identification method based on intelligent optimization algorithms proposed in this application, especially evolutionary optimization methods represented by genetic algorithms, can achieve efficient and accurate model parameter identification by relying on only a portion of basic data. This basic data includes, but is not limited to: the cell's cutoff voltage, positive and negative electrode material types, capacity, and a small amount of standard rate (e.g., 0.1C, 0.5C, 1C) charge-discharge experimental data. The above information is generally available during the cell manufacturing testing phase. Therefore, this method can be directly applied to existing processes without the need for additional testing equipment and costs. It has good practicality and scalability, and is especially suitable for application scenarios that lack first-hand battery experimental data.

[0031] It should be noted that the application scenarios or examples provided in the embodiments of this application are for ease of understanding, and the embodiments of this application do not specifically limit the application of the technical solutions. In addition, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0032] The technical solution of this application and how it solves the aforementioned technical problems are described in detail below with specific embodiments. The listed specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0033] Figure 1 A flowchart illustrating the method for generating a battery simulation model according to an embodiment of this application is shown, such as... Figure 1 As shown, the method may include any one of steps S101 to S107.

[0034] Step S101: Determine the target mapping method based on the types of parameters included in the measured data. The measured data is the data obtained from the actual measurement of the target battery. The target mapping method is used to map the measured data from the full electrode to the half electrode. The full electrode is the electrode state including the positive electrode and the negative electrode, and the half electrode is the electrode state including the positive electrode or the negative electrode.

[0035] In the embodiments provided in this application, based on the types of parameters included in the measured data of the target battery, a suitable target mapping method is selected from multiple preset mapping algorithms to achieve accurate conversion of full-electrode macroscopic data to half-electrode thermodynamic parameters. The full electrode is a combination of positive and negative electrodes, while the half-electrode is a single positive or negative electrode. This method eliminates the need for costly and destructive experimental data; mapping can be achieved using only conventional measured data such as cutoff voltage, basic charge-discharge curves, and open-circuit voltage curves. This solves the problem of difficulty in obtaining half-electrode parameters in scenarios with scarce data or limited experimental methods.

[0036] It should be noted that under general experimental conditions, it is easy to measure the macroscopic behavior of the entire battery, including the total voltage, total charge / discharge capacity, and the SOC-OCV curve of the entire battery. However, it is difficult to measure data such as the voltage and lithium concentration of the positive or negative electrode individually within the battery. Furthermore, high-precision electrochemical models (such as P2D models) are based on the physicochemical processes of each electrode and require microscopic parameters related to the half-electrode, creating an information gap. By utilizing readily available macroscopic test data of the entire battery, through calculation and reasoning, the key characteristic parameters of the positive and negative electrode materials within the battery under charge-discharge limits can be indirectly and accurately determined. This solves the problem of electrochemical parameter identification in low-information scenarios and helps improve model accuracy.

[0037] Step S102: The measured data is processed using a target mapping method to determine the thermodynamic parameters of the half-electrode, including the lithium intercalation boundary value of the half-electrode.

[0038] The purpose of using target mapping to process measured data to determine the thermodynamic parameters of the half-electrode (including lithium intercalation boundary values) is to resolve the contradiction between the ease of obtaining macroscopic data of the entire electrode and the difficulty in directly measuring the microscopic parameters of the half-electrode in low-information scenarios. Since the measured data may be provided by the battery manufacturer and includes different types of parameters, an appropriate mapping method for the thermodynamic parameters can be selected according to the parameter type. Open-circuit voltage data is crucial in electrochemical simulations and directly affects the final output of the model. The open-circuit voltage data of both the positive and negative electrodes are related to the lithium intercalation state of the electrode. However, conventional experimental methods often only obtain the overall state of charge and voltage response signal of the cell, and cannot obtain information about the positive or negative electrode individually. The open-circuit voltage curve of the half-electrode is directly determined by the electrode material and can be directly queried and used through public datasets; it is not experimental data that is difficult to obtain directly.

[0039] According to some embodiments provided in this application, the parameter types include the positive electrode potential corresponding to the positive electrode, the negative electrode potential corresponding to the negative electrode, and the cutoff voltage and solid-phase charge capacity of the target battery. The solid-phase charge capacity is the charge capacity corresponding to the amount of lithium atoms contained in the solid-phase region of the target battery. The above-mentioned target mapping method for determining thermodynamic parameters can include the following methods:

[0040] The first constraint relationship is determined based on the correlation between the potential difference between the positive and negative electrode potentials and the cutoff voltage;

[0041] Based on the correlation between the lithium content and solid-phase charge capacity of the negative and positive electrodes, respectively, the second constraint relationship is determined.

[0042] The third constraint relationship is determined based on the correlation between the amount of lithium inserted into the negative electrode and the amount of lithium extracted from the positive electrode;

[0043] Determine the first lithium intercalation boundary value of the negative electrode and the second lithium intercalation boundary value of the positive electrode that satisfy the first constraint relationship, the second constraint relationship, and the third constraint relationship;

[0044] Thermodynamic parameters are determined based on the first and second lithium intercalation boundary values.

[0045] According to the embodiments provided in this application, a method is provided that uses cutoff voltage and total lithium content of battery electrodes to solve for the aforementioned first, second, and third constraints through an algebraic approach. These constraints can be expressed by one or more equations. This embodiment utilizes measured data including the positive electrode potential (i.e., the open-circuit voltage of the positive electrode), the negative electrode potential (i.e., the open-circuit voltage of the negative electrode), the target battery cutoff voltage, and the solid-phase charge capacity. It employs an algebraic equation-based approach to construct multiple constraints. Based on the physical correlation that the potential difference between the positive and negative electrode potentials must match the cutoff voltage, a first constraint equation is established. A second constraint equation is constructed based on the quantitative correspondence between the lithium content of the negative and positive electrodes and the solid-phase charge capacity. A third constraint equation is established by combining the conservation of lithium intercalation in the negative electrode and lithium deintercalation in the positive electrode during battery charging and discharging. By simultaneously solving the above one or more algebraic equations, the first lithium intercalation boundary value of the negative electrode and the second lithium intercalation boundary value of the positive electrode that simultaneously satisfy all three constraints are selected. Finally, the thermodynamic parameters are determined based on these two types of lithium intercalation boundary values. The aforementioned first lithium intercalation boundary value can be the positive electrode lithium intercalation value including the predetermined upper and lower SOC limits, and the aforementioned second lithium intercalation boundary value can be the negative electrode lithium intercalation value including the predetermined upper and lower SOC limits. From a theoretical perspective, the potential difference determines the battery's external voltage characteristics, the lithium content relates to charge storage capacity, and the conservation of lithium migration maintains the continuous reaction. By directly linking macroscopic measured parameters with the microscopic thermodynamic properties of the half-electrode through algebraic equations, the key problem of not being able to directly obtain half-electrode parameters in low-information scenarios is solved, providing a parameter basis for electrochemical models that conforms to physical essence.

