A lithium battery SOH prediction method based on a multi-physical field coupling model

Abstract

The present application relates to lithium ion battery modeling and health state evaluation technical field, and disclose a kind of lithium battery SOH prediction method based on multi-physics coupling model, comprising the following steps, first, the electrochemical model including SEI film growth and lithium deposition is built;Then the coupling of thermal field and force field constitutes lithium battery multi-physics model;Then the coupling of SEI film thickening and lithium deposition side reaction constitutes side reaction lithium loss model of coupling multi-physics, and the SOH of lithium battery is predicted, then parameter is optimized, obtains the final side reaction lithium loss model;Finally, the feasibility of the final coupling multi-physics side reaction lithium loss model is verified, and auxiliary diagnosis, the present application is coupled modeling by electrochemistry, thermology and mechanics three physical fields, fully describes the internal side reaction behavior of lithium battery, focuses on introducing SEI film growth and two key reaction mechanisms of metal lithium deposition, can dynamically reflect the regulation and control effect of temperature and stress on reaction rate and lithium loss.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of lithium ion battery modeling and state of health evaluation, in particular to a lithium battery SOH prediction method based on a multi-physical field coupling model. BACKGROUND

[0002] Lithium ion batteries involve multiple physical fields such as electrochemical reactions, heat conduction and structural stress during use, and these coupling effects have a significant impact on battery performance, safety and life. Among them, side reactions such as SEI film growth and metal lithium deposition are the main source of irreversible lithium loss, directly leading to battery capacity decline and SOH (state of health) degradation. Although existing models partially consider thermal-electric-mechanical coupling, most do not systematically establish the physical path between side reactions and SOH, and lack a dynamic parameter identification mechanism that combines voltage response and impedance spectrum. Impedance spectrum changes (such as SEI impedance and charge transfer impedance) as an important observation dimension of side reaction strength and aging degree, the integration value in modeling needs to be further explored. It is of great significance to build a model that can integrate multi-physical field modeling, dynamic parameter updating, SOH prediction and impedance response output, to improve the accuracy and engineering applicability of battery aging modeling.

[0003] Most existing lithium ion battery modeling methods focus on a single physical field, or only achieve partial coupling of electrochemistry, heat and force, making it difficult to fully capture the dynamic feedback effects between physical fields. Especially when modeling electrochemical processes, mechanical stress is often ignored in lithium ion intercalation / deintercalation and side reaction behavior (such as SEI film growth and lithium deposition), resulting in inaccurate prediction of capacity loss and its evolution process on SOH. At the same time, traditional thermal models are mostly run independently and do not form an effective closed-loop feedback between temperature rise and stress, reaction rate, affecting the accuracy of hot spot and aging prediction. In addition, existing methods generally do not dynamically associate model output with impedance spectrum changes, making it difficult to achieve external verification and perception of modeling results. Therefore, it is urgent to build a highly coupled multi-physical field modeling method that can describe side reaction mechanisms and simultaneously output SOH estimation and impedance characteristics, to improve the prediction ability, diagnostic accuracy and engineering applicability of the model.

[0004] Most of the existing lithium-ion battery modeling methods focus on a single physical field or only achieve partial coupling of electrochemistry, heat, and force, making it difficult to fully capture the dynamic feedback effects between various physical fields. Especially when modeling the secondary reaction mechanism, the regulatory effect of stress and temperature on lithium ion intercalation / deintercalation and secondary reaction behavior (such as SEI film growth and lithium deposition) is often ignored, resulting in a large deviation in the prediction of capacity loss and SOH evolution. At the same time, traditional thermal models are mostly run independently and do not form an effective closed-loop feedback between heat sources and secondary reaction processes, affecting the accuracy of hot spot and aging estimation. In addition, the current model parameters lack a dynamic identification process based on experimental observation data, especially the combination of voltage response and frequency domain impedance spectrum as joint constraints for parameter inversion, resulting in insufficient model accuracy and external verifiability. Therefore, it is urgent to build a SOH prediction model that is highly coupled with multiple physical fields, can describe the secondary reaction mechanism, and realizes the feedback of voltage and impedance in two domains, in order to improve the credibility and diagnostic value of modeling. SUMMARY

[0005] The present application is based on a multi-physical field coupling model, which comprehensively considers the mutual feedback mechanism between electrochemistry, heat conduction and mechanical stress, dynamically corrects the reaction parameters, can more comprehensively predict the performance change of the battery during charging and discharging, adjusts the secondary reaction rate through the temperature field and the stress field, captures potential degradation signals such as abnormal temperature rise, structural expansion or stress concentration, realizes early perception of the battery health state, further, the model introduces a parameter identification method of double objective function, jointly considers the voltage response and impedance spectrum characteristics, realizes accurate inversion of key parameters such as structural parameters, diffusion coefficient and secondary reaction rate through optimization algorithm, thereby improving the prediction accuracy of the model on the SOH change trend, this method is suitable for various types of lithium batteries and different working conditions, has good engineering adaptability, the model also combines the frequency domain disturbance response output impedance spectrum, which can reflect the influence of electrolyte, charge transfer, SEI film thickening and lithium deposition on the internal resistance structure of the battery, through the fusion of temperature, stress and impedance three types of physical observation data, improves the accuracy of fault diagnosis and SOH prediction, provides mechanism support and model basis for lithium battery life management and safety monitoring, and a lithium battery SOH prediction method based on a multi-physical field coupling model is proposed.

