Lithium battery state of health estimation method under shaking condition by fusing physical information technology

Abstract

The present application relates to the technical field of lithium battery state of health estimation, and particularly relates to a lithium battery state of health estimation method fusing physical information technology under a shaking working condition, comprising the following steps: S1: determining main aging characteristics of the shaking working condition; S2: establishing an aging degradation model based on the main aging characteristics of the shaking working condition; S3: establishing a pseudo two-dimensional electrochemical model; S4: coupling the aging degradation model and the pseudo two-dimensional electrochemical model to obtain a lithium battery coupling degradation model; S5: extracting physical characteristics related to the lithium battery state of health based on the lithium battery coupling degradation model, and constructing a multi-step physical characteristic sequence; S6: training a state of health estimation model based on the multi-step physical characteristic sequence and corresponding battery state of health true labels; and S7: inputting the multi-step physical characteristic sequence of a lithium battery to be predicted into the trained state of health estimation model to output corresponding lithium battery state of health estimation values. The present application can improve the physical consistency and interpretability of lithium battery state of health estimation.

Description

Technical Field

[0001] This invention relates to the field of lithium battery health status estimation technology, specifically to a lithium battery health status estimation method that integrates physical information technology under shaking conditions. Background Technology

[0002] With the development of ship electrification and low-carbon marine transportation, lithium batteries, due to their advantages such as high energy density, long cycle life, fast power response, and good environmental adaptability, are widely used in electric ships, hybrid-powered ships, ferries, tugboats, offshore platforms, and ship auxiliary power systems. As a crucial component of ship propulsion and energy storage systems, the health status of lithium batteries directly affects the safety, reliability, and economy of the ship's propulsion system. Therefore, accurately estimating the health status of lithium batteries under ship operating conditions is of great significance.

[0003] Compared to land-based vehicles or stationary energy storage systems, ships operate in a more complex environment. During navigation, berthing, dynamic positioning, and maritime operations, ships are affected by factors such as wind, waves, and currents, resulting in multi-degree-of-freedom motions including rolling, pitching, bow rolling, swaying, heave, and heave. These ship movements subject onboard lithium batteries to a prolonged low-frequency swaying environment. The mechanical stress generated by this swaying alters the electrolyte distribution within the lithium battery and may cause electrode particle cracking, active material shedding, interfacial film rupture and regeneration, changes in pore structure, deterioration of conductive paths, and enhanced interfacial side reactions, thereby accelerating lithium battery capacity decay and impedance changes.

[0004] Currently, methods for estimating the health status of lithium batteries mainly include those based on capacity testing, open-circuit voltage, incremental capacity curves, equivalent circuit models, electrochemical models, and data-driven approaches. Capacity testing and open-circuit voltage testing typically require a complete charge-discharge process or a long resting time, making them unsuitable for online estimation in continuous ship operation scenarios. Incremental capacity analysis can reflect some aging characteristics, but it has high requirements for test conditions and data quality and is easily affected by noise, rate variations, and complex operating conditions. Equivalent circuit models have high computational efficiency, but their parameters often lack clear granular degradation physical meanings, making it difficult to explain the internal aging process of lithium batteries under shaking conditions. Electrochemical models can describe lithium-ion diffusion, potential distribution, and interfacial reaction processes, offering good physical interpretability. However, when electrochemical models are further coupled with aging mechanisms such as interfacial film growth, lithium deposition or stripping, and electrode particle cracking, the number of model equations increases, leading to higher solution complexity and making them unsuitable for direct application in practical ship battery management systems.

[0005] In recent years, data-driven methods such as neural networks have been widely used for lithium battery health status estimation. These methods can establish a nonlinear mapping relationship between operating data and health status, but they usually rely on a large number of training samples and lack clear physical constraints. Under complex operating conditions such as shaking, rate changes and aging stage changes, they are prone to problems such as insufficient generalization ability and uninterpretable estimation results.

[0006] Therefore, how to propose a health status estimation method that can simultaneously consider the effects of shaking mechanical stress, the internal electrochemical reaction process of the battery, and multiple aging and degradation mechanisms is a technical problem that urgently needs to be solved. Summary of the Invention

[0007] To address the shortcomings of the existing technologies, the technical problem this invention aims to solve is: how to provide a lithium battery health status estimation method that integrates physical information technology under shaking conditions. This method determines battery aging characteristics through cyclic aging test data under shaking conditions, describes the internal lithium-ion diffusion, interfacial side reactions, and particle structure damage processes through a lithium battery coupled degradation model, and outputs the lithium battery health status through a health status estimation model that integrates the physical constraints of lithium-ion diffusion. This approach reduces the dependence of purely data-driven models on a large number of samples and improves the physical consistency and interpretability of lithium battery health status estimation under shaking conditions.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0009] A method for estimating the health status of lithium batteries under shaking conditions by integrating physical information technology includes:

[0010] S1: Conduct cyclic aging tests on lithium batteries under shaking conditions to determine the main aging characteristics under shaking conditions.

[0011] S2: Based on the main aging characteristics of shaking conditions, an aging degradation model is established to describe the growth of the internal interface film, lithium deposition or stripping, and the cracking process of electrode particles in lithium batteries.

[0012] S3: Establish a pseudo-two-dimensional electrochemical model to describe the solid-phase lithium-ion diffusion, electrolyte lithium-ion transport, solid-phase potential distribution, liquid-phase potential distribution, and electrode or electrolyte interface reaction processes inside a lithium battery.

[0013] S4: Couple the aging degradation model and the pseudo-two-dimensional electrochemical model to obtain the lithium battery coupled degradation model;

[0014] S5: Extract physical features related to the health status of lithium batteries based on the lithium battery coupling degradation model, and construct a multi-step physical feature sequence as training samples;

[0015] S6: Based on the multi-step physical feature sequence as training samples and their corresponding real labels of battery health status, a health status estimation model is trained in combination with the physical constraints of lithium-ion diffusion.

[0016] S7: Construct a multi-step physical feature sequence of the lithium battery to be predicted based on the operating data of the lithium battery during the application process, and input it into the trained health status estimation model to output the corresponding lithium battery health status estimate.

[0017] Preferably, step S1 specifically includes the following processing steps:

[0018] S101: Construct a test environment including a shaking test platform, a battery charge and discharge test device and an electrochemical impedance test device, and conduct a cycle aging test on the lithium battery under preset shaking parameters to obtain cycle aging test data.

[0019] S102: Electrochemical impedance testing and electrode material characterization were performed on lithium batteries at different aging stages to obtain electrochemical impedance test results and electrode material characterization results.

[0020] S103: Based on cycle aging test data, electrochemical impedance test results, and electrode material characterization results, combined with capacity decay law, incremental capacity curve changes, and impedance evolution law, the main aging characteristics of lithium batteries under shaking mechanical stress are identified. The main aging characteristics include conductivity loss, lithium inventory loss, active material loss, interface film growth, electrode particle cracking, by-reaction product deposition, and mass transfer capability.