[0046] Through the above processing, leveraging the characteristics of solving algebraic equations, the boundary values ​​and thermodynamic parameters of half-electrode lithium intercalation can be quickly derived from measured data, reducing the data dependency threshold and computational complexity, and adapting to the practical application needs of low-information scenarios. By strictly limiting the parameter solution range through multiple constraint equations, it is ensured that the obtained first and second lithium intercalation boundary values, as well as the final thermodynamic parameters, all conform to physical principles such as lithium-ion conservation and electrochemical reaction potential laws, effectively avoiding model divergence problems caused by unreasonable parameters. The output thermodynamic parameters can serve as precise prior constraints for subsequent electrochemical model parameter identification, helping to narrow the parameter search space, improve the overall simulation accuracy and identification efficiency of the model, and provide reliable support for lithium battery state assessment and performance prediction.

[0047] For example, the known information required for the above method of determining the lithium insertion boundary by the cutoff voltage method may include: battery size information, the battery's maximum and minimum cutoff voltages, the open-circuit voltages (i.e., electrode potentials) of the positive and negative electrode materials, and the lithium content in the active particles under equilibrium conditions. In the following subscripts, n represents the negative electrode (or a), and p represents the positive electrode (or c).

[0048] Based on the first constraint, the algebraic equation can be established as follows:

[0049]

[0050]

[0051] Based on the second constraint, the algebraic equation can be established as follows:

[0052]

[0053]

[0054] Based on the third constraint, the algebraic equation can be established as follows:

[0055]

[0056]

[0057] This embodiment contains a total of 5 equations and 5 unknowns. , , , , It can be directly solved through a system of equations, based on the principle of the conservation of recyclable lithium in the positive and negative electrode active particles.

[0058] in, and These represent the total amount of neutral lithium at the initial equilibrium state of the positive and negative electrodes, respectively. (Subscript) This represents the initial state, and the unit is mol. This represents the amount of neutral lithium that can be cyclically moved during charging and discharging, expressed in mol. and These represent the total amount of neutral lithium in the positive and negative electrodes, respectively, in moles. The charge capacity (i.e., solid phase charge capacity) represents the amount of neutral lithium in the total solid phase, and is expressed in Ah. and These represent the minimum and maximum cutoff voltages of the target battery, respectively, in volts (V). Indicates positive electrode potential. This represents the negative electrode potential, measured in volts (V). and These represent the lithium intercalation stoichiometry of the positive electrode at SOC=100% and 0% states of charge (i.e., the first lithium intercalation boundary value of the positive electrode mentioned above). and These represent the lithium intercalation stoichiometry of the negative electrode at SOC=100% and 0% states (i.e., the second lithium intercalation boundary value of the aforementioned negative electrode). F represents the Faraday constant, which represents the total charge carried by 1 mole of electrons.

[0059] in, and It can be directly calculated from the electrode geometry and the maximum lithium intercalation concentration of the active particles. The calculation formula is as follows:

[0060]

[0061]

[0062] in, Indicates the size of the positive electrode plate. Indicates the size of the negative electrode plate. This indicates the volume fraction of active particles in the positive electrode. This indicates the volume fraction of active particles in the negative electrode. This indicates the maximum lithium intercalation concentration in the aforementioned cathode material. This represents the maximum lithium intercalation concentration in the aforementioned negative electrode material. The above processing method involves obtaining a definite solution through a system of equations.

[0063] According to another embodiment provided in this application, for parameters including the positive electrode potential corresponding to the positive electrode, the negative electrode potential corresponding to the negative electrode, and the constant current charge-discharge data of the target battery, the measured data is processed using a target mapping method to determine the thermodynamic parameters of the half-electrode, which may include the following steps:

[0064] Based on the charging current included in the constant current charge and discharge data, determine the change in ampere-hours of the negative electrode and the change in ampere-hours of the positive electrode;

[0065] Based on the open-circuit voltage, negative electrode potential, positive electrode potential, negative electrode ampere-hour change and positive electrode ampere-hour change included in the constant current charge-discharge data, the first lithium intercalation boundary value of the negative electrode and the second lithium intercalation boundary value of the positive electrode are determined.

[0066] The thermodynamic parameters are determined based on the first lithium intercalation boundary value of the negative electrode and the second lithium intercalation boundary value of the positive electrode.

[0067] In some embodiments provided in this application, for scenarios where the measured parameters include positive electrode potential, negative electrode potential, and constant current charge-discharge data of the target battery, a dynamic data-driven approach combined with potential correlation is used for boundary derivation. The charging current is extracted from the constant current charge-discharge data, and the changes in ampere-hours (Ah) of the negative and positive electrodes are calculated by integrating the current over time. The Ah change of the negative electrode reflects the charge change in the amount of lithium intercalated at the negative electrode, and the Ah change of the positive electrode reflects the charge change in the amount of lithium deintercalated at the positive electrode. Combined with the open-circuit voltage in the constant current charge-discharge data, and the known positive electrode potential (i.e., the open-circuit voltage of the positive electrode) and negative electrode potential (i.e., the open-circuit voltage of the negative electrode), a correlation is established between the open-circuit voltage and the potential difference between the positive and negative electrodes, and between the changes in Ah changes of the positive and negative electrodes and the amount of lithium intercalated. Through optimization methods such as curve fitting, the first lithium intercalation boundary value of the negative electrode and the second lithium intercalation boundary value of the positive electrode are derived, such as the lithium intercalation stoichiometry corresponding to the SOC extreme value. Based on these two types of lithium intercalation boundary values, the thermodynamic parameters of the half-electrode are determined. Since current integrals reflect charge changes, and charge changes correspond to fluctuations in lithium insertion / deintercalation at the positive and negative electrodes, and changes in lithium insertion directly affect the potentials of the positive and negative electrodes and the open-circuit voltage of the entire battery, this method correlates macroscopic electrical signals with the microscopic lithium insertion characteristics of the half-electrode using dynamic charge-discharge data. This solves the problem of capturing dynamic lithium insertion boundaries by relying solely on static parameters, providing a technical path consistent with electrochemical reaction laws for identifying half-electrode thermodynamic parameters in low-information scenarios with sufficient dynamic data. The above involves continuously adjusting parameters to match the model results with the set curve, which is an optimization problem.