[0006] The technical solution of the present application to solve the above technical problems is as follows:

[0007] A lithium battery SOH prediction method based on a multi-physical field coupling model, comprising the following steps:

[0008] S1: constructing an electrochemical model including SEI film growth and lithium deposition;

[0009] S2: constructing a lithium battery multi-physical field model by coupling the thermal field and the force field with the electrochemical model in S1;

[0010] S3: coupling the SEI film thickening and lithium deposition side reaction to the lithium battery multi-physical field model in S2 to form a coupled multi-physical field side reaction lithium loss model, and using the coupled multi-physical field side reaction lithium loss model to predict the SOH of the lithium battery;

[0011] S4: optimizing the parameters of the coupled multi-physical field side reaction lithium loss model in S3 to obtain a final side reaction lithium loss model;

[0012] S5: constructing a model impedance spectrum for verifying the feasibility of the final coupled multi-physical field side reaction lithium loss model in S4 and assisting in diagnosis.

[0013] On the basis of the above technical solutions, the application can also be improved as follows.

[0014] Preferably, the S1 comprises the following steps:

[0015] S1.1, constructing a pseudo two-dimensional model as a basic electrochemical model of the lithium ion battery;

[0016] S1.2, using a diffusion-interface reaction cooperative control rate expression to construct a SEI film growth and lithium deposition side reaction flux model, wherein the SEI film thickening side reaction occurs on the negative electrode surface and is represented as:

[0017]

[0018] wherein, is the SEI reaction current density, is the SEI solid volume fraction, is the liquid concentration, is the SEI growth reaction rate constant, is the SEI growth reaction lithium ion transfer coefficient, is the diffusion rate of the electrolyte in the SEI, is the SEI thickness, is the SEI layer overpotential, F is the Faraday constant, R is the universal gas constant, and T is the battery temperature;

[0019] wherein the lithium deposition side reaction adopts a bidirectional kinetics model based on the Butler-Volmer equation, and the metal lithium deposition process is represented as:

[0020]

[0021] wherein, is the lithium deposition reaction current density, and represents the rate of the lithium deposition process, is the overpotential of the lithium deposition reaction, and the reaction only occurs when the voltage is below 0V, and are the charge transfer coefficients of the anode and the cathode, respectively, The reference current density for lithium deposition reaction.

[0022] Preferably, the S2 comprises the following steps:

[0023] S2.1, by Ohmic heat Q J , reaction heat Q M and entropy change term Q rev Solve the temperature field distribution, the heat conduction process in the model is simplified as one-dimensional heat transfer along the thickness direction of the battery, and the temperature change in the battery can be described by the following energy conservation equation:

[0024]

[0025] wherein, represents the equivalent density of the battery, C p is the equivalent specific heat capacity, is the equivalent thermal conductivity;

[0026] The change of the battery temperature is not only affected by the internal heat source, but also affected by the heat exchange between the environment, and the boundary heat exchange can be expressed as:

[0027]

[0028] wherein, is the heat exchange coefficient, is the ambient temperature, is the battery temperature;

[0029] S2.2, the temperature change is fed back to the electrochemical model based on the Arrhenius equation, and the changes of the side reaction rate and the electrochemical parameters are dynamically corrected, and the correction expression of the temperature related parameters is as follows:

[0030]

[0031] wherein, T ref is the reference temperature, Y ref is the parameter value at the reference temperature, Y(T) represents the temperature related parameter, Ea is the activation energy corresponding to Y(T), and R is the universal gas constant;

[0032] S2.3, the lithium ion concentration field in the positive and negative active materials distributed along the thickness direction of the battery is calculated according to the electrochemical model, the stress-strain distribution in the electrode is derived due to the volume expansion effect caused by the embedding and stripping of lithium, and the stress field distribution is established by combining the volume strain caused by the embedding of lithium ions and the thermal strain, which is represented as:

[0033]

[0034] wherein, is the total strain tensor, is the elastic strain tensor, is the diffusion strain tensor, is the thermal strain tensor.

