Dynamic prediction method and system for SOE of lithium iron phosphate battery pack and life evaluation based on P2D model

Abstract

The present invention discloses a method and system for dynamic prediction of SOE and life assessment of lithium iron phosphate battery packs that integrates a P2D model, which relates to the field of battery management technology, including collecting operating data to establish a P2D model that couples multiple physical fields, performing dynamic prediction of SOE based on the P2D model, building a life assessment model in combination with a performance decay mechanism, and optimizing operating strategies. Based on historical operating data and real-time state parameters, the present invention generates dynamic prediction results of SOE for battery packs using a P2D model that couples multiple physical fields of electrochemistry, thermo-mechanics, and dynamic performance; analyzes the prediction results in combination with a long-term performance decay mechanism, evaluates the remaining life of the battery pack, and outputs the evaluation results. The accuracy of SOE prediction is improved, and the life assessment model is used to comprehensively reflect the law of performance changes, significantly improving operating efficiency and extending service life through real-time monitoring and optimization. The present invention can meet the needs of efficient and reliable battery management under complex working conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery management and life assessment, and in particular, relates to a method and system for dynamically predicting SOE and life assessment of a lithium iron phosphate battery pack by integrating a P2D model. Background Art

[0002] With the widespread application of lithium iron phosphate batteries in the energy storage sector, the demand for performance prediction and lifespan assessment is increasing. Accurate dynamic SOE (State of Energy) prediction and lifespan assessment can not only improve the operating efficiency of battery packs but also extend their service life, thereby reducing overall operating costs. However, existing technologies for dynamic SOE prediction and lifespan assessment of lithium iron phosphate battery packs still lack model accuracy, dynamic response capabilities, and multi-physics coupling analysis, making them difficult to meet the requirements for efficient management under complex operating conditions.

[0003] In the prior art, publication number CN119382299B discloses a method and system for intelligent control of photovoltaic energy storage battery pack operations. This patent collects the status data of each battery cell in the photovoltaic energy storage battery pack in real time, and uses an intelligent agent model to perform status data analysis and target predictive evaluation, thereby achieving refined control and parameter optimization of the single-cell charging and discharging actions. However, this technical solution mainly relies on the intelligent agent model for state prediction and lacks in-depth modeling of the multi-physical field coupling process within the battery. In particular, in the dynamic prediction of the SOE of lithium iron phosphate batteries, the interaction between electrochemical reactions and thermodynamic behaviors is not fully considered, which may lead to insufficient prediction accuracy. In addition, this solution pays less attention to battery life evaluation, making it difficult to fully reflect the performance degradation law of the battery pack during long-term operation.

[0004] Another patent, with publication number CN118446684B, discloses an artificial intelligence-based optimization method for chemical battery energy storage systems. This patent uses a machine learning algorithm to evaluate the status of recycled batteries, and through deep learning and optimization algorithms combined with active balancing and passive balancing technologies, it achieves overall optimization configuration of the battery pack and dynamic charge and discharge strategy adjustment. However, this technical solution focuses on the overall optimization management of the battery pack, and the dynamic prediction accuracy of the SOE of single cells is relatively limited, and no electrochemical model is introduced to accurately describe the internal state of the battery. In addition, although this solution can evaluate the remaining life of battery cells, it lacks in-depth analysis of the performance degradation mechanism during the long-term operation of the battery pack, making it difficult to achieve high-precision life assessment.

[0005] The above issues demonstrate that existing technologies for dynamic SOE prediction and lifespan assessment of lithium iron phosphate battery packs still have certain deficiencies in terms of multi-physics field coupling modeling, dynamic response capabilities, and long-term performance degradation analysis. Therefore, the present invention provides a system for dynamic SOE prediction and lifespan assessment of lithium iron phosphate battery packs that integrates a P2D model. This system aims to improve the accuracy of dynamic SOE prediction by introducing a P2D model that couples electrochemical, thermodynamic, and multi-physics fields. Furthermore, combined with analysis of long-term performance degradation mechanisms, it achieves more accurate lifespan assessment, thereby meeting the needs for efficient and reliable battery management under complex operating conditions. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for dynamic prediction of SOE and life assessment of lithium iron phosphate battery packs integrating P2D model, mainly to solve the deficiencies in the existing technology in multi-physics field coupling modeling, dynamic response capability and long-term performance degradation analysis.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model includes the following steps:

[0009] S1, collect basic operating data of the lithium iron phosphate battery pack through the sensor network arranged in the battery pack, and establish a dedicated communication link between the battery pack and the monitoring equipment;

[0010] S2, uploads the battery pack’s historical operating data, real-time status parameters, and environmental variables via a dedicated communication link;

[0011] S3, based on historical operating data and real-time state parameters, uses the electrochemical-thermal-dynamic multi-physics field coupled P2D model to generate the SOE dynamic prediction results of the battery pack;

[0012] S4 , analyzes the prediction results in combination with the long-term performance degradation mechanism, evaluates the remaining life of the battery pack, and outputs the evaluation results.

[0013] In step S1, the method for establishing the dedicated communication link is as follows:

[0014] S11, obtaining an identifier of a monitoring device connected to the battery pack;

[0015] S12, with the center point of the battery pack as the center and half of the maximum geometric dimension of the battery pack as the radius, define a monitoring area for communication coverage;

[0016] S13, determine whether there is a relay node in the monitoring area; if not, the monitoring device directly communicates with the battery pack main control unit to complete the establishment of a dedicated communication link; if it exists, proceed to step S14;

[0017] S14, with the center point of the battery pack as the center, a sub-area that includes the relay node and is concentric with the monitoring area is delineated as the relay coverage area; the monitoring device communicates with the battery pack main control unit through the relay node to complete the establishment of a dedicated communication link.

[0018] In step S3, the SOE dynamic prediction result is calculated by the P2D model of electrochemical-thermal-dynamic multi-physics field coupling, and the specific steps are as follows:

[0019] S31, collects the operating parameters of the lithium iron phosphate battery pack, including voltage, current, temperature, internal resistance and electrolyte concentration distribution;

[0020] S32, divide the battery into multiple micro-units, each containing a positive electrode region, a negative electrode region, and a separator region. Construct electrochemical reaction equations, heat conduction equations, and mechanical equilibrium equations, and define the electrochemical reaction rate, heat conduction coefficient, and mechanical stress distribution of each region to describe the electrochemical reaction process, heat transfer behavior, and mechanical deformation inside the battery, respectively.

[0021] S33, based on the collected operating parameter data, the electrochemical reaction equation, heat conduction equation and mechanical equilibrium equation inside the battery are discretized using the finite difference method;

[0022] S34, through the coupling of electrochemical reaction rate, thermal conductivity and mechanical stress distribution, a multi-physics field coupled P2D model inside the battery is formed;

[0023] "Coupling" refers to the mutual independence, interaction and correlation between different physical fields (electrochemical reactions, heat conduction, mechanical stress). Specifically:

[0024] The electrochemical reaction rate (involving electrochemical processes such as charge transfer and ion transport), thermal conductivity (heat generation and diffusion laws) and mechanical stress distribution (material deformation, structural stress, etc.) do not exist independently, but influence each other through physical laws.