[0068] This embodiment provides a dual-tank method for determining the lithium intercalation boundary. The physical significance of the dual-tank method lies in visualizing the lithium storage and migration process of the positive and negative electrode active particles in a lithium-ion battery through an intuitive analogy of water tanks and water volume. The active materials of the positive and negative electrodes are considered as two independent water tanks. The total volume of the tanks corresponds to the maximum capacity of the positive and negative electrode active particles to hold neutral lithium, representing the upper limit of the total amount of lithium stored in the positive and negative electrodes. The water volume in the tanks corresponds to the amount of neutral lithium actually intercalated into the positive and negative electrode active particles. The flow of water between the two tanks can be categorized as lithium delithiation at the positive electrode and lithium intercalation at the negative electrode during charging and discharging, illustrating the migration of lithium between the positive and negative electrodes. During charging and discharging, the continuous input or output of current will cause changes in charge, corresponding to an increase or decrease in the amount of neutral lithium in the positive and negative electrodes, i.e., changes in the water volume in the tanks. The change in the amount of neutral lithium will then cause changes in the potential of the positive and negative electrodes themselves, ultimately manifested as the open-circuit voltage response of the entire battery. Based on the aforementioned physical correlation, the dual-tank method can combine measured constant current charge-discharge data with the inherent potential characteristics of the positive and negative electrode materials to inversely deduce the lithium storage capacity of the positive and negative electrodes (i.e., and ), initial lithium intercalation amount (i.e. The parameters, such as the battery internal resistance (denoted as R) representing water flow resistance, transform the complex internal lithium migration and electrochemical reaction processes of the battery into a physical model that can be solved by data fitting, thereby achieving indirect characterization of the thermodynamic properties of the half-electrode that cannot be directly measured.

[0069] For example, This represents the degree of lithium ion intercalation in the negative electrode active material at a certain moment during the charging and discharging process, i.e., the real-time lithium intercalation state of the negative electrode. It represents the degree of lithium ion insertion in the positive electrode active material at a certain moment during the charging and discharging process, that is, the real-time lithium insertion state of the positive electrode.

[0070]

[0071]

[0072]

[0073] Where Q is the ampere-hour integral of the charging and discharging current, and I represents the value obtained by the charging circuit from the above constant current charging and discharging data, which is calculated as follows:

[0074]

[0075] R is the internal resistance, which can be considered as the resistance to flow between the two water tanks. By fitting the real-time response voltage during constant current charging and discharging, the final value is obtained. Parameters such as these.

[0076] In some embodiments provided in this application, the parameter types include first correlation data, second correlation data, and third correlation data. The first correlation data represents the correlation between the open-circuit voltage and the state of charge of the target battery; the second correlation data represents the correlation between the lithium intercalation content and the positive electrode potential; and the third correlation data represents the correlation between the lithium intercalation content and the negative electrode potential. Using a target mapping method to process the measured data and determine the thermodynamic parameters of the half-electrode may include the following steps:

[0077] Based on the predetermined state of charge of the target battery and the first candidate boundary value, the first lithium intercalation content of the cathode is determined, where the first candidate boundary value is the lithium intercalation boundary value of the cathode in the predetermined boundary state of charge.

[0078] Based on the predetermined state of charge and the second candidate boundary value, the second lithium intercalation content of the negative electrode is determined. The second candidate boundary value is the lithium intercalation boundary value of the negative electrode in the predetermined boundary state of charge.

[0079] Based on the second associated data, query the candidate positive electrode potential corresponding to the first lithium intercalation content; and based on the third associated data, query the candidate negative electrode potential corresponding to the second lithium intercalation content.

[0080] Based on the difference between the candidate positive electrode potential and the candidate negative electrode potential, the candidate open-circuit voltage corresponding to the predetermined state of charge is determined.

[0081] Generate fitted data relating each candidate open-circuit voltage to its corresponding predetermined state of charge;

[0082] The first candidate lithium intercalation boundary value with the smallest deviation between the fitted data and the first associated data is determined as the first lithium intercalation boundary value, and the second candidate lithium intercalation boundary value is determined as the first lithium intercalation boundary value.

[0083] In the embodiments provided in this application, based on the predetermined state of charge of the target battery, the first lithium intercalation content of the positive electrode is derived by combining the assumed first candidate boundary value of the positive electrode, and the second lithium intercalation content of the negative electrode is derived by combining the assumed second candidate boundary value of the negative electrode. The first and second candidate boundary values ​​are assumed unknowns. The candidate positive electrode potential corresponding to the first lithium intercalation content is queried based on the second correlation data, and the candidate negative electrode potential corresponding to the second lithium intercalation content is queried based on the third correlation data. The difference between the two is used to obtain the candidate open-circuit voltage corresponding to the predetermined state of charge. Fitting data between the candidate open-circuit voltage and the predetermined state of charge is generated. Finally, the candidate boundary value that minimizes the deviation between the fitted data and the first correlation data is selected, meaning the fitted data is close to the measured data. These are used as the first lithium intercalation boundary value of the positive electrode and the second lithium intercalation boundary value of the negative electrode, respectively, thereby determining the thermodynamic parameters of the half-electrode. Since the state of charge determines the lithium intercalation range of the positive and negative electrodes, and the lithium intercalation amount is related to the potential through material properties, the potential difference between the positive and negative electrodes constitutes the open-circuit voltage of the full cell. Through fitting optimization, the macroscopic open-circuit voltage characteristics are correlated with the lithium intercalation boundary of the half-electrode. The above processing method also involves continuously adjusting parameters to match the model results with the set curve.

[0084] For example, the first correlation data mentioned above can be the full cell open-circuit voltage versus state of charge (SOC-OCV) curve, and the second correlation data can be the relationship between the positive electrode open-circuit voltage and its lithium intercalation content. - The curve, the third correlation data is the relationship between the negative electrode open-circuit voltage and its lithium intercalation content. - Curve. For a given specific state of charge (SOC) (i.e., the predetermined state of charge mentioned above) and the potential curves of the positive and negative electrodes in a quasi-equilibrium state, the open-circuit voltage of the full cell can be expressed as:

[0085]

[0086] in and Consistent with the method described above, the potential curves are measured separately for the positive and negative electrodes under equilibrium conditions. The open-circuit voltage of the electrodes is related to their lithium intercalation state (SOL), while the state of charge (SOC) of the half-electrode can be expressed as:

[0087]

[0088]

[0089] in and These represent the lithium intercalation stoichiometry of the cathode at SOC=100% and 0% states, respectively (i.e., the first lithium intercalation boundary value of the cathode). and These represent the lithium intercalation stoichiometry of the negative electrode at 100% and 0% SOC, respectively (i.e., the second lithium intercalation boundary value of the negative electrode). The lithium intercalation concentration of the positive and negative electrodes at 100% and 0% SOC of the entire electrode can be obtained by fitting the open-circuit voltage of the matched full cell.

[0090] like Figure 2 As shown, using the above-mentioned full-cell open-circuit voltage matching method, the green curve represents the second correlation data, the blue curve represents the third correlation data, the fitted open-circuit voltage and state of charge are represented by the red curve, and the measured battery open circuit is represented by the black dashed line. It can be seen that the fitted data has good follow-up to the measured data.

[0091] Step S103: Based on thermodynamic parameters, generate an initial model of the target battery. The initial model contains undetermined parameters other than thermodynamic parameters.

[0092] In the embodiments provided in this application, thermodynamic parameters obtained by mapping measured data are used to set the initial model. The initial model also includes a variety of parameters, such as geometric parameters, kinetic parameters, ohmic parameters, etc., which need to be further determined. The parameters that need to be determined are the parameters to be determined.