[0021] Preferably, in step S2, the aging degradation model includes an interface film growth model, a lithium deposition or stripping model, and an electrode particle cracking model:

[0022] (1) Interfacial film growth model

[0023] The change in interfacial film thickness is expressed as:

[0024] ;

[0025] In the formula: Indicates the thickness of the interfacial film; Indicates solvent molecule flux; Indicates the molar volume of the interfacial membrane; Indicates solvent molecule concentration; Indicates the solvent molecule diffusion coefficient;

[0026] The solvent molecule flux satisfies: , Indicates the temperature or thermodynamic temperature of the lithium battery;

[0027] The solvent molecule diffusion coefficient satisfies: ;

[0028] The boundary conditions are: ; ;

[0029] In the formula: Indicates the inherent diffusion coefficient of the electrolyte; Indicates the integral number of the electrolyzed liquid; Indicates the initial solvent molecule concentration; Indicates the coordinates of the interface film thickness direction;

[0030] The porosity change caused by interfacial film growth is as follows:

[0031] ;

[0032] In the formula: Indicates electrode porosity; This represents the specific surface area of ​​the negative electrode;

[0033] The overpotential of the interface film caused by the growth of the interface film is:

[0034] ;

[0035] In the formula: Indicates the overpotential of the interface film; This represents the interfacial reaction current density; Indicates the ohmic resistivity of the interface film; Indicates the interfacial membrane impedance;

[0036] (2) Lithium deposition or stripping model

[0037] The change in lithium-ion concentration in the negative electrode solid phase is expressed as:

[0038] ;

[0039] In the formula: This indicates the lithium-ion concentration in the negative electrode solid phase; Indicates the lithium-ion stripping flux of the negative electrode; This represents the dead lithium decay rate related to the interface film thickness;

[0040] The dead lithium decay rate is expressed as:

[0041] ;

[0042] In the formula: This represents the initial dead lithium decay parameter; Indicates the initial interface film thickness;

[0043] Lithium inventory losses caused by lithium deposition or stripping processes are expressed as follows:

[0044] ;

[0045] In the formula: This indicates lithium inventory loss; This indicates the initial maximum solid-phase lithium-ion concentration at the negative electrode; This indicates the current maximum solid-phase lithium-ion concentration at the negative electrode;

[0046] The lithium deposition overpotential is expressed as:

[0047] ;

[0048] In the formula: Indicates the overpotential of lithium deposition; Indicates the negative electrode solid-phase potential; Indicates the potential of the electrolyte phase;

[0049] (3) Electrode particle cracking model

[0050] The hydrostatic stress of the electrode particles is expressed as:

[0051] ;

[0052] In the formula: Indicates hydrostatic stress; Indicates radial stress; Indicates tangential stress;

[0053] The change in particle crack length is represented as:

[0054] ;

[0055] In the formula: Indicates the crack length; This parameter represents the particle cracking rate. Indicates the stress intensity factor; Indicates the particle cracking constant;

[0056] The change in specific surface area of ​​the negative electrode caused by particle cracking is expressed as:

[0057] ;

[0058] In the formula: Indicates the number of cracks; Indicates the crack width; Indicates the initial time parameter;

[0059] The change in particle radius is expressed as:

[0060] ;

[0061] In the formula: Indicates the molar volume of the negative electrode material; Indicates particle radius; Indicates the average lithium-ion concentration;

[0062] The change in interfacial film thickness at the crack is represented as follows:

[0063] ;

[0064] In the formula: Indicates the thickness of the interfacial film at the crack;

[0065] The change in volume fraction of the negative electrode active material is expressed as:

[0066] ;

[0067] In the formula: Indicates the volume fraction decay rate; Indicates the decay index; Indicates the yield strength of the negative electrode material;

[0068] The loss of active material is expressed as:

[0069] ;

[0070] In the formula: This indicates the loss of active materials; This indicates the initial volume fraction of the active material in the negative electrode. This indicates the current volume fraction of the active material in the negative electrode.

[0071] Preferably, in step S3, the pseudo-two-dimensional electrochemical model includes a solid-phase lithium-ion diffusion model, an electrolyte lithium-ion transport model, a solid-phase potential model, a liquid-phase potential model, and an interfacial reaction kinetic model.

[0072] (1) Solid-phase lithium-ion diffusion model

[0073] The solid-phase lithium-ion diffusion process is represented as follows:

[0074] ;

[0075] In the formula: These represent the negative and positive electrodes, respectively. Indicates the lithium-ion concentration in the electrode solid phase; Indicates the effective diffusion coefficient of the electrode solid phase; Indicates the radial coordinates of the electrode particles; Indicates time;

[0076] (2) Electrolyte lithium-ion transport model

[0077] In the thickness direction of the battery, the lithium-ion transport process in the electrolyte can be represented as follows:

[0078] ;

[0079] In the formula: Indicates the integral number of the electrolyzed liquid; Indicates the lithium-ion concentration in the electrolyte; Indicates the effective diffusion coefficient of the electrolyte; Indicates the coordinates along the battery thickness direction; Represents the lithium-ion transference number; Denotes Faraday's constant; Indicates the specific surface area of ​​the electrode; This indicates the current density at the electrode interface.

[0080] (3) Solid-state potential model

[0081] The potential distribution process in the electrode solid phase can be represented as follows:

[0082] ;

[0083] In the formula: Indicates the effective conductivity of the electrode solid phase; Indicates the solid-state potential of the electrode; Indicates the specific surface area of ​​the electrode; Denotes Faraday's constant; This indicates the current density at the electrode interface.

[0084] (4) Liquid phase potential model

[0085] The potential distribution process in the electrolyte phase can be represented as follows:

[0086] ;

[0087] In the formula: Indicates the effective conductivity of the electrolyte phase; Indicates the potential of the electrolyte phase; Represents the gas constant; Represents thermodynamic temperature; Denotes Faraday's constant; Represents the lithium-ion transference number; Indicates the lithium-ion concentration in the electrolyte;

[0088] (5) Interfacial reaction kinetic model

[0089] The electrochemical reaction process at the electrode or electrolyte interface is represented as follows:

[0090] ;

[0091] In the formula: This indicates the current density at the electrode interface. Indicates the exchange current density; and These represent the anodic and cathodic reaction transfer coefficients, respectively. Indicates electrode overpotential;

[0092] The electrode overpotential satisfies: ;

[0093] In the formula: This represents the open-circuit potential function related to the lithium-ion concentration at the electrode surface. This indicates the lithium ion concentration on the electrode surface.

[0094] Preferably, in step S4, the lithium battery coupling degradation model includes a battery terminal voltage model, a battery state of charge model, and a battery capacity model.

[0095] (1) Battery terminal voltage model:

[0096] ;

[0097] In the formula: Indicates the terminal voltage of the lithium battery; Indicates the solid-phase potential at the positive electrode; Indicates the negative electrode solid-phase potential; Indicates the positive electrode overpotential; Indicates the negative electrode overpotential; Indicates ohmic impedance; Indicates the application of current; Indicates the battery length;

[0098] (2) Battery state of charge model:

[0099] ;

[0100] In the formula: Indicates the state of charge of the lithium battery; This indicates the lithium-ion concentration on the negative electrode surface; This indicates the maximum solid-phase lithium-ion concentration at the negative electrode. Indicates particle radius;

[0101] (3) Battery capacity model:

[0102] ;

[0103] In the formula: Indicates the capacity of the lithium battery; This represents the specific surface area of ​​the negative electrode; This indicates the volume fraction of the negative electrode active material; This indicates the minimum solid-phase lithium-ion concentration at the negative electrode.

[0104] Preferably, step S5 specifically includes the following steps:

[0105] S501: Based on the initial state parameters and operating data of the lithium battery, the coupled degradation model of the lithium battery is solved in each cycle stage to obtain the internal electrochemical state variables and aging state variables of the battery in each cycle stage.