[0025] S35, inputting the discretized multi-physics field coupled P2D model into the computing platform of the main control unit to complete the initialization of the P2D model;

[0026] The electrochemical reaction equation, heat conduction equation and mechanical equilibrium equation constitute a coupling model, and the coupling model here is the P2D model.

[0027] Discretization refers to the process of converting a continuous mathematical model (such as a partial differential equation describing a physical field, or a continuous space or time domain) into a discrete numerical model for numerical computation. The core of this process is to approximate the originally continuous physical field or mathematical relationship using a finite set of discrete points or cells.

[0028] S36, defines the initial state parameters of the battery pack, including SOC, SOH and current ambient temperature;

[0029] S37, based on the P2D model, calculates the energy change rate of the battery pack under different working conditions, and derives the dynamic change curve of SOE through the energy change rate;

[0030] S38, introduces a dynamic response algorithm, combines the current and voltage data collected in real time, and corrects the dynamic change curve of the SOE;

[0031] S39 outputs the corrected SOE dynamic change curve to the main control unit for guiding the charge and discharge management of the battery pack.

[0032] In step S4, the battery life assessment method includes the following steps:

[0033] S41, collecting capacity attenuation data of the battery pack at different cycle times, including capacity retention rate and internal resistance change rate;

[0034] S42, first collects capacity decay data of the battery pack at different cycle times, including capacity retention rate and internal resistance change rate, analyzes the performance decay mechanism of the battery pack in long-term operation, and identifies key influencing factors, including electrochemical side reactions, thermal runaway risk, and mechanical fatigue;

[0035] S43, based on the performance degradation mechanism, defines life assessment indicators, including cycle life, calendar life and safety life;

[0036] S44, using machine learning algorithms to train the collected decay data to generate a life assessment model;

[0037] S45, embedding the life assessment model into the life assessment module to predict the remaining life of the battery pack.

[0038] The machine learning algorithm is a support vector machine or a neural network algorithm, which is used to perform nonlinear fitting and prediction on capacity decay data.

[0039] The monitoring device and the main control unit transmit data via a dedicated communication link. The communication coverage of the dedicated communication link is an area with the center point of the battery pack as the center and a radius of half of the maximum geometric dimension of the battery pack. The SOE dynamic prediction module receives the output results of the P2D model and generates a dynamic SOE prediction result of the battery pack based on the result. The life assessment module evaluates the remaining life of the battery pack in combination with the long-term performance degradation mechanism and outputs the evaluation result.

[0040] When the finite difference method is used to discretize the electrochemical reaction equation, the heat conduction equation and the mechanical equilibrium equation, a numerical calculation method with a fixed step size is adopted.

[0041] In the S32, the electrochemical reaction equation, heat conduction equation and mechanical equilibrium equation are constructed:

[0042] The electrochemical reaction process is modeled through electrochemical field, including the lithium ion solid phase diffusion equation, liquid phase transport equation and Butler-Volmer equation;

[0043] Lithium ion solid phase diffusion equation:

[0044] ,

[0045] The boundary conditions include the particle centrosymmetry:

[0046] ,

[0047] Surface reaction flux:

[0048] ,

[0049] Where t represents time, c s is the solid phase lithium concentration, D s is the solid phase diffusion coefficient, j is the electrode reaction current density, F is the Faraday constant, r is the radial coordinate in the spherical coordinate system, R s represents the radius of solid lithium particles;

[0050] The liquid phase transport equation includes the conservation equation for the coupling of lithium ion concentration and potential in the electrolyte and the charge conservation equation. The conservation equation for the coupling of lithium ion concentration and potential in the electrolyte is:

[0051] ,

[0052] The charge conservation equation is:

[0053] ,

[0054] Among them, ▽ is the vector differential operator, ce is the liquid lithium concentration, ε e is the electrolytic liquid volume fraction, is the effective liquid diffusion coefficient, is the lithium ion migration number, is the liquid phase potential, K eff 、 is the effective conductivity of liquid lithium and the diffusion conductivity of liquid lithium;

[0055] The Butler-Volmer equation first describes the electrode / electrolyte interface reaction kinetics:

[0056] ,

[0057] Exchange current density j0 expression:

[0058] ,

[0059] The overpotential η is defined as:

[0060] ,

[0061] Among them, α a , α c are the anode and cathode transfer coefficients, U ocp is the open circuit potential, is the surface concentration of solid lithium, is the solid phase potential; R is the gas constant, T is the absolute temperature, and k is the intrinsic reaction rate constant;

[0062] The heat transfer behavior is represented by thermodynamic field coupling, including the energy conservation equation and temperature-sensitive parameter correction. The total heat generation rate Q of the energy conservation equation includes:

[0063] Joule heat: ,

[0064] Reaction heat: ,

[0065] Polarization heat: ,

[0066] Then the heat conduction equation:

[0067] ,

[0068] in, is the material density, is the specific heat capacity at constant pressure, represents volume heat capacity, λ is thermal conductivity, is the effective conductivity of solid-phase lithium;

[0069] The temperature-sensitive parameter correction is expressed by the temperature dependence of the diffusion coefficient and the conductivity:

[0070] ,

[0071] Where D represents the diffusion coefficient, D(T) represents the diffusion coefficient at temperature T, Indicates the reference temperature The diffusion coefficient under a is the activation energy, T ref is the reference temperature;

[0072] Mechanical equilibrium equation: Calculate the internal stress of particles based on the solid phase diffusion equation and the elastic mechanics equation :

[0073] ,

[0074] Where E is the elastic modulus, c s Solid phase lithium concentration, c0 is the initial solid phase lithium concentration, represents the maximum intercalation concentration of solid-phase lithium, It is the relative deformation of the material under stress or environmental changes.

[0075] In said S32, defining the electrochemical reaction rate, thermal conductivity and mechanical stress distribution of each region means:

[0076] The electrochemical reaction rate was described by the Butler-Volmer equation to describe the electrode / electrolyte interface reaction kinetics;

[0077] The thermal conductivity is determined by the positive and negative electrode regions and the diaphragm region. The positive and negative electrode regions are composed of active materials, conductive agents, binders and pore electrolytes. Their equivalent thermal conductivity w eff Calculated by the mixed model:

[0078] ,

[0079] Among them, μ i is the volume fraction of each component, w i is the intrinsic thermal conductivity of the material;

[0080] The energy conservation equation defines the source of heat generation, while the heat conduction equation describes the transfer and accumulation of heat. The heat conduction equation is the "heat transfer end," directly linking heat generation Q to the temperature distribution T within the battery, reflecting how heat is transferred through conduction and causes temperature changes. Temperature-sensitive correction establishes a bidirectional coupling. As temperature T increases, the diffusion coefficient D and conductivity increase (due to an increase in the exp term), leading to an increase in the electrochemical reaction rate (increase in j) and a change in Joule heating Qohm (due to the correlation between σeff and κeff and conductivity). These parameter changes, in turn, affect the heat generation rate Q (for example, an increase in j leads to increases in Qrxn and Qpol), which in turn changes the temperature distribution T through the heat conduction equation.