[0093] For example, the initial model described above can be used to model a lithium-ion battery (LIB), which consists of a positive electrode (also called an anode), a negative electrode (also called a cathode), a separator, and an electrolyte. The positive and negative electrodes are typically composed of current collectors, active materials, conductive additives, and polymer binders. The separator is a porous polymer membrane that acts as an electronic insulator sandwiched between the positive and negative electrodes. The electrolyte is an electronic insulator, but it is an ion conductor, providing ion channels between the positive and negative electrodes. Depending on their composition, electrolytes can be classified as liquid, solid polymer, and solid inorganic electrolytes.

[0094] During charging, electrons released from the positive electrode flow to the negative electrode through an external connection circuit. Electrochemical oxidation and reduction reactions occur simultaneously at both the positive and negative electrodes, accompanied by the insertion and extraction of Li+ ions. Subsequently, to maintain electroneutrality, lithium ions migrate from the positive to the negative electrode via diffusion and migration through the electrolyte. This creates a potential difference between the two electrodes. Therefore, lithium-ion batteries were initially called "rocking chair batteries."

[0095] A quasi-two-dimensional (P2D) model of a lithium-ion battery was established, describing the chemical quantity variations within the battery's two phases (solid and liquid) and three regions (positive electrode, negative electrode, and separator). The electrode region is a porous medium structure composed of active particles, fillers, and electrolyte, and the influence of pores is described using average tortuosity and porosity. Figure 3 As shown, using LixC6 and LiCoO2 as electrolytes, the pseudo-two-dimensional model assumes that spherical particles of the same radius are distributed along the thickness direction of the electrode, and the particle surface is surrounded by the liquid electrolyte. The lithium concentration inside the particles is distributed along the radial direction as follows: Its concentration on the particle surface is The model equations and description are summarized as follows:

[0096] Table 1. Part of the introduction and description of the electrochemical model.

[0097]

[0098] Table 2. Another part of the introduction and description of the electrochemical model.

[0099]

[0100] The model uses many effective coefficients because the electrode structure is a porous medium, and the model treats each micro-element in the electrode as a mixed phase. Therefore, the coefficients used need to be corrected. The correction method is as follows: For diffusion activation energy, The activation energy of the reaction. The reference temperature is indicated by the superscript "ref", which represents the nominal value of the material parameter.

[0101]

[0102] ,

[0103] , .

[0104] in, For the effective conductivity of the solid phase, The nominal conductivity is for the solid phase. This represents the volume fraction of the solid phase. The effective diffusion coefficient of the solid phase is... R is the nominal diffusion coefficient of the solid phase, T is the gas constant, and T is the actual temperature. The effective reaction rate constant, The effective diffusion coefficient of the liquid phase is... The nominal diffusion coefficient of the liquid phase is... The liquid phase volume percentage in the electrode region is represented by 'brug', where 'brug' is the tortuosity correction factor. This represents the percentage of solid volume in the electrode region. The effective conductivity of the liquid phase, This refers to the nominal conductivity of the liquid phase.

[0105] In the embodiments provided in this application, low-sensitivity or unidentifiable parameters can be eliminated by sensitivity screening, reducing the dimension of the parameters to be identified, reducing the search range of the parameter space, reducing the computational resource consumption of heuristic algorithms (such as genetic algorithms) and gradient optimization algorithms, improving the speed and efficiency of parameter identification, and avoiding wasting computational power on meaningless parameters.

[0106] Low-sensitivity parameters have a negligible impact on model output, and their numerical fluctuations are difficult to accurately invert from measured data. If included in the identification process, they easily become optimization noise. This not only fails to improve model accuracy but may also cause the algorithm to get stuck in local optima or reduce identification stability due to redundant correlations between parameters. By focusing on highly sensitive parameters (such as cathode thickness, solid-phase diffusion coefficient, maximum lithium intercalation concentration, etc.) through sensitivity screening, the optimization objective becomes clearer, allowing the algorithm to concentrate on adjusting parameters that play a key role in the model output. This improves the fitting accuracy between the final parameter set and the measured data, and reduces simulation errors. For example, in the example, the optimized voltage error was reduced from 2mV to 1mV.

[0107] Parameter sensitivity analysis, as a preliminary method for hierarchical analysis of model parameters, is used to identify key parameters that significantly affect model output, guiding subsequent parameter identification processes. Parameter sensitivity analysis only needs to be performed once after the model structure is determined, and is not a necessary component of every parameter identification process. This method is used to quantitatively assess the degree to which the model output responds to the uncertainty of input parameters, thereby identifying a subset of key parameters that significantly affect model behavior. This avoids wasting computational resources on insensitive parameters. These key parameters typically include the physicochemical properties of electrode materials, electrolyte properties, etc., and directly affect core performance indicators such as temperature, open-circuit voltage, and state of charge (SOC).

[0108] According to the embodiments provided in this application, before generating an initial population based on the initial values ​​of each undetermined parameter, the method further includes:

[0109] Determine the sensitivity index corresponding to each of the multiple initial parameters included in the initial model. The sensitivity index is the degree of coupling between the corresponding initial parameter and the simulation parameters output by the initial model.

[0110] Based on the sensitivity indices corresponding to multiple initial parameters, each parameter to be determined is selected.

[0111] In the embodiments provided in this application, the core objective is to select undetermined parameters from the initial model parameters. This is achieved through a progressive process of sensitivity index calculation and parameter priority selection. For multiple initial parameters included in the initial model, a sensitivity index is quantitatively calculated for each initial parameter. This index measures the degree of coupling between the initial parameters and the simulation parameters output by the initial model. It is expressed as the magnitude and strength of the impact of changes in the initial parameter values, such as observable macroscopic response parameters like battery terminal voltage and temperature, on the simulation parameters. For example, adjustments to parameters like cathode thickness and solid-phase diffusion coefficient significantly alter the voltage simulation results, thus these parameters are considered to have high sensitivity indices. Conversely, changes in parameters like separator tortuosity and transfer coefficient have minimal impact on the simulation parameters, thus these parameters are considered to have low sensitivity indices.

[0112] After calculating the sensitivity index of all initial parameters, all initial parameters are sorted according to the magnitude of the sensitivity index. A sensitivity threshold is set as a screening criterion to remove low-sensitivity parameters or unidentifiable parameters whose sensitivity index is lower than the threshold. Since the coupling with the simulation parameters is extremely weak, the improvement effect on the accuracy of the model output is limited.

[0113] Before parameter identification, the Morris method, a global sensitivity analysis approach, can be used to further screen key parameters, define the value range or probability distribution of each input parameter, and determine the target output variable. The Morris method is used to initially screen all parameters, eliminating those with low sensitivity or those that are unidentifiable. The Sobol method is then used to perform a global sensitivity quantitative analysis on the screened key parameters, evaluating the impact of the parameters and their interactions on the model output. Based on the sensitivity index ranking, a subset of key parameters for subsequent parameter identification is determined.

[0114] In some embodiments provided in this application, determining the sensitivity index corresponding to each of the multiple initial parameters included in the initial model may include the following steps:

[0115] Within the range of the second predetermined parameters corresponding to the first parameter, the first parameter is sampled to generate the parameter trajectory of the first parameter. The parameter trajectory includes multiple sampling points of the first parameter, and the first parameter is any one of multiple initial parameters.