[0035] Preferably, the S3 comprises the following steps:

[0036] S3.1, further introducing the thermal field and force field coupling mechanism on the basis of the lithium battery multi-physical field model in S2, using the Arrhenius correction of temperature on reaction rate constant, the adjustment effect of stress on reaction overpotential, and the thermal-force double feedback correction of diffusion coefficient, so as to dynamically capture the evolution characteristics of SEI growth behavior with the change of environment and working condition, which is specifically expressed as:

[0037]

[0038]

[0039]

[0040] wherein, is the SEI growth reaction rate constant coupled with the temperature field, is the initial SEI growth reaction rate constant, is the SEI reaction activation energy, R is the universal gas constant, and T is the battery temperature; is the SEI reaction overpotential coupled with the force field, is the initial SEI reaction overpotential, β is the stress coupling coefficient, and σ h is the volume stress at the calculation point; is the diffusion coefficient of EC in SEI coupled with the force field and the temperature field, represents the initial diffusion coefficient of EC in SEI, E D represents the activation energy of EC diffusion, and γ is the stress feedback coefficient;

[0041] S3.2, introducing the temperature-modified exchange current density term and the stress-modulated overpotential term, so as to construct the deposition reaction rate expression reflecting the dynamic change of negative microenvironment and physical field;

[0042]

[0043]

[0044] wherein, is the lithium deposition reaction reference exchange current density coupled with the temperature field, is the initial lithium deposition reaction reference exchange current density, is the lithium deposition reaction activation energy; is the lithium deposition reaction overpotential coupled with the force field, is the initial lithium deposition reaction overpotential, β is the stress coupling coefficient, σ h is the volume stress at the calculation point;

[0045] S3.3, the SEI film thickening in the battery and the current density caused by lithium deposition reaction are taken as lithium loss source terms, and the irreversible consumption of lithium is estimated by time integration through Faraday's law, and combined with the rated capacity of the battery, a health state estimation formula for the whole life cycle is established:

[0046]

[0047]

[0048] where F is the Faraday constant, and is the SEI thickening and lithium deposition reaction current density varying with time, is the lithium loss term, is the rated capacity of the battery, is the effective area of the negative electrode reaction.

[0049] Preferably, the S4 comprises the following steps:

[0050] S4.1, the parameters involved in the coupled multi-physical field side reaction lithium loss model in S3 are classified, including four categories of geometric parameters, electrochemical parameters, thermal parameters and mechanical parameters;

[0051] S4.2, a parameter identification method based on a double objective function is adopted, the time domain voltage response and the frequency domain impedance spectrum characteristics are considered jointly, the fitting error of the two types of data is weighted and fused by constructing a comprehensive loss function, the particle swarm optimization algorithm is used for global search inversion, so as to obtain the optimal parameter combination, and the consistency of the model simulation output and the experimental results is improved; in the parameter identification process, the objective function constructed is:

[0052]

[0053] wherein, is the objective function of parameter identification, θ is the set of parameters to be identified, λ is the error weighting factor, which is used to balance the influence of time domain and frequency domain data on the identification result, is the simulation voltage response of the model at time t i when the parameter is θ, is the voltage response of the experiment at time t i ; is the simulation impedance value of the model at frequency ω j when the parameter is θ, is the experimental impedance value at frequency ω jImpedance value on the upper side;

[0054] To minimize the loss function, the particle swarm algorithm is used to search for the optimal parameter combination by simulating the updating of the position of particles in a multidimensional space under the guidance of "individual experience" and "group optimization". The particle position updating rule is as follows:

[0055]

[0056] The inertia weight is adjusted dynamically and the boundary constraint strategy is adopted to improve the global search ability and convergence efficiency of the particle swarm optimization algorithm. The inertia weight is gradually reduced from the initial value 0.9 to 0.4 according to the linear decreasing rule, so that the particle swarm algorithm has strong global search ability in the early stage and strong local convergence ability in the later stage. The boundary constraint adopts a hybrid strategy combining "rebound and clamping", which ensures that the parameters are always within the reasonable physical range. The particle swarm size is set to 30 to 50 to cover the parameter space while controlling the computational burden. The maximum number of iterations is set to 200 to 300 to ensure the convergence accuracy.