[0081] Describe the heat transfer behavior:

[0082] Heat transfer behavior is a dynamic process of "heat generation-conduction-accumulation", which can be divided into the following three steps:

[0083] 1. Heat generation stage (dominated by energy conservation equation)

[0084] When the battery is working, electrochemical reaction (reaction heat Qrxn), current resistance (Joule heat Qohm) and entropy change (polarization heat Qpol) generate heat simultaneously, and the total heat generation rate Q is the sum of these three parts. For example:

[0085] When large current is discharged, Joule heat Qohm increases significantly;

[0086] At low temperatures, the reaction rate decreases (j decreases), but when the activation energy Ea is high, the polarization heat Qpol may dominate the heat generation.

[0087] 2. Heat conduction stage (dominated by the heat conduction equation)

[0088] Heat generation Q causes local temperature to rise, forming a temperature gradient Heat is transferred through heat conduction Transfer from high temperature area (such as the center of the electrode) to low temperature area (such as the battery surface). For example:

[0089] When the thermal conductivity λ inside the battery is low (such as the separator material), heat is difficult to diffuse and local hot spots are easily formed;

[0090] In liquid-cooled batteries, the cooling medium increases the , accelerating heat dissipation.

[0091] 3. Parameter feedback stage (dominated by temperature-sensitive parameter correction)

[0092] The change in temperature T corrects the diffusion coefficient D and the conductivity using the Arrhenius equation:

[0093] Temperature rises → D increases → lithium ion diffusion accelerates → reaction current j increases → Qrxn and Qpol increase → temperature rises further (positive feedback);

[0094] Temperature decreases → D decreases → diffusion hysteresis → lithium ion concentration gradient increases on the electrode surface → overpotential η increases → Qrxn increases (which may exacerbate local overheating at low temperatures). The diaphragm region is dominated by material properties;

[0095] The distribution of mechanical stress is jointly determined by the positive and negative electrode regions and the diaphragm region; among them, the sources of stress in the positive and negative electrode regions are: lithium ion insertion / extraction causes the volume change of active particles, inducing local stress; the sources of stress in the diaphragm region are external pressure conduction and thermomechanical stress; external pressure conduction: the battery packaging pressure is transmitted to the diaphragm through the electrode, and its compression resistance must be considered; thermomechanical stress: temperature changes cause differences in the thermal expansion coefficients of the diaphragm and the electrode, resulting in interfacial shear stress.

[0096] Furthermore, implementation scenarios of the system include power battery management systems for electric vehicles or lithium iron phosphate battery pack management in photovoltaic energy storage systems.

[0097] A P2D model-integrated SOE dynamic prediction and life assessment system for a lithium iron phosphate battery pack includes a lithium iron phosphate battery pack, a monitoring device communicating with the lithium iron phosphate battery pack via a dedicated communication link, a main control unit connected to the monitoring device, a P2D model-integrated SOE dynamic prediction module connected to the main control unit, and a life assessment module connected to the main control unit.

[0098] Compared with the prior art, the present invention has the following beneficial effects:

[0099] (1) The present invention accurately describes the internal state of the lithium iron phosphate battery pack by introducing the P2D model of electrochemical-thermal-mechanical multi-physics field coupling. The traditional P2D model only focuses on the electrochemical process, while this method innovatively integrates thermodynamics (temperature field) and mechanical (stress / strain) analysis to more comprehensively reflect the changes in the internal state of the battery. The P2D model can describe the electrochemical reaction process, heat conduction behavior and mechanical stress distribution inside the battery in detail, thereby improving the accuracy of SOE dynamic prediction. At the same time, the present invention combines the long-term performance degradation mechanism of the battery pack to construct a life assessment model that can comprehensively reflect the performance change law of the battery pack under complex working conditions. In addition, by real-time monitoring and adjustment and optimization of the battery pack's operating strategy, the operating efficiency of the battery pack is significantly improved and its service life is extended. The technical means of the present invention are specific and clear, and can meet the needs of efficient and reliable battery management under complex working conditions.

[0100] (2) The present invention introduces a dynamic response algorithm into the dynamic prediction process of SOE, and realizes closed-loop prediction of SOE through the dynamic response algorithm (such as data-driven model update or adaptive filtering), thereby solving the error accumulation problem caused by parameter aging or sudden change of working conditions in traditional models.

[0101] The dynamic SOE curve calculated using the P2D model is corrected by combining real-time current and voltage data. Compared to the shortcomings of existing technologies in terms of dynamic response, this approach can promptly capture real-time changes during battery operation and quickly adjust the prediction results, making the dynamic SOE prediction more closely aligned with actual operating conditions. This significantly enhances the system's dynamic response capabilities, providing strong support for more efficient charge and discharge management of the battery pack and improving its operating efficiency.

[0102] (3) The present invention collects capacity decay data of battery packs at different cycle times, deeply analyzes the performance decay mechanism of battery packs in long-term operation, identifies key influencing factors such as electrochemical side reactions, thermal runaway risks, and mechanical fatigue, and defines life assessment indicators such as cycle life, calendar life, and safety life based on this. The decay data is trained using machine learning algorithms such as support vector machines or neural network algorithms to generate a life assessment model. Compared with the existing technology that lacks in-depth analysis of long-term performance decay mechanisms and has difficulty in achieving high-precision life assessment, the present invention can more comprehensively and accurately assess the remaining life of the battery pack, providing an important reference for the full life cycle management of the battery pack, helping to extend the service life of the battery pack and reduce overall operating costs.

[0103] (4) Combining SOE dynamic prediction with life assessment, by analyzing the cumulative effects of energy loss (such as cycle aging and calendar aging), differentiated management of battery packs (such as balanced charge and discharge, abnormal cell warning) can be achieved. Traditional life assessment relies on fixed thresholds (such as the number of cycles), while this method captures the attenuation mechanism under actual operating conditions (such as lithium dendrite growth and electrolyte decomposition) through a multi-physics field coupling model.

[0104] In summary, the multi-physics coupling model comprehensively considers the cross-influence of electrochemistry, heat and mechanics, avoids the limitations of a single physics model, and significantly improves the SOE prediction accuracy (especially under high-rate charge and discharge and wide temperature range conditions).

[0105] Through dynamic correction algorithms, the model can quickly adapt to changes in operating conditions (such as sudden high currents and temperature fluctuations), reduce prediction delays, and is suitable for scenarios with high real-time requirements such as autonomous driving and energy storage power stations.