[0106] S502: Extract physical characteristics related to battery health status from electrochemical state variables and aging state variables; physical characteristics related to battery health status include lithium ion concentration on the negative electrode surface, interfacial film overpotential, lithium deposition overpotential, volume fraction of negative electrode active material, interfacial reaction current density, and battery capacity.

[0107] S503: Serialize physical features according to the number of iterations or sampling time, defining the... The physical features extracted in each cyclic stage are:

[0108] ;

[0109] In the formula: Indicates the first Physical features extracted in each cyclic stage; , , , , , They represent the first The lithium ion concentration on the negative electrode surface, the interfacial film overpotential, the lithium deposition overpotential, the volume fraction of the negative electrode active material, the interfacial reaction current density, and the battery capacity were extracted in each cycle stage.

[0110] S504: Using a preset length Constructing the sliding time window, the first The multi-step physical feature sequence corresponding to each cyclic stage is as follows:

[0111] ;

[0112] In the formula: Indicates the first The multi-step physical feature sequence corresponding to each cycle stage serves as the health status estimation model in the 1st cycle. The input for each loop stage.

[0113] Preferably, step S6 specifically includes the following processing steps:

[0114] S601: Obtain the training samples of the current batch. Each training sample contains a multi-step physical feature sequence of the corresponding cycle stage and the real label of the lithium battery health status.

[0115] S602: Input the multi-step physical feature sequence of the corresponding cycle stage into the health status estimation model, and output the estimated value of the lithium battery health status of the corresponding cycle stage.

[0116] S603: Calculate the physical residual loss of lithium-ion diffusion based on the concentration of lithium-ion in the negative electrode solid phase, the effective diffusion coefficient of the negative electrode solid phase, its time and radial spatial derivative in the multi-step physical characteristic sequence within the corresponding cycle stage.

[0117] S604: The residual loss of the data fitting is calculated based on the estimated state of health of the lithium battery at the corresponding cycle stage and the actual label of the state of health of the lithium battery.

[0118] S605: Calculate the joint training loss of the health status estimation model based on the physical residual loss of lithium-ion diffusion and the residual loss of data fitting;

[0119] S606: Backward optimization of the parameters of the health status estimation model based on joint training loss;

[0120] S607: Repeat steps S601 to S606 to iteratively train the health status estimation model until it converges, and obtain the trained health status estimation model.

[0121] Preferably, in step S603, the formula for calculating the physical residual loss of lithium-ion diffusion is:

[0122] ;

[0123] In the formula: This represents the physical residual loss due to lithium-ion diffusion. Indicates the number of samples in the current batch; Indicates the first The concentration of lithium ions in the negative electrode solid phase during each cycle stage; Indicates the first The effective diffusion coefficient of the negative electrode solid phase in each cycle stage; Indicates the radial coordinates of the electrode particles; Indicates time.

[0124] Preferably, in step S604, the formula for calculating the data fitting residual loss is:

[0125] ;

[0126] In the formula: This represents the data fitting residual loss; , , These represent the weighting coefficients for the state-of-charge residual, the health-state residual, and the terminal voltage residual, respectively. Indicates the first Lithium ion concentration on the negative electrode surface during each cycle stage; Indicates the first The maximum solid-phase lithium-ion concentration at the negative electrode during each cycle stage; Indicates the application of current; Indicates the first The current available capacity for each cycle stage; Indicates the first Estimated state of health of lithium-ion batteries at each cycle stage; Indicates the first A true label of the health status of lithium batteries at each cycle stage; Indicates the first The terminal voltage for each cycle stage is calculated from the battery terminal voltage model in the lithium battery coupling degradation model; Indicates the first Measured terminal voltage for each cycle stage;

[0127] In step S605, the formula for calculating the joint training loss is:

[0128] ;

[0129] In the formula: Represents the joint loss function; The weight representing the loss of physical residuals in lithium-ion diffusion; The weights represent the loss of the data fitting residuals.

[0130] Preferably, in step S6, during the training of the health status estimation model, the aging state parameters in the lithium battery coupling degradation model are updated based on the lithium battery health status estimate value under the current cycle stage output by the health status estimation model.

[0131] The updated aging state parameters are fed back to the lithium battery coupled degradation model to correct the electrochemical state variables and aging state variables of the next cycle stage, extract the physical characteristics of the next cycle stage, and generate a multi-step physical characteristic sequence of the next cycle stage.

[0132] The aging state parameters include one or more of the following: maximum solid-phase lithium-ion concentration, available capacity, negative electrode active material volume fraction, and interface film thickness. The aging state parameters are updated in the following ways: the maximum solid-phase lithium-ion concentration is updated based on lithium inventory loss; the current available capacity is updated based on the product of the estimated lithium battery health status at the current cycle stage and the initial capacity or rated capacity; the negative electrode active material volume fraction is updated based on active material loss; and the interface film thickness is updated based on the interface film growth model.

[0133] The lithium battery health status estimation method integrating physical information technology under shaking conditions in this invention has the following advantages compared with the prior art:

[0134] (1) By acquiring the cyclic aging test data of lithium batteries under shaking conditions, this invention analyzes the influence of shaking mechanical stress on the capacity decay, impedance evolution and degradation mode of lithium batteries, and can effectively determine the main aging characteristics of lithium batteries under shaking conditions, providing a basis for the degradation of working condition adaptability for health status estimation.

[0135] (2) The present invention establishes an aging degradation model for describing the growth of interfacial film, lithium deposition or stripping and electrode particle cracking process inside lithium battery, and a pseudo two-dimensional electrochemical model for describing the solid phase lithium ion diffusion, electrolyte lithium ion transport, solid phase potential distribution, liquid phase potential distribution and electrode or electrolyte interface reaction process inside lithium battery, and couples them to obtain a lithium battery coupled degradation model that integrates electrochemical reaction, interfacial film growth, lithium deposition or stripping and electrode particle cracking process, which can effectively characterize the lithium ion diffusion, interfacial side reaction and particle structure damage process inside lithium battery, and improve the model's ability to explain the degradation process inside battery.

[0136] (3) Based on the lithium battery coupling degradation model, the present invention extracts physical features with clear physical meaning, constructs a multi-step physical feature sequence as training samples, and combines lithium ion diffusion physical constraints to train the health status estimation model. This can reduce the dependence of the simple data-driven method on a large number of samples, and make the estimation results satisfy both data consistency and physical consistency, thereby improving the accuracy, computational efficiency and interpretability of lithium battery health status estimation under shaking conditions.

[0137] In summary, this invention determines battery aging characteristics through cyclic aging test data under shaking conditions, describes the internal lithium-ion diffusion, interfacial side reactions, and particle structure damage processes through a lithium battery coupled degradation model, and outputs the lithium battery health status through a health status estimation model that integrates the physical constraints of lithium-ion diffusion. This can reduce the dependence of a purely data-driven model on a large number of samples and improve the physical consistency and interpretability of the health status estimation. Attached Figure Description

[0138] To make the objectives, technical solutions, and advantages of the invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0139] Figure 1 This is a logic block diagram of a lithium battery health status estimation method that integrates physical information technology under shaking conditions.

[0140] Figure 2 The network structure diagram for training the health status estimation model.