[0057] Preferably, the S5 comprises the following steps:

[0058] The dynamic response data of the battery at different aging stages are collected through the final coupling multi-physical field side reaction lithium loss model in S4. The battery is excited and analyzed using EIS technology, that is, a small amplitude sinusoidal current disturbance is applied at multiple set frequencies, and the corresponding voltage response signal is collected synchronously to obtain the complex impedance Z(ω) of the battery at each frequency:

[0059]

[0060] wherein, and are the complex representations of the voltage and current of the battery at the frequency

[0061] The amplitude and phase information in the frequency domain are extracted by performing Fourier transform on the time domain voltage and current signals to obtain complete impedance spectrum data. The complex impedance at multiple frequency points is plotted into a Nyquist plot to visualize the trend of changes in the electrochemical properties of the battery. In combination with the multi-physical field coupling information of the temperature field, electric field and concentration field of the battery, the changes in key parameters during the aging process are analyzed, including the interface reaction rate constant k, the interface resistance R film , and the lithium ion diffusion coefficient D s ​, reveal the evolution mechanism of the internal reaction kinetics and mass transfer behavior of the battery; the above parameter changes are manifested as response changes of characteristic frequency bands in the impedance spectrum, by comparing the impedance characteristics with the model output, the relationship between the changes of key parameters and the health state is understood, the comprehensive analysis from the time domain voltage response and the frequency domain impedance spectrum characteristics to the evolution process of the multi-physical mechanism is realized, and the accuracy and physical interpretability of the SOH prediction are improved.

[0062] Compared with the prior art, the technical scheme of the present application has the following beneficial technical effects:

[0063] 1. The present application comprehensively describes the internal side reaction behavior of lithium battery through the coupling modeling of three physical fields of electrochemistry, thermodynamics and mechanics, focuses on introducing two key reaction mechanisms of SEI film growth and metal lithium deposition, and can dynamically reflect the regulation effect of temperature and stress on reaction rate and lithium loss.

[0064] 2. The present application further accurately estimates irreversible lithium loss by integrating the side reaction current density, realizes the physical driving prediction of capacity attenuation and SOH evolution trend.

[0065] 3. The present application builds a time domain and frequency domain double-target comprehensive loss function, and uses particle swarm optimization algorithm to inverse multiple key parameters, improves the modeling accuracy and robustness of SOH prediction, and is especially suitable for health monitoring and life management scenarios of lithium battery in multiple working conditions such as electric vehicles and energy storage systems. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 The flowchart of the present application is a lithium battery SOH prediction method based on a multi-physical field coupling model;

[0067] Figure 2 The voltage curve graph of the model simulation and experimental results;

[0068] Figure 3 The SEI and lithium deposition reaction current density graph of the model;

[0069] Figure 4 The temperature curve graph of the model simulation and experimental results;

[0070] Figure 5 The impedance spectrum graph of the model simulation and experimental results;

[0071] Figure 6 The SOH prediction and experimental measurement SOH graph of the model. DETAILED DESCRIPTION

[0072] With reference to the accompanying drawings: clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0073] A lithium battery SOH prediction method based on a multi-physical field coupling model, comprising the following steps:

[0074] S1: constructing an electrochemical model containing SEI film growth and lithium deposition;

[0075] S2: constructing a lithium battery multi-physical field model by coupling the thermal field and the force field in combination with the electrochemical model in S1;

[0076] S3: constructing a coupled multi-physical field side reaction lithium loss model by coupling the SEI film thickening and lithium deposition side reaction in combination with the lithium battery multi-physical field model in S2, and using the coupled multi-physical field side reaction lithium loss model to predict the SOH of the lithium battery;

[0077] S4: optimizing the parameters of the coupled multi-physical field side reaction lithium loss model in S3 to obtain the final side reaction lithium loss model;

[0078] S5: constructing a model impedance spectrum for verifying the feasibility of the final coupled multi-physical field side reaction lithium loss model in S4 and assisting in diagnosis.

[0079] The S1 comprises the following steps:

[0080] S1.1, constructing a pseudo two-dimensional model as a basic electrochemical model of the lithium ion battery;

[0081] S1.2, constructing an SEI film growth and lithium deposition side reaction flux model using a diffusion-interface reaction cooperative control rate expression, wherein the SEI film thickening side reaction occurs on the negative electrode surface and is expressed as:

[0082]

[0083] wherein, is the SEI reaction current density, is the SEI solid volume fraction, is the liquid concentration, is the SEI growth reaction rate constant, is the SEI growth reaction lithium ion transfer coefficient, is the diffusion rate of the electrolyte in the SEI, is the SEI thickness, for the SEI layer overpotential, F is the Faraday constant, R is the universal gas constant, and T is the battery temperature;

[0084] where the lithium extraction side reaction adopts a two-way kinetic model based on the Butler-Volmer equation, and the process of metal lithium deposition is expressed as:

[0085]

[0086] wherein, is the lithium extraction reaction current density, indicating the rate of the lithium extraction process, is the overpotential of the lithium extraction reaction, and the reaction only occurs below 0V, and are the charge transfer coefficients of the anode and the cathode, respectively, is the reference current density of the lithium deposition reaction.