[0106] SOE dynamic prediction and life assessment share multi-physics model parameters to avoid repeated modeling. At the same time, aging factor decomposition (such as distinguishing the contributions of cycle aging and calendar aging) is achieved through energy loss analysis, providing data support for battery pack maintenance.

[0107] The SOE-based charge and discharge strategy can avoid deep charge and discharge (DOD), reduce the risk of electrode stress accumulation, and inhibit lithium dendrite phenomenon; temperature field optimization reduces the risk of thermal runaway and improves safety.

[0108] The model framework is applicable to lithium iron phosphate battery packs of different specifications. It only requires adjusting the micro-unit division density and parameter calibration, and has the potential for engineering application. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] FIG1 is a block diagram of the principle of the present invention.

[0110] FIG2 is a flow chart of the system implementation of the present invention. DETAILED DESCRIPTION

[0111] The present invention will be further described below with reference to the accompanying drawings and examples. The embodiments of the present invention include but are not limited to the following examples.

[0112] Example 1

[0113] As shown in Figure 1, the present invention provides a P2D model-integrated SOE dynamic prediction and life assessment system for lithium iron phosphate battery packs, comprising a lithium iron phosphate battery pack, a monitoring device communicating with the lithium iron phosphate battery pack via a dedicated communication link, a main control unit connected to the monitoring device, a P2D model-integrated SOE dynamic prediction module connected to the main control unit, and a life assessment module connected to the main control unit.

[0114] The lithium iron phosphate battery pack is the core component of the entire system, storing and releasing electrical energy. The monitoring device connects to the battery pack via a physical interface, collecting basic operating data in real time, including voltage, current, temperature, and other parameters. The monitoring device also exchanges data with the main control unit via a dedicated communication link. The dedicated communication link is established as follows: First, the monitoring device's identifier is obtained. Then, a monitoring area is defined for communication coverage, centered at the center of the battery pack and with a radius equal to half the pack's maximum geometric dimension. If there are no relay nodes within the monitoring area (relay nodes are commonly used in wireless communications to forward signals and expand network coverage), the monitoring device communicates directly with the main control unit. If a relay node is present, a subarea containing the relay node and concentric with the monitoring area is further defined as the relay coverage area. The monitoring device establishes a communication link with the main control unit via the relay node. This communication link design ensures stable and real-time data transmission. There are many ways to determine whether there is a relay node in the monitoring area, such as detecting whether there is an area with abnormally stable signal strength or analyzing the protocol type in the data packet. Some relay devices will retain specific protocol headers.

[0115] The main control unit receives historical operating data and real-time status parameters from monitoring equipment and transmits this data to the P2D model for processing. As shown in Figure 2, the data acquisition process for the lithium iron phosphate battery pack begins with the sensor network deployed within the pack, acquiring operating parameters such as voltage, current, temperature, internal resistance, and electrolyte concentration distribution. These operating parameters are transmitted to the main control unit's computing platform for subsequent modeling and analysis. During the P2D model construction process, the battery interior is divided into multiple microcells, each containing a positive electrode region, a negative electrode region, and a separator region. Ion conduction between the positive and negative electrode regions is achieved through the separator region, which also serves to prevent direct electron flow. Within each microcell, the electrochemical reaction rate, thermal conductivity coefficient, and mechanical stress distribution are defined. These parameters describe the electrochemical reaction process, heat transfer behavior, and mechanical deformation within the battery, respectively. Based on the acquired operating parameters, the finite difference method is used to discretize the electrochemical reaction equations, heat conduction equations, and mechanical equilibrium equations, thereby forming a multi-physics coupled model of the battery interior. After the model is initialized through the computing platform, it can accurately characterize the multi-physical field coupling relationship inside the battery.

[0116] In the SOE dynamic prediction module, the initial state parameters of the battery pack are defined, including SOC (State of Charge), SOH (State of Health), and the current ambient temperature. Based on the established P2D model, the energy change rate of the battery pack under different operating conditions is calculated, and the SOE dynamic curve is derived. To improve prediction accuracy, a dynamic response algorithm is introduced to correct the SOE dynamic curve using real-time current and voltage data. The corrected SOE dynamic curve is output to the monitoring system to guide battery pack charge and discharge management. In this process, the P2D model provides fundamental support for SOE dynamic prediction. By accurately describing the changes in electrochemical reaction rates, thermal conductivity, and mechanical stress distribution within the battery, the accuracy of the SOE prediction results is ensured.

[0117] The life assessment module is constructed based on capacity decay data of battery packs at different cycle times, including capacity retention and internal resistance change rate. By analyzing the performance decay mechanisms during long-term operation, key factors influencing battery life, such as electrochemical side reactions, thermal runaway risk, and mechanical fatigue, are identified. Based on these performance decay mechanisms, life assessment metrics such as cycle life, calendar life, and safety life are defined. A machine learning algorithm is used to train the collected decay data to generate a life assessment model. The machine learning algorithm, typically a support vector machine or neural network algorithm, is used to perform nonlinear fitting and prediction of capacity decay data. This model is embedded in the life assessment module to predict the remaining life of the battery pack. During this process, the life assessment module and the P2D model work together, with the former relying on the multi-physics coupling information provided by the latter, and the latter optimizing its predictive capabilities through feedback from the former.

[0118] In practical applications, this system can be deployed in the power battery management system of electric vehicles. For example, while the vehicle is in operation, monitoring equipment collects real-time voltage, current, and temperature data from the lithium iron phosphate battery pack and transmits this data to the main control unit via a dedicated communication link. The main control unit inputs this data into the P2D model, which generates a dynamic state of equilibrium (SOE) prediction using multi-physics coupling analysis and transmits this result to the SOE dynamic prediction module. Simultaneously, the lifespan assessment module evaluates the remaining life of the battery pack by incorporating long-term performance degradation mechanisms. This assessment is fed back to the vehicle control system, allowing the driver to promptly understand the battery status and take appropriate measures.

[0119] This system also has broad application value in industrial energy storage scenarios. For example, in photovoltaic energy storage systems, lithium iron phosphate battery packs, serving as energy storage devices, require frequent charging and discharging. Monitoring equipment collects real-time operating data from the battery packs and transmits it to the main control unit via a dedicated communication link. The main control unit uses a P2D model to dynamically predict the battery pack's status and, in conjunction with a lifespan assessment module, estimates the remaining lifespan of the battery packs. This real-time monitoring and assessment mechanism effectively extends the battery pack's service life and improves system efficiency.

[0120] This system accurately models the complex multi-physics interactions within batteries by introducing a P2D model that couples electrochemical, thermodynamic, and multi-physics fields. Data is transmitted between the monitoring equipment and the battery pack's main control unit via a dedicated communication link, ensuring real-time and accurate data acquisition. The P2D model combines historical operating data with real-time status parameters to generate dynamic SOE predictions. Its simple and efficient structural design enables faster convergence and higher prediction accuracy during model training. Through in-depth analysis of the long-term performance degradation mechanisms of battery packs, more accurate lifespan assessments are achieved, meeting the needs of efficient and reliable battery management under complex operating conditions.