[0116] Perturb each sampling point and determine the initial model's perturbation response rate to each sampling point;

[0117] Based on the disturbance response rate of each sampling point, the average value and standard deviation of multiple sampling points are determined;

[0118] The sensitivity index of the first parameter is determined based on the average and standard deviation of multiple sampling points.

[0119] Using the sensitivity index obtained from the first parameter, sensitivity indices corresponding to multiple initial parameters are obtained respectively.

[0120] In some embodiments provided in this application, for any first parameter among multiple initial parameters, sampling is performed within its corresponding second predetermined parameter range to generate a parameter trajectory containing multiple sampling points, achieving comprehensive coverage of the parameter's value range. A perturbation is applied to each sampling point in the parameter trajectory, and by analyzing the output change of the initial model after the perturbation, the perturbation response rate of the model to each sampling point is calculated to reflect the strength of the parameter's influence on the model under specific values. By generating a parameter trajectory within a specific parameter range and combining it with perturbation response rate analysis, the one-sidedness of evaluating a single parameter value is avoided. Based on the perturbation response rates of all sampling points, the average and standard deviation of this set of response rates are calculated. The average reflects the overall influence level of the parameter, and the standard deviation reflects the degree of fluctuation in the parameter's influence. Combining the above average and standard deviation, a sensitivity index for the first parameter is determined and applied to other initial parameters of the model to obtain the sensitivity indices corresponding to each initial parameter.

[0121] Step S104: Generate an initial population based on the initial values ​​of each undetermined parameter, and perform the first iteration within the first predetermined parameter range, starting with the initial population, to determine the target individual that minimizes the objective function; the initial values ​​are random initialization values ​​within the first predetermined parameter range, the objective function is used to represent the difference between the simulation parameters of the initial model and the measured parameters of the target battery, and the target individual is generated based on the intermediate values ​​of each undetermined parameter;

[0122] In this embodiment, each undetermined parameter in the model is randomly initialized within a first predetermined parameter range, generating an initial value for each parameter, i.e., a set of parameter combinations, and an initial population is constructed based on these initial values. Starting from the initial population, a first iteration is performed within the first predetermined parameter range. By continuously updating the parameter combinations in the population using evolutionary algorithms such as selection, crossover, and mutation, the parameter values ​​are gradually optimized. During the iteration process, the objective function representing the difference between the initial model simulation parameters and the measured parameters of the target battery is used as the evaluation criterion. The parameter combination that minimizes the objective function value is continuously screened, and this optimal combination is the target individual, which is composed of the intermediate values ​​of each undetermined parameter after iterative optimization. Through the above processing, the optimization method with population iteration as the core is more likely to escape local optima than single-point search. Combined with the constraint of the first predetermined parameter range, it can quickly converge to a globally better solution within a reasonable parameter space, providing an optimized parameter combination scheme for battery modeling. The obtained target individual enables the model output to be closest to the actual battery characteristics.

[0123] According to the embodiments provided in this application, a heuristic global optimization algorithm under multiple constraints, such as simulated annealing, particle swarm optimization, and genetic algorithms, is used to perform a global search in the parameter space to achieve efficient identification of the parameter set in a high-dimensional space. By setting appropriate iteration numbers, objective function convergence criteria, and population update strategies, the minimum value of the objective function is gradually approximated. The following description uses a genetic algorithm. The first iteration, starting with an initial population within a first predetermined parameter range, may include the following steps:

[0124] Within a first predetermined parameter, determine the iteration values ​​of each individual in the current iteration population, which is obtained based on the initial population iteration. Extract a predetermined number of candidate individuals from the current iteration population and determine the fitness function value corresponding to each candidate individual. The fitness function value represents the degree to which the corresponding candidate individual matches the objective function's iteration toward the minimum value. Determine the parent individual with the largest fitness function value among the candidate individuals. Perform a predetermined recombination operation on each parent individual to obtain offspring individuals. The predetermined recombination operation includes at least crossover and mutation operations. The crossover operation is used to interpolate the iteration values ​​included in each parent individual, and the mutation operation is used to perturb the iteration values ​​included in each parent individual. Perform the first iteration based on each offspring individual to obtain the next iteration population.

[0125] In the embodiments provided in this application, within a first predetermined parameter range, the iterative value of each individual in the current iterative population is determined to be the current parameter combination. The current population is obtained from the initial population through previous iterations. A predetermined number of candidate individuals are extracted from the current population, and the fitness function value of each candidate individual is calculated. The larger the value, the more it conforms to the direction of the objective function iterative towards the minimum value. The individual with the largest fitness function value is selected from the candidate individuals as the parent individual. Selecting high-quality parent individuals based on the fitness function ensures that the iteration direction always points to the minimum value of the objective function, avoiding blindness in the optimization process. Optionally, a predetermined recombination operation can be performed on the parent individuals to generate offspring individuals. Alternatively, a crossover operation can be used to interpolate and fuse new parameter combinations by the iterative values ​​of the parent individuals. An optional mutation operation can be used to introduce randomness by perturbing the iterative values, introducing new parameter combinations. Based on the generated offspring individuals, the next round of iteration is performed to form a new iterative population. The above process is repeated until the iteration termination condition is met. Through this iterative mechanism simulating biological evolution, the parameter combination is gradually optimized to approach the minimum value of the objective function.

[0126] According to the embodiments provided in this application, a penalty function is used to control the iteration direction in the first iteration process. The penalty function is used to set the fitness function value of the individual to a predetermined minimum value when the objective function has not converged.

[0127] In the embodiments provided in this application, during the first iteration, for each individual in the population, the convergence of its corresponding objective function is evaluated. If the objective function fails to converge, it indicates that the parameter combination corresponding to that individual deviates from the optimal direction, and a penalty function is triggered, forcibly setting the fitness function value of that individual to a predetermined minimum value, such as zero, so that it is eliminated in subsequent parent selection. If the objective function has converged, its fitness function value is calculated normally, and it is retained to participate in the recombination iteration. The penalty function excludes inferior individuals from participating in subsequent iterations, preventing interference with the iteration path of superior parameter combinations, strengthening the screening pressure on superior individuals, accelerating the convergence speed, and constraining the population evolution direction through explicit penalty rules, ensuring that the iteration always revolves around the goal of reducing the difference between model simulation and actual measurement.

[0128] For example, the model's operating conditions are set to match the measured current, and the output is interpolated to match the measured time series. Its objective function is...

[0129]

[0130] in, For time The measured voltage at that time For the set of model parameters, The measured process current is the same as the model input operating condition. This represents the current parameter set as Input operating conditions are At that time, the time was The simulated voltage. Upper and lower bounds have been set for each parameter to be identified, but some parameter combinations in the current parameter space may still cause the model calculation to fail to converge. These combined parameters are clearly not the optimal set of parameters sought; therefore, f(χ) for the non-convergence scenario can be directly set to a very large penalty value, such as f(χ) = 10. 6 .