[0087] The S2 includes the following steps:

[0088] S2.1, solving the temperature field distribution by Ohmic heat Q J , reaction heat Q M and entropy change Q rev The heat conduction process in the model is simplified as one-dimensional heat transfer along the thickness direction of the battery, and the temperature change in the battery can be described by the following energy conservation equation:

[0089]

[0090] wherein, represents the equivalent density of the battery, C p is the equivalent specific heat capacity, is the equivalent thermal conductivity;

[0091] The change of the battery temperature is not only affected by the internal heat source, but also affected by the heat exchange between the environment, and the boundary heat exchange can be expressed as:

[0092]

[0093] wherein, is the heat exchange coefficient, is the environment temperature, is the battery temperature;

[0094] S2.2, feedback the temperature change to the electrochemical model based on the Arrhenius equation, dynamically correct the changes of the side reaction rate and the electrochemical parameters, and the correction expression of the temperature related parameters is as follows:

[0095]

[0096] wherein, T ref is the reference temperature, Yref Y(T) represents a parameter related to temperature, Ea is an activation energy corresponding to Y(T), and R is a general gas constant;

[0097] S2.3, the lithium ion concentration field in the positive and negative active materials along the thickness direction of the battery is calculated according to the electrochemical model, the volume expansion effect caused by the insertion and extraction of lithium is derived to obtain the stress-strain distribution inside the electrode, and the volume strain caused by the insertion of lithium ions is combined with the thermal strain to establish the stress field distribution, which is represented as:

[0098]

[0099] wherein, is a total strain tensor, is an elastic strain tensor, is a diffusion strain tensor, is a thermal strain tensor.

[0100] The S3 includes the following steps:

[0101] S3.1, the thermal field and force field coupling mechanism is further introduced on the basis of the lithium battery multi-physical field model in S2, the Arrhenius correction of temperature on reaction rate constant, the adjustment effect of stress on reaction overpotential, and the thermal-force double feedback correction of diffusion coefficient are used, so that the evolution characteristics of SEI growth behavior with the change of environment and working condition are dynamically captured, which is specifically represented as:

[0102]

[0103]

[0104]

[0105] wherein, is the SEI growth reaction rate constant coupled with the temperature field, is the initial SEI growth reaction rate constant, is the SEI reaction activation energy, R is the general gas constant, and T is the battery temperature; is the SEI reaction overpotential coupled with the force field, is the initial SEI reaction overpotential, β is the stress coupling coefficient, σ h is the volume stress at the calculation point; is the diffusion coefficient of EC (ethylene carbonate) in SEI coupled with the force field and the temperature field, represents the initial diffusion coefficient of EC (ethylene carbonate) in SEI, E D represents the activation energy of EC diffusion, and γ is the stress feedback coefficient;

[0106] S3.2, introduce the temperature-corrected exchange current density term and the stress-modulated overpotential term, thereby constructing a deposition reaction rate expression that reflects the dynamic changes in the negative microenvironment and physical field;

[0107]

[0108]

[0109] wherein, is the reference exchange current density of lithium deposition reaction coupled with temperature field, is the initial reference exchange current density of lithium deposition reaction, is the activation energy of lithium deposition reaction; is the overpotential of lithium deposition reaction coupled with force field, is the initial overpotential of lithium deposition reaction, β is the stress coupling coefficient, σ h is the volume stress at the calculation point;

[0110] S3.3, take the SEI film thickening and the current density caused by lithium deposition reaction in the battery as lithium loss source terms, and perform time integration through Faraday's law to estimate the irreversible consumption of lithium, and then combine the rated capacity of the battery to establish a health state estimation formula facing the whole life cycle:

[0111]

[0112]

[0113] wherein, F is the Faraday constant, and are the SEI thickening and the lithium deposition reaction current density varying with time, is the lithium loss term, is the rated capacity of the battery, is the effective area of the negative reaction.

[0114] The S4 comprises the following steps:

[0115] S4.1, classify the parameters involved in the multi-physical field coupled side reaction lithium loss model in S3, including four categories of geometric parameters, electrochemical parameters, thermal parameters and mechanical parameters;

[0116] S4.2, adopt a parameter identification method based on a double-objective function, jointly consider the voltage response in the time domain and the impedance spectrum characteristics in the frequency domain, fuse the fitting errors of the two types of data by constructing a comprehensive loss function, use a particle swarm optimization algorithm for global search inversion to obtain the optimal parameter combination, and improve the consistency between the model simulation output and the experimental results; in the parameter identification process, the objective function constructed has the form:

[0117]

[0118] wherein, is the objective function for parameter identification, θ is the set of parameters to be identified, λ is the error weighting factor for balancing the influence of time-domain and frequency-domain data on the identification result, is the simulated voltage response of the model at time t i with parameters θ, is the voltage response of the experiment at time t i ; is the simulated impedance value of the model at frequency ω j with parameters θ, is the impedance value of the experiment at frequency ω j ;