[0121] Example 2

[0122] In order to better enable relevant personnel in this technical field to fully understand and implement the present invention, the specific implementation method of the present invention is further described below in conjunction with a specific application scenario.

[0123] A method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model includes the following steps:

[0124] S1 collects basic operating data from the lithium iron phosphate battery pack through a network of sensors (monitoring devices) deployed in the battery pack. A dedicated communication link is established between the battery pack and the monitoring device sensors. The monitoring device connects to the lithium iron phosphate battery pack via a physical interface and collects basic operating data in real time, including parameters such as voltage, current, temperature, internal resistance, and electrolyte concentration distribution. This data is transmitted to the main control unit via a dedicated communication link.

[0125] The method for establishing a dedicated communication link is as follows:

[0126] S11, obtaining an identifier of a monitoring device connected to the battery pack;

[0127] S12, with the center point of the battery pack as the center and half of the maximum geometric dimension of the battery pack as the radius, define a monitoring area for communication coverage;

[0128] S13, determine whether there is a relay node in the monitoring area; if not, the monitoring device directly communicates with the battery pack main control unit to complete the establishment of a dedicated communication link; if it exists, proceed to step S14;

[0129] S14, with the center point of the battery pack as the center, a sub-area that includes the relay node and is concentric with the monitoring area is delineated as the relay coverage area; the monitoring device communicates with the battery pack main control unit through the relay node to complete the establishment of a dedicated communication link.

[0130] S2, uploads the battery pack’s historical operating data, real-time status parameters, and environmental variables via a dedicated communication link;

[0131] S3, based on historical operating data and real-time state parameters, uses the electrochemical-thermal-dynamic multi-physics field coupled P2D model to generate the SOE dynamic prediction results of the battery pack;

[0132] The SOE dynamic prediction results are calculated using the P2D model that couples electrochemical, thermal, and mechanical multi-physics fields. The specific steps are as follows:

[0133] S31, collects the operating parameters of the lithium iron phosphate battery pack, including voltage, current, temperature, internal resistance and electrolyte concentration distribution;

[0134] S32, divide the battery into multiple micro-units, each containing a positive electrode region, a negative electrode region, and a separator region. Construct electrochemical reaction equations, heat conduction equations, and mechanical equilibrium equations, and define the electrochemical reaction rate, heat conduction coefficient, and mechanical stress distribution of each region to describe the electrochemical reaction process, heat transfer behavior, and mechanical deformation inside the battery, respectively.

[0135] S33, based on the collected operating parameter data, the electrochemical reaction equation, heat conduction equation and mechanical equilibrium equation inside the battery are discretized using the finite difference method;

[0136] When the finite difference method is used to discretize the electrochemical reaction equation, the heat conduction equation and the mechanical equilibrium equation, a numerical calculation method with a fixed step size is adopted.

[0137] S34, by coupling the electrochemical reaction rate, thermal conductivity and mechanical stress distribution, a multi-physics field coupled P2D model inside the battery is formed;

[0138] S35, inputting the discretized multi-physics field coupled P2D model into the computing platform of the main control unit to complete the initialization of the P2D model;

[0139] S36, defines the initial state parameters of the battery pack, including SOC, SOH and current ambient temperature;

[0140] S37, based on the P2D model, calculates the energy change rate of the battery pack under different working conditions, and derives the dynamic change curve of SOE through the energy change rate;

[0141] S38, introduces a dynamic response algorithm, combines the current and voltage data collected in real time, and corrects the dynamic change curve of the SOE;

[0142] S39 outputs the corrected SOE dynamic change curve to the main control unit for guiding the charge and discharge management of the battery pack.

[0143] S4 , analyzes the prediction results in combination with the long-term performance degradation mechanism, evaluates the remaining life of the battery pack, and outputs the evaluation results.

[0144] In step S4, the battery life assessment method includes the following steps:

[0145] S41, collecting capacity attenuation data of the battery pack at different cycle times, including capacity retention rate and internal resistance change rate;

[0146] S42, first collects capacity decay data of the battery pack at different cycle times, including capacity retention rate and internal resistance change rate, analyzes the performance decay mechanism of the battery pack in long-term operation, and identifies key influencing factors, including electrochemical side reactions, thermal runaway risk, and mechanical fatigue;

[0147] S43, based on the performance degradation mechanism, defines life assessment indicators, including cycle life, calendar life and safety life;

[0148] S44, using machine learning algorithms to train the collected decay data to generate a life assessment model;

[0149] S45, embedding the life assessment model into the life assessment module to predict the remaining life of the battery pack.

[0150] The machine learning algorithm is a support vector machine or a neural network algorithm, which is used to perform nonlinear fitting and prediction on capacity attenuation data.

[0151] Data is transmitted between the monitoring device and the main control unit via a dedicated communication link. The communication coverage of the dedicated communication link is an area with the center point of the battery pack as the center and half of the maximum geometric dimension of the battery pack as the radius; the SOE dynamic prediction module receives the output results of the P2D model and generates a dynamic SOE prediction result of the battery pack based on the result; the life assessment module evaluates the remaining life of the battery pack in combination with the long-term performance degradation mechanism and outputs the evaluation result.

[0152] When building a P2D model, the battery's interior is first divided into multiple microcells. Each microcell contains a positive electrode region, a negative electrode region, and a separator region. Ion conduction between the positive and negative electrode regions occurs through the separator region, while the separator region also prevents direct electron flow.

[0153] Construct electrochemical reaction equations, heat conduction equations and mechanical equilibrium equations, define the electrochemical reaction rate, heat conduction coefficient and mechanical stress distribution for each microunit, and describe the electrochemical reaction process, heat transfer behavior and mechanical deformation inside the battery respectively.

[0154] Among them, the electrochemical reaction process is modeled through electrochemical field, including the lithium ion solid phase diffusion equation, liquid phase transport equation and Butler-Volmer equation.

[0155] First, the conservation equations of the internal electrodes (positive / negative electrodes) and electrolytes of lithium-ion batteries are established based on the Newman framework, including the lithium ion solid phase diffusion equation:

[0156] ,

[0157] The boundary conditions include the particle centrosymmetry:

[0158] ,

[0159] Surface reaction flux:

[0160] ,

[0161] Among them, c s is the solid phase lithium concentration, D s is the solid phase diffusion coefficient, j is the electrode reaction current density, F is the Faraday constant, r is the radial coordinate in the spherical coordinate system.