[0131] Taking genetic algorithms as an example, such as Figure 4 The algorithm can employ methods such as tournament selection or sorting selection to choose superior individuals from the current population based on their fitness to serve as parents for the next generation. Crossover operations are performed among the selected parents, using methods such as single-point crossover, double-point crossover, or uniform crossover to recombine parameter vectors and generate new individuals. Certain parameter positions of the newly generated individuals are perturbed with a certain probability to ensure population diversity and avoid getting trapped in local optima. The mutation operation must ensure that the mutated parameters remain within a predetermined physically feasible range. An upper limit on the number of iterations or a convergence criterion for the fitness function is set; for example, the algorithm is considered convergent and the iteration terminates when the change in the optimal fitness is less than a certain threshold for several consecutive generations. The individuals with the best fitness in the final population are selected as the initial optimal parameter set and used as input to the gradient optimization algorithm.

[0132] To avoid getting trapped in local optima and struggling to find the global optimal solution in a parameter space with limited information, high dimensions, complexity, and multiple local optima, we simulate natural selection and genetic mechanisms, make full use of the identified thermodynamic parameters, refer to the sampled data from the experimental battery, and use a combination of heuristic algorithms such as genetic algorithms and gradient optimization algorithms to efficiently identify and correct the operating parameters, effectively improving the globality and robustness of parameter identification.

[0133] After completing the initial global search and obtaining the optimal individual (i.e., the parameter set with the best performance), further local fine-tuning is performed in the vicinity of this individual. Specifically, local optimization algorithms such as quasi-Newton methods (e.g., BFGS) are used, with the optimal parameter set obtained from the global search as the initial values, to further explore local optimal solutions in the parameter space. The resulting parameter set improves the accuracy and stability of model parameter identification while taking into account both global optimality and local convergence.

[0134] Step S105: Based on the second derivative information of the objective function, determine the iteration direction of the objective function, which is the direction pointing to the minimum value of the objective function;

[0135] In this application, by utilizing the second derivative information of the objective function to determine the iteration direction, the curvature characteristics of the function can be captured more accurately, enabling the iteration process to converge to the minimum more efficiently. The second derivative reflects the changing trend of the function's rate of change. The second derivative of the objective function at the current iteration point is calculated to obtain the curvature characteristics of the function at that point. The second derivative can be in matrix form, i.e., the Hessian matrix. Combining the analysis of the second derivative with the local shape of the function, if the second derivative is positive, it indicates that the function is convex at that point, and the minimum value is located in the gradient descent direction at the current point. If the second derivative is negative, it indicates that the function is concave, and the direction needs to be adjusted to avoid moving away from the minimum value. By comprehensively considering the curvature information reflected by the second derivative, the iteration direction pointing towards the minimum value of the objective function is determined, which is used to further approximate the optimal solution.

[0136] Step S106: Starting with the current gradient value of the objective function, perform the second iteration based on the iteration direction to determine the target value of each undetermined parameter when the objective function satisfies the predetermined termination condition. The current gradient value is the gradient value of the objective function changing towards the minimum value when each undetermined parameter is an intermediate value.

[0137] With all undetermined parameters taking intermediate values, the current gradient value of the objective function is calculated, i.e., the direction and magnitude of the gradient of the function towards its minimum at that point. Starting from this current gradient value, the second iteration is performed along a pre-determined iterative direction pointing towards the minimum of the objective function. In each iteration, the values ​​of the undetermined parameters are adjusted based on the gradient information. The value of the objective function is continuously determined during the iteration process. Iteration stops when the function value satisfies the condition, and the values ​​of the undetermined parameters at this point are the target values. The predetermined termination conditions include reaching a preset precision, an upper limit for the number of iterations, or a sufficiently small gradient value. Through this process, the parameter combination that brings the objective function to its optimal state is determined by the synergistic guidance of gradient information and iterative direction.

[0138] Step S107: Set the initial model based on the target values ​​of each undetermined parameter to obtain the target model.

[0139] By setting the initial model based on the target values ​​of each undetermined parameter to obtain the target model, the results of previous parameter optimization can be directly applied to model construction, giving the target model better performance and practicality. The target values ​​of the undetermined parameters are the optimal solutions obtained through multiple rounds of iterative optimization. Using these as the basis for configuring the initial model solves problems such as low simulation accuracy and poor adaptability caused by unreasonable parameters in the initial model, thereby improving the reliability and accuracy of the obtained target model.

[0140] Based on the above embodiments, this application also provides an optional implementation method, such as... Figure 4The following details the parameter identification of a 6.7Ah battery cell with NMC811 as the positive electrode, Graphite as the negative electrode, LiPF6 in EC:EMC (3:7) electrolyte, and a specific capacity. The specific capacity density of this cell can be calculated from the theoretical specific capacity density. The concentration is 0.75 mol, and the cutoff voltages of this battery are 2.5V and 4.2V, respectively. Open-circuit voltage data for the battery materials used in this battery can be found in publicly available datasets. For example... Figure 2 The diagram illustrates the connection between NMC811 (positive electrode, blue curve), Graphite (negative electrode, green curve), and the open-circuit voltage of the full cell (red curve). , The diagram illustrates the positive and negative open-circuit voltages at different states of charge (SOC).

[0141] As shown in Table 3 and Figure 6 As shown, different types of operating condition parameters, such as geometric parameters, kinetic parameters, and ohmic parameters in the electrochemical model, are considered as the initial parameter set. After sensitivity analysis, the parameters are sorted in descending order and divided into high-sensitivity, medium-sensitivity, and low-sensitivity parameters for illustration.

[0142] Table 3. Parameter descriptions and sensitivity classifications of NMC811 / graphite batteries

[0143]

[0144] The identification results of the three methods for mapping thermodynamic boundary parameters from full cells to half cells are compared in the table below:

[0145] Table 4 Comparison of results from three methods for mapping thermodynamic parameters from full cells to half cells

[0146]

[0147] Figure 2 and Figure 7 The curves after optimization using the full-cell open-circuit voltage matching method and the dual-tank method are shown respectively. Figure 7 The black dashed curve represents the simulation result (corresponding to the simulation model), and the red solid line represents the measured battery response data (corresponding to the experimental data). Among the three methods, assuming known information and disregarding errors, Method 1, by directly solving the system of equations, can be considered as nominal parameters. Since the dual-tank model uses data from discharging the battery at 100% SOC, the dual-tank method... and That means and As shown in Table 4, the three methods are applicable to scenarios with different known information, and all can accurately reflect the true thermodynamic boundary of the battery cell. Furthermore, the identification result of the cutoff voltage method is used as the preset parameter for subsequent parameter identification.

[0148] This implementation example demonstrates a case tested at a constant ambient temperature of 298.15 K. Therefore, it does not involve the identification process of parameters such as activation energy and entropy-thermal coefficient that change due to temperature variations. Furthermore, each time a new individual or parameter set is generated, the volume percentage of the solid phase region needs to be manually adjusted to satisfy material conservation at the electrode region.