[0119] To minimize this loss function, a particle swarm optimization algorithm is used to search for the optimal parameter combination by simulating the updating of particle positions in a multidimensional space guided by both individual experience and group optimization. The particle position updating rule is as follows:

[0120]

[0121] A dynamically adjusted inertia weight and boundary constraint strategy are used to improve the global search ability and convergence efficiency of the particle swarm optimization algorithm. The inertia weight is gradually reduced from the initial value of 0.9 to 0.4 according to the linear decreasing rule, which enables the particle swarm optimization algorithm to have strong global search ability in the early stage and strong local convergence ability in the later stage. The boundary constraint adopts a hybrid strategy combining "rebound" and "clamping" to ensure that the parameters are always within a physically reasonable range. The particle swarm size is set to 30 to 50 to cover the parameter space while controlling the computational burden. The maximum number of iterations is set to 200 to 300 to ensure convergence accuracy. If necessary, an early termination mechanism can be introduced to automatically stop iteration based on the convergence trend of the loss function, thereby improving the identification efficiency and avoiding falling into local optimum.

[0122] The S5 includes the following steps:

[0123] The dynamic response data of the battery at different aging stages are collected through the final coupling multi-physics side reaction lithium loss model in S4. The battery is excited and analyzed using Electrochemical Impedance Spectroscopy (EIS) technology, i.e., a small amplitude sinusoidal current disturbance is applied at multiple set frequencies, and the corresponding voltage response signal is synchronously collected to obtain the complex impedance Z(ω) of the battery at each frequency:

[0124]

[0125] wherein, and are the complex representation of the voltage and current of the battery at frequency

[0126] By Fourier transforming the time domain voltage and current signals, the amplitude and phase information in the frequency domain is extracted, and the complete impedance spectrum data is obtained. The complex impedance at multiple frequency points is plotted into a Nyquist plot to visualize the trend of the electrochemical characteristics of the battery. Furthermore, by analyzing the changes of key parameters during the aging process, including the interfacial reaction rate constant k, the interfacial resistance R film , and the lithium ion diffusion coefficient D s , the changes of these physical parameters will be manifested in the impedance spectrum characteristic changes (such as the rise of the interfacial resistance in the low frequency band, the change of the diffusion tail in the medium frequency band, and the change of the high frequency arc, etc.), revealing the evolution mechanism of the reaction kinetics and mass transfer behavior inside the battery. The above parameter changes in the impedance spectrum are manifested as the response changes of the characteristic frequency bands. By comparing the impedance characteristics with the model output, the relationship between the changes of key parameters and the state of health is understood, realizing the comprehensive analysis of the evolution process from the time domain voltage response and the frequency domain impedance spectrum characteristics to the multi-physical mechanism, and improving the accuracy and physical interpretability of SOH prediction.

[0127] Example One:

[0128] The nominal capacity 10000mAh lithium ion battery with lithium iron phosphate material was cycled at 1C constant current (CC) protocol at 25°C ambient temperature, the charging cutoff voltage was 4.2V, and the discharging cutoff voltage was 2V. When the battery voltage exceeded the cutoff voltage, the charging and discharging program was terminated. The impedance spectrum measurement was measured by an electrochemical workstation at 100 SOC (state of charge) of the battery, the measurement frequency range was 0.1Hz-1000Hz, and the perturbation current amplitude was 0.5A.

[0129] The specific test results of Example One are as follows:

[0130] Figure 2 The experimental measurement and the model prediction of the battery at 0 state of charge (SOC) state under 1C rate CC charging process are shown. It can be seen that the model calculated voltage can maintain high consistency with the experimental voltage, and the voltage characteristics such as the voltage inflection point at the initial stage of charging and the voltage platform at the middle stage can be well predicted. Figure 3 The current density of SEI film thickening and lithium deposition reaction calculated by the model during the charging process is shown. SEI reaction occurs simultaneously with charging, while lithium deposition reaction occurs with a delay at the beginning of charging because the required reaction overpotential is less than 0V. Figure 4 ​The temperature curves of the battery during charging process, experimental measurement and model output. The model calculated temperature can well fit the actual battery temperature, verifying the accuracy of the model in predicting battery temperature in multiple physical fields. Figure 5 The electrochemical impedance spectra of experimental measurement and model calculation can be seen. The model calculated impedance can well simulate the impedance characteristics of the battery in each frequency range. Figure 6 The change of battery SOH with cycle number for experimental measurement and model calculation can be seen. In the early and middle stages of the cycle, the model can well predict the change of SOH. The maximum percentage error of the predicted SOH is not more than 1.2% throughout the cycle.

[0131] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... " does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0132] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, alternatives, and variations can be made in the embodiments without departing from the spirit and scope of the present application as defined by the appended claims and their equivalents.