[0162] The liquid phase transport equation includes the conservation equation for the coupling of lithium ion concentration and potential in the electrolyte and the charge conservation equation. The conservation equation for the coupling of lithium ion concentration and potential in the electrolyte is:

[0163] ,

[0164] The charge conservation equation is:

[0165] ,

[0166] Among them, ▽ is the vector differential operator, c e is the liquid lithium concentration, ε e is the electrolytic liquid volume fraction, is the effective liquid diffusion coefficient, is the lithium ion migration number, is the liquid phase potential, K eff 、 is the effective conductivity of liquid lithium and the diffusion conductivity of liquid lithium;

[0167] The Butler-Volmer equation first describes the electrode / electrolyte interface reaction kinetics:

[0168] ,

[0169] Exchange current density j0 expression:

[0170] ,

[0171] The overpotential η is defined as:

[0172] ,

[0173] Among them, α a , α c are the anode and cathode transfer coefficients, U ocp is the open circuit potential, is the surface concentration of solid lithium, is the solid phase potential. R is the gas constant, T is the absolute temperature, and k is the intrinsic reaction rate constant;

[0174] The heat transfer behavior is represented by thermodynamic field coupling, including the energy conservation equation and temperature-sensitive parameter correction. The total heat generation rate Q of the energy conservation equation includes:

[0175] Joule heat: ,

[0176] Reaction heat: ,

[0177] Polarization heat: ,

[0178] Then the heat conduction equation:

[0179] ,

[0180] in, is the material density, is the specific heat capacity at constant pressure, represents volume heat capacity, λ is thermal conductivity, is the effective conductivity of solid-phase lithium;

[0181] The temperature-sensitive parameter correction is expressed by the temperature dependence of the diffusion coefficient and the conductivity:

[0182] ,

[0183] Where D represents the diffusion coefficient, D(T) represents the diffusion coefficient at temperature T, Indicates the reference temperature The diffusion coefficient under a is the activation energy, T ref is the reference temperature.

[0184] Mechanical equilibrium equation: Calculate the internal stress of particles based on the solid phase diffusion equation and the elastic mechanics equation :

[0185] ,

[0186] Where E is the elastic modulus, c s Solid phase lithium concentration, c0 is the initial solid phase lithium concentration, represents the maximum intercalation concentration of solid-phase lithium, It is the relative deformation of the material under stress or environmental changes.

[0187] In S32, the electrochemical reaction rate, thermal conductivity, and mechanical stress distribution of each region are defined as follows:

[0188] The electrochemical reaction rate was described by the Butler-Volmer equation to describe the electrode / electrolyte interface reaction kinetics;

[0189] The thermal conductivity is determined by the positive and negative electrode regions and the diaphragm region. The positive and negative electrode regions are composed of active materials, conductive agents, binders and pore electrolytes. The equivalent thermal conductivity w eff Calculated by the mixed model:

[0190] ,

[0191] Where μ i is the volume fraction of each component, w i is the intrinsic thermal conductivity of the material (e.g., about 150 W / (m·K) for graphite anode and about 1.5 W / (m·K) for lithium iron phosphate cathode).

[0192] The diaphragm region is dominated by material properties: the diaphragm is typically a microporous polyethylene (PE) or polypropylene (PP) membrane with low thermal conductivity (approximately 0.3–0.5 W / (m·K)), and the closed-pore effect (pore closure at high temperatures leading to increased thermal resistance) must be considered.

[0193] The distribution of mechanical stress is also determined by the positive and negative electrode regions and the diaphragm region. Among them, the stress sources in the positive and negative electrode regions are: lithium ion insertion / extraction causes the volume change of active particles (such as the expansion rate of graphite is about 10%, and the silicon-based material can reach 300%), which causes local stress. The stress sources in the diaphragm region are external pressure conduction and thermomechanical stress. External pressure conduction: The battery packaging pressure is transmitted to the diaphragm through the pole piece, and its compression resistance needs to be considered. Thermomechanical stress: Temperature changes lead to differences in the thermal expansion coefficients of the diaphragm and the electrode (such as the expansion coefficient of the aluminum collector is 23×10⁻ 6 / K, PE diaphragm about 200×10⁻ 6 / K), generating interfacial shear stress.

[0194] Based on the collected operating parameters, the finite difference method is used to discretize the electrochemical reaction equations, heat conduction equations, and mechanical equilibrium equations, forming a P2D model of the multi-physics field coupling within the battery. The solid-phase diffusion equation uses the implicit Euler method to discretize the time term or the central difference method to discretize the spatial term. The nonlinear terms of the Butler-Volmer equation are linearized using the Newton iteration method.

[0195] After initialization on the computing platform, the model accurately depicts the multi-physics coupling relationship within the battery. Through this process, the P2D model achieves a highly accurate description of the battery's internal state, laying the foundation for subsequent SOE dynamic prediction and lifespan assessment.

[0196] When predicting SOE dynamically, the battery pack's initial state parameters, including SOC (State of Charge), SOH (State of Health), and current ambient temperature, are first defined. Based on the established P2D model, the battery pack's energy change rate under different operating conditions is calculated, and the SOE dynamic curve is derived. To improve prediction accuracy, a dynamic response algorithm is introduced to correct the SOE dynamic curve using real-time current and voltage data. This corrected SOE dynamic curve is output to the monitoring system to guide battery pack charge and discharge management. During this process, the P2D model ensures the accuracy of SOE predictions by accurately describing electrochemical reaction rates, thermal conductivity, and mechanical stress distribution. For example, when an electric bus accelerates, the battery pack's current demand suddenly increases. The P2D model can quickly respond and adjust the SOE dynamic curve, thereby preventing performance degradation caused by overcharging or over-discharging.

[0197] During the lifespan assessment process, we first collect capacity decay data for the battery pack at different cycle times, including capacity retention and internal resistance change rate. By analyzing the performance decay mechanism during long-term operation, we identify key factors affecting battery life, such as electrochemical side reactions, thermal runaway risk, and mechanical fatigue.

[0198] 1. Performance degradation driven by electrochemical side reactions

[0199] 1. Collapse of cathode material structure

[0200] Cathode materials (such as ternary lithium and lithium iron phosphate) undergo repeated lithium ion insertion and extraction during charge and discharge, resulting in volume expansion and contraction. Long-term stress accumulation can cause the material's lattice to collapse, blocking lithium ion insertion pathways and leading to capacity degradation. For example, lithium cobalt oxide can decompose into cobalt oxide at high temperatures, releasing oxygen, further exacerbating structural damage.

[0201] 2. Lithium plating on negative electrode and SEI film thickening

[0202] Under low temperature or overcharge conditions, lithium ion insertion into the negative electrode slows down, leading to the precipitation of metallic lithium on the negative electrode surface. This precipitation not only consumes active lithium but also punctures the separator, causing internal short circuits. Furthermore, the SEI (solid electrolyte interface) membrane, which repeatedly breaks and repairs during cycling, increases in thickness, hindering lithium ion transport and increasing internal resistance.

[0203] 3. Electrolyte decomposition and conductive salt degradation

[0204] Organic solvents in the electrolyte, such as carbonates, undergo oxidative decomposition at high temperatures or voltages, producing flammable gases such as hydrogen and ethylene, as well as unstable intermediates. Conductive salts such as LiPF6 degrade to HF acid, which corrodes electrode materials and accelerates capacity decay. The accumulation of decomposition products can also clog the pores of the separator and reduce ion mobility.