[0149]

[0150] in This refers to the volume percentage of conductive adhesives and other fillers.

[0151] During parameter identification, the constructed objective function is used to measure the deviation between the model output and the experimentally measured data. To simplify the method and facilitate the explanation of the implementation process, the objective function in this embodiment is calculated only based on the voltage response error of the constant current-constant voltage (CCCV) experiment, and other variables such as temperature are not included in the objective function. A total of 5513 experimental data points were collected at a frequency of 1 Hz, including input current, terminal voltage, and acquisition time.

[0152] The P2D model has many parameters. Based on the results of the previous sensitivity analysis and some measured results, a set of parameters to be identified is selected.

[0153] Table 5. Information on the known parameter set

[0154]

[0155] Table 6 Information on parameters to be identified

[0156]

[0157] Tables 5 and 6 above list the complete parameter sets, and based on the parameter sensitivity analysis results and known information, the known parameter set and the undetermined parameter set are divided. For the known parameter set, the specific values ​​and acquisition methods are explained. For the undetermined parameter set, this application embodiment selects 10 highly sensitive parameters that are extremely difficult to obtain experimentally for identification, and provides reasonable ranges for the undetermined parameters based on their intrinsic material properties.

[0158] This section uses genetic algorithms as a representative of heuristic search to detail the specific operation process in parameter identification.

[0159] Parameter encoding and population initialization: The electrochemical model parameters to be identified (such as diffusion coefficient, porosity, reaction rate constant, etc.) are generated based on a pre-defined parameter generation strategy. Within the given parameter value range Generate a set of parameter vectors according to the rules. This is then encapsulated into the individual structure of a genetic algorithm. They represent parameters respectively The upper and lower boundaries of the values, in this embodiment, the function The parameters are generated uniformly and randomly, where i represents the i-th parameter. The selectable values ​​for the parameter to be identified. Uniformly distributed across the interval.

[0160]

[0161] Call the repeated generation operation Where M is the population size, i.e., the total number of individuals. Each The initial population is generated through the individual initialization process, providing a search starting point for subsequent genetic operations (selection, crossover, mutation). 'm' is used to identify different individuals. In this embodiment, M individuals are repeatedly generated based on the aforementioned individual initialization method to form the initial population.

[0162] In this embodiment, the objective function is the weighted mean square error (WMSE) of the voltage response under constant current-constant voltage (CCCV) experimental conditions, which is used to measure the error between the model simulation output and the experimental data.

[0163] For individuals that do not meet the electrochemical principles or cause the simulation model to fail to converge, a penalty function method is used to reduce their fitness. A tournament selection method is employed. In each selection process, k individuals are randomly selected from the current population (k=3 in this example), their fitness values ​​are compared, and the individual with the highest fitness is selected as one of the parent individuals. This process is repeated until the required set of parent individuals is generated. This method can maintain a high selection probability for superior individuals while preserving a certain degree of population diversity, thereby avoiding premature convergence. The Tournament Set is the tournament candidate set, which is a subset consisting of k randomly selected individuals from the entire population. Represents an individual The ranking of Tournament Set members based on their fitness values. Generally, the higher the fitness value, the higher the ranking. Representative of individuals The probability of being selected as a parent individual is positively correlated with the individual's ranking in the candidate set.

[0164]

[0165] A blend crossover (BLX-α) operation is performed between selected parent individuals to generate offspring individuals through linear interpolation of the corresponding parameter components. The m1th and m2th individuals are selected as parent individuals for crossover. The parameter component i within each individual is denoted as... , The parameter values ​​of its crossover offspring are in the range Randomly generated internally.

[0166]

[0167] in , As an expansion coefficient, α=0.5 is used in this embodiment. This method can increase the diversity of the search space and improve the global optimization capability while preserving the parent features.

[0168] For each parameter in an individual With probability (i.e., mutation probability) Perform Gaussian perturbation, The result after perturbation:

[0169]

[0170] in The mean of a Gaussian distribution is given. This represents the standard deviation. This step generates a local random search in the parameter space to help the algorithm escape local optima.

[0171] After mutation, if Exceeding the preset physical boundary range Then, a truncation method is used for correction:

[0172]

[0173] This correction mechanism ensures that all generated individuals satisfy the physical rationality and constraints of the model in subsequent evolution.

[0174] After the g-th iteration, record the optimal fitness value for that iteration. The algorithm terminates its iteration when any of the following conditions are met:

[0175] Iteration upper limit condition That is, to reach the preset maximum algebra Or, according to the convergence and stability condition:

[0176]

[0177] in, The relative convergence threshold (e.g., 1e-4) is given, and k is the continuous stable algebra. This condition sets a minimum number of consecutive stable algebras (e.g., 5 generations). This condition guarantees that the algorithm is considered convergent when the variation in the optimal value is sufficiently small and has persisted for several generations.

[0178] When the convergence condition is met, the individual with the best fitness is selected from the final population. This serves as the initial optimal parameter solution for the electrochemical model, and then proceeds to the local gradient optimization stage.

[0179] The initial optimal parameter vector obtained by the genetic algorithm

[0180]

[0181] Based on this, a quasi-Newton type L-BFGS-B algorithm is used for local gradient optimization to further approximate the global optimum. The optimization process is as follows:

[0182] The objective function and constraints are the same as in the genetic algorithm. In this embodiment, the optimization algorithm is set to: maximum number of iterations. =1000, convergence threshold (change in objective function) ftol=1e-12, gradient difference step size eps=1e 8. The iterative update formula using the L-BFGS-B algorithm:

[0183]

[0184] in, This is the parameter vector at the k-th iteration. This represents the residual of the objective function when the parameter vector is substituted. For a finite-memory quasi-Newton approximation of the inverse Hessian matrix, The step size factor is determined by line search.

[0185] The iteration terminates when any of the following conditions are met:

[0186]

[0187] in ftol represents the maximum number of iterations, and ftol represents the maximum tolerance error.

[0188] The final output optimal parameter vector serves as the accurate identification result of the electrochemical model, and its specific values ​​are listed in Table 6. Substituting the preliminary and final optimal parameter vectors into the model, the mean squared errors (MSEs) calculated for CCCV experimental data are 2 mV and 1 mV, respectively. The identification results are as follows: Figure 8As shown. Further, based on the final optimal parameter vector, the model was validated under dynamic operating conditions. The red curve represents the measured response data of the battery, the black curve represents the simulation results after population optimization, and the green curve represents the final simulation results obtained after gradient optimization. Results under different operating conditions are shown below. Figure 9 The black and red curves represent the measured and simulated battery response voltages, respectively, indicating that the identified parameters have good robustness and applicability under different operating conditions.