Claims

1. A lithium battery SOH prediction method based on a multi-physical field coupling model, characterized in that, Comprising the following steps: S1: constructing an electrochemical model containing SEI film growth and lithium deposition; S2: combining the electrochemical model in S1 through coupling of thermal field and force field to form a lithium battery multi-physical field model; S3: combining the lithium battery multi-physical field model in S2 through coupling of SEI film thickening and lithium deposition side reactions to form a side reaction lithium loss model of coupled multi-physical fields, and using the side reaction lithium loss model of coupled multi-physical fields to predict the SOH of the lithium battery; S4: optimizing the parameters of the side reaction lithium loss model of coupled multi-physical fields in S3 to obtain a final side reaction lithium loss model; S5: constructing a model impedance spectrum for verifying the feasibility of the final side reaction lithium loss model of coupled multi-physical fields in S4 and assisting in diagnosis.

2. The lithium battery SOH prediction method based on a multi-physical field coupling model according to claim 1, characterized in that, The S1 comprises the following steps: S1.1, constructing a pseudo two-dimensional model as a basic electrochemical model of a lithium ion battery; S1.2, using a rate expression of diffusion-interface reaction cooperative control to construct a SEI film growth and lithium deposition side reaction flux model, wherein the SEI film thickening side reaction occurs on the negative electrode surface and is represented as: wherein, is the SEI reaction current density, is the SEI solid volume fraction, is the liquid phase concentration, is the SEI growth reaction rate constant, is the SEI growth reaction lithium ion transfer coefficient, is the electrolyte diffusivity in the SEI, is the SEI thickness, is the SEI layer overpotential, F is the Faraday constant, R is the universal gas constant, and T is the battery temperature; wherein the lithium deposition side reaction adopts a bidirectional kinetic model based on the Butler-Volmer equation, and the metal lithium deposition process is represented as: wherein, is the lithium plating reaction current density, which indicates the rate of the lithium plating process, is the overpotential of the lithium plating reaction, which only occurs below 0 V, and are the charge transfer coefficients of the anode and cathode, respectively, is the lithium deposition reaction reference current density.

3. The lithium battery SOH prediction method based on a multi-physical field coupling model according to claim 1, characterized in that, The S2 comprises the following steps: S2.1, by ohmic heat Q J , reaction heat Q M and entropy change term Q rev The temperature field distribution is solved, and the heat transfer process in the model is simplified as one-dimensional heat transfer along the thickness direction of the battery. The temperature change in the battery can be described by the following energy conservation equation: wherein, represents the equivalent density of the battery, C p is the equivalent specific heat capacity, is the equivalent thermal conductivity; The change of the battery temperature is not only affected by the internal heat source, but also affected by the heat exchange with the environment, and the boundary heat exchange can be represented as: wherein, is a heat exchange coefficient, is an ambient temperature, is a battery temperature; S2.2, feeding back the temperature change to the electrochemical model based on the Arrhenius equation to dynamically correct the changes of the side reaction rate and the electrochemical parameters, and the correction expression of the temperature related parameters is as follows: wherein T ref is a reference temperature, Y ref is a parameter value at the reference temperature, Y(T) indicates a temperature-dependent parameter, Ea is an activation energy corresponding to Y(T), and R is a universal gas constant; S2.3, calculating the lithium ion concentration field in the positive and negative active materials distributed along the thickness direction of the battery according to the electrochemical model, deriving the stress-strain distribution in the electrode interior due to the volume expansion effect caused by the insertion and extraction of lithium, and establishing the stress field distribution by combining the volume strain caused by the insertion of lithium ions with the thermal strain, which is represented as: wherein, is the total strain tensor, is the elastic strain tensor, is the diffusion strain tensor, is the thermal strain tensor.