[0205] 2. Thermal runaway risk and chain reaction

[0206] 1. Thermal runaway trigger mechanism

[0207] Internal short circuits (such as lithium plating piercing the diaphragm), overcharge, overdischarge, or mechanical damage can cause the local heat generation rate to exceed the heat dissipation capacity, triggering the following chain reaction:

[0208] 2.90-150℃ stage: SEI film decomposes, and the electrolyte reacts with the negative electrode to release heat.

[0209] 3.150-250℃ stage: The diaphragm melts and causes a larger-scale short circuit, and the positive electrode material decomposes and releases oxygen.

[0210] 4.>250℃ stage: The electrolyte burns violently, the battery shell ruptures and releases combustible gas, causing an explosion.

[0211] 5. Heat diffusion and module-level loss of control

[0212] The heat generated by thermal runaway in a single cell is transferred to adjacent cells through conduction, convection, and radiation, creating a "domino effect" of thermal runaway propagation. For example, the propagation speed of thermal runaway in a lithium iron phosphate battery pack can reach 5-10 cm / min.

[0213] 3. Mechanical fatigue and material degradation

[0214] Current collector corrosion and mechanical failure

[0215] During overdischarge (voltage > 1.5V), the copper current collector oxidizes and dissolves, and the deposited copper metal damages the negative electrode structure. Aluminum current collectors are corroded by HF in the electrolyte, resulting in decreased conductivity. During long-term cycling, interfacial delamination between the current collector and the active material can also increase internal resistance.

[0216] Microcracks in electrode material particles

[0217] During the charging and discharging process, the electrode material repeatedly expands and contracts (for example, the volume of the silicon-based negative electrode changes by up to 300%), resulting in particle breakage, increased contact resistance, and rapid capacity decay.

[0218] Structural support fatigue

[0219] The fixing brackets and connectors of the battery module are prone to stress fatigue due to vibration and temperature cycles, which may cause the battery cells to shift and the connections to loosen, leading to local overheating or short circuit.

[0220] Based on these performance degradation mechanisms, life assessment indicators such as cycle life, calendar life, and safe life are defined. Machine learning algorithms are used to train the collected degradation data to generate a life assessment model.

[0221] This process generally uses existing technology and is implemented through software algorithms. The general process includes:

[0222] 1. Data collection and cleaning

[0223] 2. Feature extraction

[0224] 3. Data Labeling

[0225] 4. Model selection and training

[0226] 5. Model Validation and Evaluation

[0227] 6. Model deployment and online update

[0228] This model is embedded in the monitoring system to predict the remaining life of the battery pack. During this process, the life assessment model and the P2D model work together. The former relies on the multi-physics coupling information provided by the latter, while the latter uses the former's feedback to optimize its predictive capabilities. For example, after long periods of operation in electric buses, the battery pack may experience capacity degradation. The life assessment module can predict the remaining life based on historical data and provide a basis for subsequent maintenance.

[0229] During actual operation, when the vehicle is in high-load conditions such as acceleration or climbing a slope, the monitoring equipment collects real-time voltage, current, and temperature data from the lithium iron phosphate battery pack and transmits it to the main control unit via a dedicated communication link. The main control unit uses a P2D model to dynamically predict the battery pack's status and, in conjunction with the SOE dynamic prediction module, generates the current SOE value. Simultaneously, the lifespan assessment module evaluates the battery pack's remaining life based on long-term performance degradation mechanisms and feeds the results back to the vehicle control system. The driver can view the battery status in real time on the onboard display and take appropriate actions based on the prompts, such as adjusting driving mode or scheduling charging. This real-time monitoring and assessment mechanism not only improves the battery pack's operating efficiency but also extends its service life.

[0230] This system also has broad application value in industrial energy storage scenarios. For example, in photovoltaic energy storage systems, lithium iron phosphate battery packs, serving as energy storage devices, require frequent charging and discharging. Monitoring equipment collects real-time operating data from the battery pack and transmits it to the main control unit via a dedicated communication link. The main control unit uses a P2D model to dynamically predict the battery pack's status and, in conjunction with a lifespan assessment module, estimates the remaining lifespan of the battery pack. Through accurate dynamic prediction of SOE and lifespan assessment, the system optimizes the battery pack's charge and discharge strategies, avoiding overcharging or overdischarging, thereby effectively extending the battery pack's service life and improving system efficiency.

[0231] In summary, this system achieves accurate modeling of the complex multi-physics interaction processes within the battery by introducing a P2D model that couples electrochemical, thermodynamic, and multi-physics fields. Data is transmitted between the monitoring equipment and the battery pack's main control unit via a dedicated communication link, ensuring real-time and accurate data acquisition. The P2D model combines historical operating data with real-time status parameters to generate dynamic SOE prediction results. Its simple and efficient structural design enables faster convergence and higher prediction accuracy during model training. Through in-depth analysis of the long-term performance degradation mechanisms of battery packs, more accurate lifespan assessments are achieved, meeting the needs of efficient and reliable battery management under complex operating conditions.

Claims

1. A method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model, characterized in that: The following steps are involved: S1, collecting basic operating data of the lithium iron phosphate battery pack through the sensor network arranged in the battery pack, and establishing a dedicated communication link between the battery pack and the monitoring equipment; S2, uploads the battery pack’s historical operating data, real-time status parameters, and environmental variables via a dedicated communication link; S3, based on historical operating data and real-time state parameters, uses the P2D model of electrochemical-thermal-dynamic multi-physics field coupling to generate the SOE dynamic prediction results of the battery pack; S4, analyzing the prediction results in combination with the long-term performance degradation mechanism, evaluating the remaining life of the battery pack, and outputting the evaluation results; In step S3, the SOE dynamic prediction result is calculated by the P2D model of electrochemical-thermal-dynamic multi-physics field coupling, and the specific steps are as follows: S31, collecting operating parameters of the lithium iron phosphate battery pack, including voltage, current, temperature, internal resistance and electrolyte concentration distribution; S32, dividing the interior of the battery into multiple micro-units, each micro-unit containing a positive electrode region, a negative electrode region, and a separator region, constructing electrochemical reaction equations, heat conduction equations, and mechanical equilibrium equations, and defining the electrochemical reaction rate, heat conduction coefficient, and mechanical stress distribution of each region, which are used to describe the electrochemical reaction process, heat transfer behavior, and mechanical deformation inside the battery, respectively; S33, based on the collected operating parameter data, using a finite difference method to discretize the electrochemical reaction equation, heat conduction equation, and mechanical equilibrium equation inside the battery; S34, through the coupling of electrochemical reaction rate, thermal conductivity and mechanical stress distribution, a P2D model of multi-physics field coupling inside the battery is formed; S35, inputting the discretized multi-physics field coupled P2D model into the computing platform of the main control unit to complete the initialization of the P2D model; S36, defining the initial state parameters of the battery pack, including SOC, SOH and current ambient temperature; S37, based on the P2D model, calculates the energy change rate of the battery pack under different working conditions, and derives the dynamic change curve of SOE through the energy change rate; S38, introduces a dynamic response algorithm, combines the current and voltage data collected in real time, and corrects the dynamic change curve of the SOE; S39, outputting the corrected SOE dynamic change curve to the main control unit for guiding the charge and discharge management of the battery pack.

2. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating the P2D model according to claim 1 is characterized in that: In step S1, the method for establishing the dedicated communication link is as follows: S11, obtaining an identifier of a monitoring device connected to the battery pack; S12, defining a monitoring area for communication coverage with the center point of the battery pack as the center and half of the maximum geometric dimension of the battery pack as the radius; S13, determining whether there is a relay node in the monitoring area; if not, the monitoring device communicates directly with the battery pack main control unit to complete the establishment of a dedicated communication link; if so, proceeding to step S14; S14, with the center point of the battery pack as the center of the circle, a sub-area containing the relay node and concentric with the monitoring area is delineated as the relay coverage area; the monitoring device communicates with the battery pack main control unit through the relay node to complete the establishment of a dedicated communication link.

3. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating the P2D model according to claim 1 is characterized in that: In step S4, the battery life assessment method includes the following steps: S41, collecting capacity attenuation data of the battery pack at different cycle times, including capacity retention rate and internal resistance change rate; S42 first collects capacity decay data of the battery pack at different cycle times, including capacity retention and internal resistance change rate, to analyze the performance decay mechanism of the battery pack in long-term operation and identify key influencing factors, including electrochemical side reactions, thermal runaway risks, and mechanical fatigue; S43, based on the performance degradation mechanism, defines life assessment indicators, including cycle life, calendar life and safety life; S44, using a machine learning algorithm to train the collected decay data to generate a life assessment model; S45, embedding the life assessment model into the life assessment module to predict the remaining life of the battery pack.

4. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model according to claim 3 is characterized in that: The machine learning algorithm is a support vector machine or a neural network algorithm, which is used to perform nonlinear fitting and prediction on capacity attenuation data.

5. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating the P2D model according to claim 2 is characterized in that: The monitoring device and the main control unit transmit data via a dedicated communication link. The communication coverage of the dedicated communication link is an area with the center point of the battery pack as the center and half of the maximum geometric dimension of the battery pack as the radius; the SOE dynamic prediction module receives the output result of the P2D model and generates the SOE dynamic prediction result of the battery pack based on the result; the life assessment module evaluates the remaining life of the battery pack in combination with the long-term performance degradation mechanism and outputs the evaluation result.

6. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model according to claim 1 is characterized in that: When the finite difference method is used to discretize the electrochemical reaction equation, the heat conduction equation and the mechanical equilibrium equation, a numerical calculation method with a fixed step size is adopted.

7. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model according to claim 1 is characterized in that: In the S32, the electrochemical reaction equation, the heat conduction equation and the mechanical equilibrium equation are constructed: The electrochemical reaction process is modeled through electrochemical field, including the lithium ion solid phase diffusion equation, liquid phase transport equation and Butler-Volmer equation; Lithium ion solid phase diffusion equation: The boundary conditions include the particle centrosymmetry: Surface reaction flux: Where t represents time, c s is the solid phase lithium concentration, D s is the solid phase diffusion coefficient, j is the electrode reaction current density, F is the Faraday constant, r is the radial coordinate in the spherical coordinate system, R s represents the radius of solid lithium particles; The liquid phase transport equation includes the conservation equation for the coupling of lithium ion concentration and potential in the electrolyte and the charge conservation equation. The conservation equation for the coupling of lithium ion concentration and potential in the electrolyte is: The charge conservation equation is: in, is the vector differential operator, c e is the liquid phase lithium concentration, ε e is the electrolytic liquid volume fraction, is the effective liquid diffusion coefficient, is the lithium ion migration number, is the liquid phase potential, is the effective conductivity of liquid lithium and the diffusion conductivity of liquid lithium; The Butler-Volmer equation first describes the electrode / electrolyte interface reaction kinetics: Exchange current density j0 expression: The overpotential η is defined as: Among them, α a , α c are the anode and cathode transfer coefficients, U ocp is the open circuit potential, is the surface concentration of solid lithium, is the solid phase potential; R is the gas constant, T is the absolute temperature, and k is the intrinsic reaction rate constant; The heat transfer behavior is represented by thermodynamic field coupling, including the energy conservation equation and temperature sensitive parameter correction. The total heat generation rate Q of the energy conservation equation includes: Joule heat: Reaction heat: Q rxn =jη Polarization heat: Then the heat conduction equation: Where ρ is the material density, C p is the specific heat capacity at constant pressure, ρC p represents volume heat capacity, λ is thermal conductivity, δ eff is the effective conductivity of solid-phase lithium; The temperature-sensitive parameter correction is expressed by the temperature dependence of the diffusion coefficient and the conductivity: Where D represents the diffusion coefficient, D(T) represents the diffusion coefficient at temperature T, Indicates the reference temperature T ref The diffusion coefficient under a is the activation energy, T ref is the reference temperature; Mechanical equilibrium equation: Based on the solid phase diffusion equation and the elastic mechanics equation, the internal stress ξ of the particle is calculated: Where E is the elastic modulus, c s Solid phase lithium concentration, c0 is the initial solid phase lithium concentration, c s,max It represents the maximum intercalation concentration of solid-phase lithium, and ε is the relative deformation of the material under stress or environmental changes.

8. The method for dynamic SOE prediction and life assessment of lithium iron phosphate battery packs integrating a P2D model according to claim 1 or 7, characterized in that: In S32, defining the electrochemical reaction rate, thermal conductivity, and mechanical stress distribution of each region refers to: The electrochemical reaction rate was described using the Butler-Volmer equation to describe the electrode / electrolyte interface reaction kinetics; The thermal conductivity is determined by the positive and negative electrode regions and the diaphragm region. The positive and negative electrode regions are composed of active materials, conductive agents, binders and pore electrolytes. The equivalent thermal conductivity w eff Calculated by the mixed model: w eff =Sm i w i Among them, μ i is the volume fraction of each component, w i is the intrinsic thermal conductivity of the material; The diaphragm region is dominated by material properties; The distribution of mechanical stress is jointly determined by the positive and negative electrode regions and the diaphragm region; among them, the sources of stress in the positive and negative electrode regions are: lithium ion insertion / extraction causes the volume change of active particles, inducing local stress; the sources of stress in the diaphragm region are external pressure conduction and thermomechanical stress; external pressure conduction: the battery packaging pressure is transmitted to the diaphragm through the electrode, and its compression resistance must be considered; thermomechanical stress: temperature changes cause differences in the thermal expansion coefficients of the diaphragm and the electrode, resulting in interfacial shear stress.

Images