[0141] Figure 3 The results of the lithium battery health status prediction under shaking conditions are as follows: Figure 3 (a) Prediction results of the health status of lithium batteries at a charge / discharge rate of 0.5C; Figure 3 (b) The predicted state of health of lithium batteries at a 1C charge / discharge rate; Figure 3 (c) The predicted state of health of lithium batteries at 2C charge / discharge rate; Figure 3 (d) shows the predicted health status of lithium batteries at 3C charge / discharge rates.

[0142] Figure 4 Quantification results of the prediction error of lithium battery health status at different rates: Figure 4 (a) is the quantification result of prediction error based on the MAPE index; Figure 4 (b) is the quantification result of the prediction error based on the RMSE index. Detailed Implementation

[0143] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but only to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0144] The following detailed explanation illustrates the specific implementation methods:

[0145] Example:

[0146] This embodiment discloses a method for estimating the health status of lithium batteries under shaking conditions by integrating physical information technology.

[0147] like Figure 1 As shown, the method for estimating the health status of lithium batteries under shaking conditions by integrating physical information technology includes:

[0148] S1: Conduct cyclic aging tests on lithium batteries under shaking conditions to determine the main aging characteristics under shaking conditions.

[0149] S2: Based on the main aging characteristics of shaking conditions, an aging degradation model is established to describe the growth of the internal interface film, lithium deposition or stripping, and the cracking process of electrode particles in lithium batteries.

[0150] S3: Establish a pseudo-two-dimensional electrochemical model to describe the solid-phase lithium-ion diffusion, electrolyte lithium-ion transport, solid-phase potential distribution, liquid-phase potential distribution, and electrode or electrolyte interface reaction processes inside a lithium battery.

[0151] S4: Couple the aging degradation model and the pseudo-two-dimensional electrochemical model to obtain the lithium battery coupled degradation model;

[0152] S5: Extract physical features related to the health status of lithium batteries based on the lithium battery coupling degradation model, and construct a multi-step physical feature sequence as training samples;

[0153] S6: Based on the multi-step physical feature sequence as training samples and their corresponding real labels of battery health status, a health status estimation model is trained in combination with the physical constraints of lithium-ion diffusion.

[0154] S7: Construct a multi-step physical feature sequence of the lithium battery to be predicted based on the operating data of the lithium battery during the application process, and input it into the trained health status estimation model to output the corresponding lithium battery health status estimate.

[0155] To better illustrate the technical solution of the present invention, this embodiment will be described in more detail through the following parts.

[0156] I. Main Aging Characteristics under Shaking Conditions

[0157] In the specific implementation process, the steps for determining the main aging characteristics of the shaking condition include:

[0158] S101: Construct a test environment including a shaking test platform, a battery charge / discharge test device, and an electrochemical impedance test device, as follows: Figure 2 As shown, a cycle aging test was conducted on the lithium battery under preset shaking parameters to obtain cycle aging test data; the cycle aging test data includes current, voltage, capacity, number of cycles, charge / discharge rate and impedance data;

[0159] S102: Electrochemical impedance spectroscopy and electrode material characterization were performed on lithium batteries at different aging stages to obtain electrochemical impedance spectroscopy and electrode material characterization results. The electrochemical impedance spectroscopy and electrode material characterization results include data related to impedance evolution, interfacial film growth, electrode particle cracking, and by-reaction product deposition.

[0160] S103: Based on cycle aging test data, electrochemical impedance test results, and electrode material characterization results, combined with capacity decay law, incremental capacity curve changes, and impedance evolution law, the main aging characteristics of lithium batteries under shaking mechanical stress are identified. The main aging characteristics include conductivity loss, lithium inventory loss, active material loss, interface film growth, electrode particle cracking, by-reaction product deposition, and mass transfer capability.

[0161] In this embodiment, conductivity loss, lithium inventory loss, and active material loss are aging characteristics further identified based on cyclic aging test data, incremental capacity curve changes, impedance evolution patterns, and electrode material characterization results. Mass transfer capability is an aging characteristic obtained based on electrochemical impedance test results and electrochemical mechanism analysis.

[0162] (1) Conductivity loss is mainly characterized by impedance evolution. Impedance parameters such as ohmic internal resistance can be obtained through electrochemical impedance testing and equivalent circuit parameter identification. As the number of cycles increases, the ohmic internal resistance increases, indicating that the internal conductive path of the battery deteriorates and charge transport is hindered, which can be identified as conductivity loss.

[0163] (2) Lithium inventory loss is mainly characterized by changes in the incremental capacity curve and the capacity decay pattern. The shift, decrease, or partial disappearance of the peak position in the incremental capacity curve can reflect the phenomenon of reduced cyclic lithium ions inside the battery, and thus lithium inventory loss can be identified. At the same time, continuous capacity decay can also reflect the impact of lithium inventory loss on the battery health status from a macroscopic perspective.

[0164] (3) The loss of active material is mainly characterized by the change in incremental capacity curve and the results of electrode material characterization. The attenuation of the peak value of the incremental capacity curve can reflect the reduction of the active reaction area; at the same time, the particle cracks, particle fissures, material structure damage and by-reaction product deposition observed in the electrode material characterization indicate that some active material can no longer effectively participate in the lithium insertion / delithiation reaction, and can therefore be identified as active material loss.

[0165] (4) Mass transfer capability is an aging characteristic obtained based on electrochemical impedance spectroscopy (EIS) test results and electrochemical mechanism analysis. Through EIS test and parameter identification, the mass transfer impedance related to diffusion and mass transfer processes can be obtained. When the mass transfer impedance increases, it indicates that the active material transfer rate decreases and the lithium-ion diffusion efficiency decreases, thus it can be identified as a decrease in mass transfer capability.

[0166] In addition, electrode particle cracking, changes in pore structure, deposition of by-reaction products, and loss of active materials can also affect the transport process of lithium ions in the electrode and electrolyte. Therefore, the results of electrode material characterization can also serve as an auxiliary basis for judging the decline in mass transfer capability.

[0167] II. Aging and Degradation Model

[0168] In practice, aging degradation models include interface film growth models, lithium deposition or stripping models, and electrode particle cracking models.

[0169] (1) Interfacial film growth model

[0170] During the cycling process of a lithium battery, the interface film on the negative electrode surface undergoes growth, rupture, and regeneration. The change in the interface film thickness is expressed as follows:

[0171] ;

[0172] In the formula: Indicates the thickness of the interfacial film; Indicates solvent molecule flux; Indicates the molar volume of the interfacial membrane; Indicates solvent molecule concentration; Indicates the solvent molecule diffusion coefficient;

[0173] The solvent molecule flux satisfies: , Indicates the temperature or thermodynamic temperature of the lithium battery;

[0174] The solvent molecule diffusion coefficient satisfies: ;

[0175] The boundary conditions are: ; ;

[0176] In the formula: Indicates the inherent diffusion coefficient of the electrolyte; Indicates the integral number of the electrolyzed liquid; Indicates the initial solvent molecule concentration; Indicates the coordinates of the interface film thickness direction;

[0177] The porosity change caused by interfacial film growth is as follows:

[0178] ;

[0179] In the formula: Indicates electrode porosity; This represents the specific surface area of ​​the negative electrode;

[0180] The overpotential of the interface film caused by the growth of the interface film is:

[0181] ;

[0182] In the formula: Indicates the overpotential of the interface film; This represents the interfacial reaction current density; Indicates the ohmic resistivity of the interface film; Indicates the interfacial membrane impedance;

[0183] The interfacial film growth model is used to characterize the enhanced side reactions, decreased porosity, and changes in polarization loss caused by repeated rupture and regeneration of the interfacial film under shaking mechanical stress.