[0189] Through the above processing, costly or destructive specialized experiments are eliminated, utilizing only conventional charge-discharge data fragments and macroscopic electrical signal data from the battery. This reduces the experimental burden and is suitable for low-information conditions in practical engineering applications. Global sensitivity analysis filters a subset of key parameters, eliminating insensitive parameters, reducing noise and disturbances during optimization, effectively preventing the algorithm from getting trapped in local optima, and improving the stability and accuracy of the identification results. The optimization space dimension is reduced, lowering computational resource consumption while improving the accuracy and robustness of parameter identification. Three mapping strategies—the cutoff voltage method, the dual-tank algorithm, and the full-cell OCV matching method—are integrated, allowing for flexible selection based on actual data characteristics, expanding the applicability and practicality of the method. Combining heuristic global search and gradient optimization methods such as genetic algorithms achieves efficient and stable parameter optimization, enhancing the algorithm's convergence and result accuracy.

[0190] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0191] Figure 10 This is a block diagram of an electronic device used to implement embodiments of this application. Figure 10 As shown, the electronic device includes a memory 1001 and a processor 1002. The memory 1001 stores a computer program that can run on the processor 1002. When the processor 1002 executes the computer program, it implements the method described in the above embodiments. The number of memories 1001 and processors 1002 can be one or more. In a specific implementation, the electronic device may also include a communication interface 1003 for communicating with external devices and performing data exchange and transmission.

[0192] In practical implementation, if the memory 1001, processor 1002, and communication interface 1003 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 10 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0193] Optionally, in a specific implementation, if the memory 1001, processor 1002 and communication interface 1003 are integrated on a single chip, the memory 1001, processor 1002 and communication interface 1003 can communicate with each other through an internal interface.

[0194] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this application.

[0195] This application provides a computer program product, including a computer program that, when executed by a processor, implements the method provided in this application.

[0196] This application also provides a chip including a processor for calling and executing instructions stored in a memory, causing a communication device with the chip installed to perform the method provided in this application.

[0197] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.

[0198] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0199] Further, optionally, the aforementioned memory may include read-only memory and random access memory. The memory may be volatile memory or non-volatile memory, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Sync Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0200] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0201] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0202] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0203] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. Furthermore, the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functionality involved.

[0204] The logic and / or steps described in the flowchart or otherwise herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0205] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. All or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware, the program being stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiments.

[0206] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.

[0207] The above are merely exemplary embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope described in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of generating a battery emulation model, the method comprising: The method comprises the following steps: determining a target mapping mode according to a parameter type of a parameter included in measured data, the measured data being data measured on a target battery, the target mapping mode being a mapping mode for mapping the measured data from a full electrode to a half electrode, the full electrode being an electrode state including a positive electrode and a negative electrode, and the half electrode being an electrode state including the positive electrode or the negative electrode; processing the measured data by using the target mapping mode to determine a thermodynamic parameter of the half electrode, the thermodynamic parameter including a lithium intercalation boundary value of the half electrode, the lithium intercalation boundary value being a lithium intercalation stoichiometric ratio corresponding to a state of charge (SOC) extreme value of the battery; generating an initial model of the target battery based on the thermodynamic parameter, the initial model including each to-be-determined parameter except the thermodynamic parameter; generating an initial population according to initial values of the each to-be-determined parameter, performing a first iteration process on the initial population in a first predetermined parameter range to determine a target individual that minimizes a target function, the initial values being random initial values in the first predetermined parameter range, the target function being used to represent a difference between simulation parameters of the initial model and measured parameters of the target battery, and the target individual being generated based on intermediate values of the each to-be-determined parameter; determining an iteration direction of the target function based on second derivative information of the target function, the iteration direction being a direction pointing to a minimum value of the target function; performing a second iteration process based on the iteration direction, starting from a current gradient value of the target function, to determine target values of the each to-be-determined parameter when the target function satisfies a predetermined termination condition, the current gradient value being a gradient value of the target function changing towards the minimum value when the each to-be-determined parameter is an intermediate value; setting the initial model based on the target values of the each to-be-determined parameter to obtain a target model.

2. The method of claim 1, wherein, Before the initial population is generated according to the initial values of the each to-be-determined parameter, the method further comprises: determining sensitivity indicators corresponding to a plurality of initial parameters included in the initial model, the sensitivity indicators being coupling degrees between the corresponding initial parameters and the simulation parameters output by the initial model; determining the each to-be-determined parameter based on the sensitivity indicators corresponding to the plurality of initial parameters.

3. The method of claim 2, wherein, The determination of the sensitivity indicators corresponding to the plurality of initial parameters included in the initial model comprises: sampling a first parameter in a second predetermined parameter range corresponding to the first parameter to generate a parameter trajectory of the first parameter, the parameter trajectory including a plurality of sampling points of the first parameter, and the first parameter being any parameter in the plurality of initial parameters; determining a disturbance response rate of the initial model to each of the sampling points by perturbing the each of the sampling points; determining a mean value and a standard deviation of the plurality of sampling points based on the disturbance response rate of the each of the sampling points; determining a sensitivity indicator of the first parameter according to the mean value and the standard deviation of the plurality of sampling points. The sensitivity index of the first parameter is obtained, and a plurality of sensitivity indexes corresponding to the initial parameters are obtained.

4. The method of claim 1, wherein, The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; 5. The method of claim 1, wherein, The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; 6. The method of claim 1, wherein, The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance amount of a lithium atom contained in a solid-phase region of the target battery; The parameter types include a positive electrode potential corresponding to the positive electrode, a negative electrode potential corresponding to the negative electrode, and a cutoff voltage and a solid-phase charge capacity of the target battery, the solid-phase charge capacity being a charge capacity corresponding to a substance determining a candidate open circuit voltage corresponding to the predetermined state of charge based on a difference between the candidate positive electrode potential and the candidate negative electrode potential; generating fitting data associated between each of the candidate open circuit voltages and the corresponding predetermined state of charge; selecting a candidate boundary value from the first candidate boundary value and the second candidate boundary value, which makes the fitting data closest to the first associated data, as a first lithium intercalation boundary value of the positive electrode and a second lithium intercalation boundary value of the negative electrode, respectively, to determine the thermodynamic parameter of the half electrode.

7. The method according to any one of claims 1 to 6, characterized in that, The first iteration process based on the initial population as the starting point within the first predetermined parameter range comprises: determining an iteration value of each individual included in a current iteration population based on the initial population, within the first predetermined parameter; extracting a predetermined number of candidate individuals from the current iteration population, and determining a fitness function value corresponding to each of the candidate individuals, the fitness function value representing a degree of compliance of the corresponding candidate individual with the iteration of the target function towards a minimum value; determining a parent individual with a maximum fitness function value from the candidate individuals; performing a predetermined recombination operation on each of the parent individuals to obtain a child individual, the predetermined recombination operation at least including a crossover operation and a mutation operation, the crossover operation being used for interpolation processing of the iteration value included in each of the parent individuals, and the mutation operation being used for perturbation processing of the iteration value included in each of the parent individuals; performing a first iteration process based on each of the child individuals to obtain a next iteration population.

8. The method according to any one of claims 1 to 6, characterized in that, A penalty function is used to control an iteration direction in the first iteration process, the penalty function being used to set the fitness function value of an individual to a predetermined minimum value in the case that the target function does not converge.

9. An electronic device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 8. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 8.

Images