4. The lithium battery SOH prediction method based on a multi-physical field coupling model according to claim 1, characterized in that, The S3 comprises the following steps: S3.1, further introducing the thermal field and force field coupling mechanism in the lithium battery multi-physical field model in S2, using the Arrhenius correction of the temperature on the reaction rate constant, the adjustment effect of the stress on the reaction overpotential, and the thermal-force double feedback correction of the diffusion coefficient, to dynamically capture the evolution characteristics of the SEI growth behavior with the changes of the environment and working conditions, which is specifically represented as: wherein, is the SEI growth reaction rate constant coupled with the temperature field, is the initial SEI growth reaction rate constant, is the SEI reaction activation energy, R is the universal gas constant, and T is the battery temperature; is the SEI reaction overpotential coupled with the force field, is the initial SEI reaction overpotential, β is the stress coupling coefficient, and σ h is the volume stress at the calculation point; is the EC diffusion coefficient in the SEI coupled with the force and temperature fields, denotes the initial EC diffusion coefficient in the SEI, E D denotes the EC diffusion activation energy, and γ is the stress feedback coefficient; S3.2, introducing the temperature corrected exchange current density term and the stress modulated overpotential term to construct a deposition reaction rate expression reflecting the dynamic changes of the negative electrode microenvironment and physical field; wherein, is the lithium deposition reaction reference exchange current density coupled to the temperature field, is the initial lithium deposition reaction reference exchange current density, is the lithium deposition reaction activation energy; is the lithium deposition reaction overpotential coupled to the force field, is the initial lithium deposition reaction overpotential, β is the stress coupling coefficient, σ h is the volume stress at the calculation point; S3.3, taking the current density caused by the SEI film thickening and lithium deposition reaction in the battery as a lithium loss source term, and performing time integration through Faraday's law to estimate the irreversible consumption of lithium, and then combining the rated capacity of the battery to establish a health state estimation formula facing the whole life cycle: where F is Faraday's constant, and is the time-varying SEI thickening and lithium deposition reaction current density, is the lithium loss term, is the battery rated capacity, is the effective area of the anode reaction.

5. The lithium battery SOH prediction method based on a multi-physical field coupling model according to claim 1, characterized in that, The S4 comprises the following steps: S4.1, classifying the parameters involved in the side reaction lithium loss model of coupled multi-physical fields in S3, including four categories of geometric parameters, electrochemical parameters, thermal parameters and mechanical parameters; S4.2, the parameter identification method based on the dual objective function is adopted, the voltage response in time domain and the impedance spectrum characteristics in frequency domain are considered jointly, the fitting error of two types of data is weighted and fused by constructing a comprehensive loss function, the global search inversion is carried out by using the particle swarm optimization algorithm, so as to obtain the optimal parameter combination and improve the consistency of model simulation output and experimental results; in the parameter identification process, the objective function constructed is as follows: wherein, is the objective function for parameter identification, θ is the set of parameters to be identified, λ is an error weighting factor to balance the influence of time-domain and frequency-domain data on the identification result, is the simulated voltage response of the model at time t i for the parameters θ, is the voltage response of the experiment at time t i for the parameters θ; is the simulated impedance value of the model at frequency ω j for the parameters θ, is the impedance value of the experiment at frequency ω j for the parameters θ. In order to minimize the loss function, the particle swarm algorithm is used to update the position of the particle in the multidimensional space by simulating the dual guidance of "individual experience" and "group optimal", so as to realize efficient search of the optimal parameter combination, and the particle position updating rule is as follows: The inertia weight and boundary constraint strategy are adopted to improve the global search ability and convergence efficiency of the particle swarm optimization algorithm; wherein the inertia weight is gradually reduced from the initial value 0.9 to 0.4 according to the linear decreasing rule, so that the particle swarm algorithm has strong global search ability in the early stage and strong local convergence ability in the later stage; the boundary constraint adopts a mixed strategy combining "rebound and clamping", so as to ensure that the parameters are always in the physical reasonable range; the particle swarm size is set to 30 to 50, so as to cover the parameter space while controlling the calculation burden; the maximum iteration number is set to 200 to 300, so as to ensure the convergence accuracy.

6. The lithium battery SOH prediction method based on a multi-physical field coupling model according to claim 1, characterized in that, The S5 includes the following steps: Through the final coupling multi-physical field side reaction lithium loss model in S4, the dynamic response data of the battery in different aging stages are collected, the battery is excited and analyzed by using EIS technology, that is, small amplitude sinusoidal current disturbance is applied at multiple set frequencies, and the corresponding voltage response signal is collected synchronously, so as to obtain the complex impedance Z(ω) of the battery at each frequency: wherein, and are the complex representation of the voltage and current of the battery at frequency f, respectively. By Fourier transform of time-domain voltage and current signals, the amplitude and phase information in the frequency domain is extracted, and the complete impedance spectrum data is obtained. The complex impedance at multiple frequency points is plotted into a Nyquist plot to visualize the trend of the electrochemical characteristics of the battery. Combined with the temperature field, electric field and concentration field of the battery, the changes of key parameters during the aging process are analyzed, including the interface reaction rate constant k, the interface resistance R film , the lithium ion diffusion coefficient D s , which reveals the evolution mechanism of the internal reaction kinetics and mass transfer behavior of the battery. The changes of the above parameters in the impedance spectrum are manifested as the response changes of the characteristic frequency bands. By comparing the impedance characteristics with the model output, the relationship between the changes of key parameters and the state of health is understood, and the comprehensive analysis from the time-domain voltage response and the frequency-domain impedance spectrum characteristics to the evolution process of the multi-physical mechanism is realized, which improves the accuracy and physical interpretability of the SOH prediction.

Images