[0184] (2) Lithium deposition or stripping model

[0185] The lithium deposition or stripping process is used to describe the deposition, stripping, and irreversible dead lithium formation of lithium ions on the negative electrode surface. The change in lithium ion concentration in the solid phase of the negative electrode is expressed as:

[0186] ;

[0187] In the formula: This indicates the lithium-ion concentration in the negative electrode solid phase; Indicates the lithium-ion stripping flux of the negative electrode; This represents the dead lithium decay rate related to the interface film thickness;

[0188] The dead lithium decay rate is expressed as:

[0189] ;

[0190] In the formula: This represents the initial dead lithium decay parameter; Indicates the initial interface film thickness;

[0191] Lithium inventory losses caused by lithium deposition or stripping processes are expressed as follows:

[0192] ;

[0193] In the formula: This indicates lithium inventory loss; This indicates the initial maximum solid-phase lithium-ion concentration at the negative electrode; This indicates the current maximum solid-phase lithium-ion concentration at the negative electrode;

[0194] The lithium deposition overpotential is expressed as:

[0195] ;

[0196] In the formula: Indicates the overpotential of lithium deposition; Indicates the negative electrode solid-phase potential; Indicates the potential of the electrolyte phase;

[0197] Lithium deposition or stripping models are used to characterize the loss of cyclic lithium ions during lithium battery cycling and its impact on health degradation.

[0198] (3) Electrode particle cracking model

[0199] The electrode particle cracking process is used to describe the particle crack propagation process under the combined action of shaking mechanical stress and the cyclic expansion and contraction of lithium battery.

[0200] The hydrostatic stress of the electrode particles is expressed as:

[0201] ;

[0202] In the formula: Indicates hydrostatic stress; Indicates radial stress; Indicates tangential stress;

[0203] The change in particle crack length is represented as:

[0204] ;

[0205] In the formula: Indicates the crack length; This parameter represents the particle cracking rate. Indicates the stress intensity factor; Indicates the particle cracking constant;

[0206] The change in specific surface area of ​​the negative electrode caused by particle cracking is expressed as:

[0207] ;

[0208] In the formula: Indicates the number of cracks; Indicates the crack width; Indicates the initial time parameter;

[0209] The change in particle radius is expressed as:

[0210] ;

[0211] In the formula: Indicates the molar volume of the negative electrode material; Indicates particle radius; Indicates the average lithium-ion concentration;

[0212] The change in interfacial film thickness at the crack is represented as follows:

[0213] ;

[0214] In the formula: Indicates the thickness of the interfacial film at the crack;

[0215] The change in volume fraction of the negative electrode active material is expressed as:

[0216] ;

[0217] In the formula: Indicates the volume fraction decay rate; Indicates the decay index; Indicates the yield strength of the negative electrode material;

[0218] The loss of active material is expressed as:

[0219] ;

[0220] In the formula: This indicates the loss of active materials; This indicates the initial volume fraction of the active material in the negative electrode. This indicates the current volume fraction of the active material in the negative electrode.

[0221] The electrode particle cracking model is used to characterize particle structure damage, loss of active material, and enhanced side reaction process caused by newly exposed surfaces under shaking conditions due to mechanical stress.

[0222] III. Pseudo-two-dimensional electrochemical model

[0223] In practice, the pseudo-two-dimensional electrochemical model includes a solid-phase lithium-ion diffusion model, an electrolyte lithium-ion transport model, a solid-phase potential model, a liquid-phase potential model, and an interfacial reaction kinetic model.

[0224] (1) Solid-phase lithium-ion diffusion model

[0225] Inside the negative and positive electrode particles, lithium ions diffuse radially along the particles. The solid-phase lithium ion diffusion process can be represented as follows:

[0226] ;

[0227] In the formula: These represent the negative and positive electrodes, respectively. Indicates the lithium-ion concentration in the electrode solid phase; Indicates the effective diffusion coefficient of the electrode solid phase; Indicates the radial coordinates of the electrode particles; Indicates time;

[0228] (2) Electrolyte lithium-ion transport model

[0229] In the thickness direction of the battery, the lithium-ion transport process in the electrolyte can be represented as follows:

[0230] ;

[0231] In the formula: Indicates the integral number of the electrolyzed liquid; Indicates the lithium-ion concentration in the electrolyte; Indicates the effective diffusion coefficient of the electrolyte; Indicates the coordinates along the battery thickness direction; Represents the lithium-ion transference number; Denotes Faraday's constant; Indicates the specific surface area of ​​the electrode; This indicates the current density at the electrode interface.

[0232] (3) Solid-state potential model

[0233] The potential distribution process in the electrode solid phase can be represented as follows:

[0234] ;

[0235] In the formula: Indicates the effective conductivity of the electrode solid phase; Indicates the solid-state potential of the electrode; Indicates the specific surface area of ​​the electrode; Denotes Faraday's constant; This indicates the current density at the electrode interface.

[0236] (4) Liquid phase potential model

[0237] The potential distribution process in the electrolyte phase can be represented as follows:

[0238] ;

[0239] In the formula: Indicates the effective conductivity of the electrolyte phase; Indicates the potential of the electrolyte phase; Represents the gas constant; Represents thermodynamic temperature; Denotes Faraday's constant; Represents the lithium-ion transference number; Indicates the lithium-ion concentration in the electrolyte;

[0240] (5) Interfacial reaction kinetic model

[0241] The electrochemical reaction process at the electrode or electrolyte interface is described using Butler-Volmer kinetics:

[0242] ;

[0243] In the formula: This indicates the current density at the electrode interface. Indicates the exchange current density; and These represent the anodic and cathodic reaction transfer coefficients, respectively. Indicates electrode overpotential;

[0244] The electrode overpotential satisfies: ;

[0245] In the formula: This represents the open-circuit potential function related to the lithium-ion concentration at the electrode surface. This indicates the lithium ion concentration on the electrode surface.

[0246] IV. Lithium-ion battery coupling degradation model

[0247] In practice, the lithium battery coupling degradation model includes a battery terminal voltage model, a battery state of charge model, and a battery capacity model.

[0248] (1) Battery terminal voltage model:

[0249] ;

[0250] In the formula: Indicates the terminal voltage of the lithium battery; Indicates the solid-phase potential at the positive electrode; Indicates the negative electrode solid-phase potential; Indicates the positive electrode overpotential; Indicates the negative electrode overpotential; Indicates ohmic impedance; Indicates the application of current; Indicates the battery length;

[0251] (2) Battery state of charge model:

[0252] ;

[0253] In the formula: Indicates the state of charge of the lithium battery; This indicates the lithium-ion concentration on the negative electrode surface; This indicates the maximum solid-phase lithium-ion concentration at the negative electrode. Indicates particle radius;

[0254] (3) Battery capacity model:

[0255]

[0256] In the formula: Indicates the capacity of the lithium battery; This represents the specific surface area of ​​the negative electrode; This indicates the volume fraction of the negative electrode active material; This indicates the minimum solid-phase lithium-ion concentration at the negative electrode.

[0257] The particle-scale degradation process is correlated with the external measurable state of the battery cell by using battery terminal voltage model, state of charge model and battery capacity model.

[0258] V. Multi-step physical characteristic sequence

[0259] In the specific implementation process, such as Figure 2 As shown, the processing steps for constructing a multi-step physical feature sequence include:

[0260] S501: Based on the initial state parameters and operating data of the lithium battery, the coupled degradation model of the lithium battery is solved in each cycle stage to obtain the internal electrochemical state variables and aging state variables of the battery in each cycle stage.

[0261] S502: Extract physical characteristics related to battery health status from electrochemical state variables and aging state variables; physical characteristics related to battery health status include lithium ion concentration on the negative electrode surface, interfacial film overpotential, lithium deposition overpotential, volume fraction of negative electrode active material, interfacial reaction current density, and battery capacity.

[0262] S503: Serialize physical features according to the number of iterations or sampling time, defining the... The physical features extracted in each cyclic stage are:

[0263] ;

[0264] In the formula: Indicates the first Physical features extracted in each cyclic stage; , , , , , They represent the first The lithium ion concentration on the negative electrode surface, the interfacial film overpotential, the lithium deposition overpotential, the volume fraction of the negative electrode active material, the interfacial reaction current density, and the battery capacity were extracted in each cycle stage.

[0265] S504: Using a preset length Constructing the sliding time window, the first The multi-step physical feature sequence corresponding to each cyclic stage is as follows:

[0266] ;

[0267] In the formula: Indicates the first The multi-step physical feature sequence corresponding to each cycle stage serves as the health status estimation model in the 1st cycle. The input for each cycle stage. A multi-step physical feature sequence is used to characterize the lithium-ion diffusion state, the degree of interfacial side reactions, and the trend of active material loss during continuous cycling of lithium batteries.

[0268] VI. Model Training

[0269] In the specific implementation process, the multi-step physical feature sequence is used as the input of the health status estimation model, the lithium battery health status estimate is used as the output of the health status estimation model, and the lithium ion diffusion physical residual and data fitting residual are introduced during the model training process to construct a health status estimation model that integrates physical constraints.

[0270] like Figure 2 As shown, the steps for training a health status estimation model include:

[0271] S601: Obtain the training samples of the current batch. Each training sample contains a multi-step physical feature sequence of the corresponding cycle stage and the real label of the lithium battery health status.

[0272] In this embodiment, the true health status label of the lithium battery can be obtained through capacity calibration tests at preset cycle stages. Specifically, standard charge-discharge capacity tests are performed on the lithium battery at different cycle stages to obtain the actual usable capacity of the current cycle stage. Then, the current actual usable capacity is compared with the initial capacity or rated capacity to obtain the true health status label of the lithium battery at that cycle stage.

[0273] S602: Input the multi-step physical feature sequence of the corresponding cycle stage into the health status estimation model, and output the estimated value of the lithium battery health status of the corresponding cycle stage.

[0274] S603: Calculate the physical residual loss of lithium-ion diffusion based on the negative electrode solid phase lithium-ion concentration, negative electrode solid phase effective diffusion coefficient and its time and radial spatial derivative (combined with Fick's second law or solid phase lithium-ion diffusion equation) in the multi-step physical characteristic sequence within the corresponding cycle stage.

[0275] S604: Calculate the data fitting residual loss based on the deviations between the estimated state of charge, the estimated state of health of the lithium battery, and the model-calculated terminal voltage at the corresponding cycle stage and the corresponding true or measured values.

[0276] S605: Calculate the joint training loss of the health status estimation model based on the physical residual loss of lithium-ion diffusion and the residual loss of data fitting;

[0277] S606: Backward optimization of the parameters of the health status estimation model based on joint training loss;

[0278] S607: Repeat steps S601 to S606 to iteratively train the health status estimation model until it converges, and obtain the trained health status estimation model.

[0279] Specifically, the formula for calculating the physical residual loss of lithium-ion diffusion is as follows:

[0280] ;

[0281] In the formula: This represents the physical residual loss due to lithium-ion diffusion. Indicates the number of samples in the current batch; Indicates the first The concentration of lithium ions in the negative electrode solid phase during each cycle stage; Indicates the first The effective diffusion coefficient of the negative electrode solid phase in each cycle stage (derived from the pseudo-two-dimensional electrochemical model, which is a parameter in the solid-phase lithium-ion diffusion model). Indicates the radial coordinates of the electrode particles; Indicates time.

[0282] Specifically, the formula for calculating the data fitting residual loss is as follows:

[0283] ;

[0284] In the formula: This represents the data fitting residual loss; , , These represent the weighting coefficients for the state-of-charge residual, the health-state residual, and the terminal voltage residual, respectively. Indicates the first Lithium ion concentration on the negative electrode surface during each cycle stage; Indicates the first The maximum solid-phase lithium-ion concentration at the negative electrode during each cycle stage; Indicates the application of current; Indicates the first The current available capacity for each cycle stage; Indicates the first Health status estimates for each cycle stage; Indicates the first A true label of the health status of lithium batteries at each cycle stage; Indicates the first The terminal voltage for each cycle stage is calculated from the battery terminal voltage model in the lithium battery coupled degradation model (derived from the lithium battery coupled degradation model, calculated from the battery terminal voltage model within it). Indicates the first Measured terminal voltage for each cycle stage;

[0285] Specifically, the formula for calculating the joint training loss is as follows:

[0286] ;

[0287] In the formula: Represents the joint loss function; The weight representing the loss of physical residuals in lithium-ion diffusion; The weights represent the loss of the data fitting residuals.

[0288] By minimizing the joint loss function, the health status estimation model can simultaneously satisfy the physical laws of lithium-ion diffusion and the requirements for fitting measured data.

[0289] VII. Feedback and Correction

[0290] In the specific implementation process, during the training of the health status estimation model, the aging state parameters in the lithium battery coupled degradation model are updated based on the lithium battery health status estimate value under the current cycle stage output by the health status estimation model.

[0291] The estimated state of health of a lithium battery is determined by the ratio of its current available capacity to its initial or rated capacity.

[0292] The aging state parameters include one or more of the following: maximum solid-phase lithium-ion concentration, available capacity, negative electrode active material volume fraction, and interface film thickness. The aging state parameters are updated in the following ways: updating the current available capacity based on the product of the estimated lithium battery health status at the current cycle stage and the initial capacity or rated capacity; updating the maximum solid-phase lithium-ion concentration based on lithium inventory loss; updating the negative electrode active material volume fraction based on active material loss; and updating the interface film thickness based on the interface film growth model.

[0293] The updated aging state parameters are fed back into the lithium battery coupled degradation model to correct the electrochemical state variables and aging state variables for the next cycle stage, extract the physical characteristics of the next cycle stage, and generate a multi-step physical characteristic sequence for the next cycle stage.

[0294] VIII. Experimental Instructions

[0295] like Figure 3 As shown, the health state estimation model (PIBM) constructed based on this invention and other existing models (BPLM and OLHM) can all fit the actual health state changes of lithium batteries well under different charge / discharge rates. Under low-rate conditions (such as...), Figure 3 (a) 0.5C charge / discharge rate, Figure 3 (b) at 1C charge / discharge rate), the model prediction of this invention is in high agreement with the actual value, indicating that the model is stable and reliable under mild cycling conditions; under high rate conditions (such as... Figure 3 (c) 2C charge / discharge rate, Figure 3 (d) 3C charge / discharge rate), although the prediction curve shows short-term fluctuations, the overall trend is still accurately captured, indicating that the model of the present invention can adapt to the high power load conditions of ships.

[0296] Furthermore, with Figure 4 (a) MAPE (Mean Absolute Percentage Error) and Figure 4(b) The RMSE (Root Mean Square Error) index quantifies the prediction errors of the three models (BPLM and OLHM are existing models, and PIBM is a model based on the method of this invention) at various rates. It can clearly reflect the accuracy of different models in capturing the mechanism of lithium battery health status. Among them, PIBM (i.e. the model based on the method of this invention) has the smallest error at various rates, indicating that it is the most accurate in predicting battery capacity decay and internal mechanism evolution. This verifies the ability of the coupled degradation model and health status estimation method to quantify the mechanism results under multiple rates and shaking conditions.

[0297] In summary, this invention determines battery aging characteristics through cyclic aging test data under shaking conditions, describes the internal lithium-ion diffusion, interfacial side reactions, and particle structure damage processes through a lithium battery coupled degradation model, and outputs the lithium battery health status through a health status estimation model that integrates the physical constraints of lithium-ion diffusion. This can reduce the dependence of a purely data-driven model on a large number of samples and improve the physical consistency and interpretability of the health status estimation.

[0298] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit the technical solutions. Those skilled in the art should understand that any modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A method for estimating the health status of lithium batteries under shaking conditions by integrating physical information technology, characterized in that, include: S1: Conduct cyclic aging tests on lithium batteries under shaking conditions to determine the main aging characteristics under shaking conditions. S2: Based on the main aging characteristics of shaking conditions, an aging degradation model is established to describe the growth of the internal interface film, lithium deposition or stripping, and the cracking process of electrode particles in lithium batteries. S3: Establish a pseudo-two-dimensional electrochemical model to describe the solid-phase lithium-ion diffusion, electrolyte lithium-ion transport, solid-phase potential distribution, liquid-phase potential distribution, and electrode or electrolyte interface reaction processes inside a lithium battery. S4: Couple the aging degradation model and the pseudo-two-dimensional electrochemical model to obtain the lithium battery coupled degradation model; S5: Extract physical features related to the health status of lithium batteries based on the lithium battery coupling degradation model, and construct a multi-step physical feature sequence as training samples; S6: Based on the multi-step physical feature sequence as training samples and their corresponding real labels of battery health status, a health status estimation model is trained in combination with the physical constraints of lithium-ion diffusion. Step S6 specifically includes the following processing steps: S601: Obtain the training samples of the current batch. Each training sample contains a multi-step physical feature sequence of the corresponding cycle stage and the real label of the lithium battery health status. S602: Input the multi-step physical feature sequence of the corresponding cycle stage into the health status estimation model, and output the estimated value of the lithium battery health status of the corresponding cycle stage. S603: Calculate the physical residual loss of lithium-ion diffusion based on the concentration of lithium-ion in the negative electrode solid phase, the effective diffusion coefficient of the negative electrode solid phase, its time and radial spatial derivative in the multi-step physical characteristic sequence within the corresponding cycle stage. S604: The residual loss of the data fitting is calculated based on the estimated state of health of the lithium battery at the corresponding cycle stage and the actual label of the state of health of the lithium battery. S605: Calculate the joint training loss of the health status estimation model based on the physical residual loss of lithium-ion diffusion and the residual loss of data fitting; S606: Backward optimization of the parameters of the health status estimation model based on joint training loss; S607: Repeat steps S601 to S606 to iteratively train the health status estimation model until convergence, and obtain the trained health status estimation model. S7: Construct a multi-step physical feature sequence of the lithium battery to be predicted based on the operating data of the lithium battery during the application process, and input it into the trained health status estimation model to output the corresponding lithium battery health status estimate.

2. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 1, characterized in that: Step S1 specifically includes the following processing steps: S101: Construct a test environment including a shaking test platform, a battery charge and discharge test device and an electrochemical impedance test device, and conduct a cycle aging test on the lithium battery under preset shaking parameters to obtain cycle aging test data. S102: Electrochemical impedance testing and electrode material characterization were performed on lithium batteries at different aging stages to obtain electrochemical impedance test results and electrode material characterization results. S103: Based on cycle aging test data, electrochemical impedance test results and electrode material characterization results, combined with capacity decay law, incremental capacity curve change and impedance evolution law, the main aging characteristics of lithium battery under shaking mechanical stress are identified. The main aging characteristics include conductivity loss, lithium inventory loss, active material loss, interfacial film growth, electrode particle cracking, by-reaction product deposition, and mass transfer capability.

3. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 1, characterized in that: In step S2, the aging degradation model includes an interface film growth model, a lithium deposition or stripping model, and an electrode particle cracking model: (1) Interfacial film growth model The change in interfacial film thickness is expressed as: ； In the formula: Indicates the thickness of the interfacial film; Indicates solvent molecule flux; Indicates the molar volume of the interfacial membrane; Indicates solvent molecule concentration; Indicates the solvent molecule diffusion coefficient; The solvent molecule flux satisfies: , Indicates the temperature or thermodynamic temperature of the lithium battery; The solvent molecule diffusion coefficient satisfies: ; The boundary conditions are: ; ; In the formula: Indicates the inherent diffusion coefficient of the electrolyte; Indicates the integral number of the electrolyzed liquid; Indicates the initial solvent molecule concentration; Indicates the coordinates of the interface film thickness direction; The porosity change caused by interfacial film growth is as follows: ； In the formula: Indicates electrode porosity; This represents the specific surface area of ​​the negative electrode; The overpotential of the interface film caused by the growth of the interface film is: ； In the formula: Indicates the overpotential of the interface film; This represents the interfacial reaction current density; Indicates the ohmic resistivity of the interface film; Indicates the interfacial membrane impedance; (2) Lithium deposition or stripping model The change in lithium-ion concentration in the negative electrode solid phase is expressed as: ； In the formula: This indicates the lithium-ion concentration in the negative electrode solid phase; Indicates the lithium-ion stripping flux of the negative electrode; This represents the dead lithium decay rate related to the interface film thickness; The dead lithium decay rate is expressed as: ； In the formula: This represents the initial dead lithium decay parameter; Indicates the initial interface film thickness; Lithium inventory losses caused by lithium deposition or stripping processes are expressed as follows: ； In the formula: This indicates lithium inventory loss; This indicates the initial maximum solid-phase lithium-ion concentration at the negative electrode; This indicates the current maximum solid-phase lithium-ion concentration at the negative electrode; The lithium deposition overpotential is expressed as: ； In the formula: Indicates the overpotential of lithium deposition; Indicates the negative electrode solid-phase potential; Indicates the potential of the electrolyte phase; (3) Electrode particle cracking model The hydrostatic stress of the electrode particles is expressed as: ； In the formula: Indicates hydrostatic stress; Indicates radial stress; Indicates tangential stress; The change in particle crack length is represented as: ； In the formula: Indicates the crack length; This parameter represents the particle cracking rate. Indicates the stress intensity factor; Indicates the particle cracking constant; The change in specific surface area of ​​the negative electrode caused by particle cracking is expressed as: ； In the formula: Indicates the number of cracks; Indicates the crack width; Indicates the initial time parameter; The change in particle radius is expressed as: ； In the formula: Indicates the molar volume of the negative electrode material; Indicates particle radius; Indicates the average lithium-ion concentration; The change in interfacial film thickness at the crack is represented as follows: ； In the formula: Indicates the thickness of the interfacial film at the crack; The change in volume fraction of the negative electrode active material is expressed as: ； In the formula: Indicates the volume fraction decay rate; Indicates the decay index; Indicates the yield strength of the negative electrode material; The loss of active material is expressed as: ； In the formula: This indicates the loss of active materials; This indicates the initial volume fraction of the active material in the negative electrode. This indicates the current volume fraction of the active material in the negative electrode.

4. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 3, characterized in that: In step S3, the pseudo-two-dimensional electrochemical model includes a solid-phase lithium-ion diffusion model, an electrolyte lithium-ion transport model, a solid-phase potential model, a liquid-phase potential model, and an interfacial reaction kinetic model. (1) Solid-phase lithium-ion diffusion model The solid-phase lithium-ion diffusion process is represented as follows: ； In the formula: These represent the negative and positive electrodes, respectively. Indicates the lithium-ion concentration in the electrode solid phase; Indicates the effective diffusion coefficient of the electrode solid phase; Indicates the radial coordinates of the electrode particles; Indicates time; (2) Electrolyte lithium-ion transport model In the thickness direction of the battery, the lithium-ion transport process in the electrolyte can be represented as follows: ； In the formula: Indicates the integral number of the electrolyzed liquid; Indicates the lithium-ion concentration in the electrolyte; Indicates the effective diffusion coefficient of the electrolyte; Indicates the coordinates along the battery thickness direction; Represents the lithium-ion transference number; Denotes Faraday's constant; Indicates the specific surface area of ​​the electrode; This indicates the current density at the electrode interface. (3) Solid-state potential model The potential distribution process in the electrode solid phase can be represented as follows: ； In the formula: Indicates the effective conductivity of the electrode solid phase; Indicates the solid-state potential of the electrode; Indicates the specific surface area of ​​the electrode; Denotes Faraday's constant; This indicates the current density at the electrode interface. (4) Liquid phase potential model The potential distribution process in the electrolyte phase can be represented as follows: ； In the formula: Indicates the effective conductivity of the electrolyte phase; Indicates the potential of the electrolyte phase; Represents the gas constant; Represents thermodynamic temperature; Denotes Faraday's constant; Represents the lithium-ion transference number; Indicates the lithium-ion concentration in the electrolyte; (5) Interfacial reaction kinetic model The electrochemical reaction process at the electrode or electrolyte interface is represented as follows: ； In the formula: This indicates the current density at the electrode interface. Indicates the exchange current density; and These represent the anodic and cathodic reaction transfer coefficients, respectively. Indicates electrode overpotential; The electrode overpotential satisfies: ; In the formula: This represents the open-circuit potential function related to the lithium-ion concentration at the electrode surface. This indicates the lithium ion concentration on the electrode surface.

5. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 4, characterized in that: In step S4, the lithium battery coupling degradation model includes the battery terminal voltage model, the battery state of charge model, and the battery capacity model. (1) Battery terminal voltage model: ； In the formula: Indicates the terminal voltage of the lithium battery; Indicates the solid-phase potential at the positive electrode; Indicates the negative electrode solid-phase potential; Indicates the positive electrode overpotential; Indicates the negative electrode overpotential; Indicates ohmic impedance; Indicates the application of current; Indicates the battery length; (2) Battery state of charge model: ； In the formula: Indicates the state of charge of the lithium battery; This indicates the lithium-ion concentration on the negative electrode surface; This indicates the maximum solid-phase lithium-ion concentration at the negative electrode. Indicates particle radius; (3) Battery capacity model: ； In the formula: Indicates the capacity of the lithium battery; This represents the specific surface area of ​​the negative electrode; This indicates the volume fraction of the negative electrode active material; This indicates the minimum solid-phase lithium-ion concentration at the negative electrode.

6. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 1, characterized in that: Step S5 specifically includes the following steps: S501: Based on the initial state parameters and operating data of the lithium battery, the coupled degradation model of the lithium battery is solved in each cycle stage to obtain the internal electrochemical state variables and aging state variables of the battery in each cycle stage. S502: Extract physical characteristics related to battery health status from electrochemical state variables and aging state variables; physical characteristics related to battery health status include lithium ion concentration on the negative electrode surface, interfacial film overpotential, lithium deposition overpotential, volume fraction of negative electrode active material, interfacial reaction current density, and battery capacity. S503: Serialize physical features according to the number of iterations or sampling time, defining the... The physical features extracted in each cyclic stage are: ； In the formula: Indicates the first Physical features extracted in each cyclic stage; , , , , , They represent the first The lithium ion concentration on the negative electrode surface, the interfacial film overpotential, the lithium deposition overpotential, the volume fraction of the negative electrode active material, the interfacial reaction current density, and the battery capacity were extracted in each cycle stage. S504: Using a preset length Constructing the sliding time window, the first The multi-step physical feature sequence corresponding to each cyclic stage is as follows: ； In the formula: Indicates the first The multi-step physical feature sequence corresponding to each cycle stage serves as the health status estimation model in the 1st cycle. The input for each loop stage.

7. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 1, characterized in that: In step S603, the formula for calculating the physical residual loss of lithium-ion diffusion is: ； In the formula: This represents the physical residual loss due to lithium-ion diffusion. Indicates the number of samples in the current batch; Indicates the first The concentration of lithium ions in the negative electrode solid phase during each cycle stage; Indicates the first The effective diffusion coefficient of the negative electrode solid phase in each cycle stage; Indicates the radial coordinates of the electrode particles; Indicates time.

8. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 7, characterized in that: In step S604, the formula for calculating the data fitting residual loss is: ； In the formula: This represents the data fitting residual loss; , , These represent the weighting coefficients for the state-of-charge residual, the health-state residual, and the terminal voltage residual, respectively. Indicates the first Lithium ion concentration on the negative electrode surface during each cycle stage; Indicates the first The maximum solid-phase lithium-ion concentration at the negative electrode during each cycle stage; Indicates the application of current; Indicates the first The current available capacity for each cycle stage; Indicates the first Estimated state of health of lithium-ion batteries at each cycle stage; Indicates the first A true label of the health status of lithium batteries at each cycle stage; Indicates the first The terminal voltage for each cycle stage is calculated from the battery terminal voltage model in the lithium battery coupling degradation model; Indicates the first Measured terminal voltage for each cycle stage; In step S605, the formula for calculating the joint training loss is: ； In the formula: Represents the joint loss function; The weight representing the loss of physical residuals in lithium-ion diffusion; The weights represent the loss of the data fitting residuals.

9. The lithium battery health status estimation method integrating physical information technology under shaking conditions as described in claim 6, characterized in that: In step S6, during the training of the health status estimation model, the aging state parameters in the lithium battery coupled degradation model are updated based on the lithium battery health status estimate value under the current cycle stage output by the health status estimation model. The updated aging state parameters are fed back to the lithium battery coupled degradation model to correct the electrochemical state variables and aging state variables of the next cycle stage, extract the physical characteristics of the next cycle stage, and generate a multi-step physical characteristic sequence of the next cycle stage. The aging state parameters include one or more of the following: maximum solid-phase lithium-ion concentration, available capacity, negative electrode active material volume fraction, and interface film thickness. The aging state parameters are updated in the following ways: the maximum solid-phase lithium-ion concentration is updated based on lithium inventory loss; the current available capacity is updated based on the product of the estimated lithium battery health status at the current cycle stage and the initial capacity or rated capacity; the volume fraction of negative electrode active material is updated based on active material loss; and the interface film thickness is updated based on the interface film growth model.

Images