Battery state estimation method, device, apparatus, storage medium, and product

Abstract

This application discloses a battery state estimation method, apparatus, device, storage medium, and product, relating to the field of battery management system technology. The method includes: acquiring the original operating dataset of the target battery; performing spatiotemporal alignment and data fusion on the original operating dataset to obtain a spatiotemporally synchronized operating state field; performing multiphysics coupling reconstruction based on the operating state field to obtain an internal state field; obtaining degradation characteristics and stress field based on the internal state field through a damage evolution model; and obtaining battery state estimation results based on the degradation characteristics and stress field. Compared to existing methods, this application constructs an analytical framework combining multiphysics coupling reconstruction and a damage evolution model, mapping surface-measurable electrical and temperature data to a state field inside the battery that is difficult to measure directly, and further extracting degradation characteristics and stress evolution laws, thereby achieving estimation of battery health status and remaining life.

Description

Technical Field

[0001] This application relates to the field of battery management system technology, and in particular to a battery state estimation method, apparatus, device, storage medium and product. Background Technology

[0002] State estimation of lithium-ion batteries, especially state of health estimation, is a core function of battery management systems. Existing technologies mainly rely on external macroscopic parameters such as terminal voltage, total current, and a few discrete temperature points, combined with empirical or simplified electrochemical models. These approaches typically treat the battery interior as a temperature homogeneous entity or approximate the overall thermal state using only a single temperature point.

[0003] However, under actual operating conditions with a wide temperature range (e.g., -30℃ to 60℃) and high charge / discharge rates, a significant and dynamically changing temperature gradient will be generated inside the battery, leading to severe inhomogeneity in internal reaction kinetics and ion transport characteristics. Existing methods lack the ability to dynamically perceive the spatial distribution of the internal temperature field and degradation mechanisms of the battery, resulting in severely lagging estimation results and a sharp drop in accuracy, failing to reveal the internal physical causes of capacity decay.

[0004] Therefore, how to detect wide-temperature-range interference and accurately estimate battery status is an urgent problem to be solved. Summary of the Invention

[0005] The main objective of this application is to provide a battery state estimation method, apparatus, device, storage medium, and product, which aims to solve the technical problem of how to sense wide-temperature-range interference and accurately estimate battery state.

[0006] To achieve the above objectives, this application proposes a battery state estimation method, which is applied to a battery management system, and the method includes: Obtain the raw runtime dataset for the target battery; The original running dataset is spatiotemporally aligned and fused to obtain a spatiotemporally synchronized running state field; Based on the aforementioned operating state field, multi-physics field coupling reconstruction is performed to obtain the internal state field; Based on the internal state field, degradation characteristics and stress field are obtained through a damage evolution model; Based on the degradation characteristics and the stress field, the battery state estimation result is obtained.

[0007] In one embodiment, the step of performing spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field includes: The data points in the original running dataset are labeled with timestamps and spatial coordinates to obtain a dataset with spatiotemporal identifiers; Using a unified time axis as a reference, the voltage and current records carrying timestamps in the dataset are resampled, and the resampled voltage and current records are aligned with the multi-point temperature record time series in the dataset. The time-aligned multi-point temperature records are mapped onto the grid nodes in the battery's three-dimensional geometric model that correspond to the spatial coordinates, and spatial interpolation is performed on the remaining grid nodes to obtain the temperature field. The operating state field is composed of the voltage record and the current record, which are aligned with the temperature field and time.

[0008] In one embodiment, the step of performing multiphysics coupling reconstruction based on the operating state field to obtain the internal state field includes: Based on the operating state field and the principle of electrochemical reaction kinetics, the reaction current density distribution at the solid-liquid interface inside the target battery is obtained. Based on the reaction current density distribution and Ohm's law, the potential distribution over the entire target battery is obtained; Based on the aforementioned reaction current density distribution and the principles of electrochemical thermodynamics, the heat generation rate distribution of electrochemical reaction heat and Joule heat is obtained. By combining the potential distribution with the mass conservation equation in the electrolyte of the target battery, the ion concentration distribution is obtained; The internal state field is constituted by the potential distribution, the ion concentration distribution, and the heat generation rate distribution.

[0009] In one embodiment, the step of obtaining degradation characteristics and stress field based on the internal state field through a damage evolution model includes: Based on the ion concentration distribution and the potential distribution, the material phase interface movement is simulated using the phase field method to obtain the degradation characteristics of material structural damage. Based on the heat generation rate distribution, the temperature field inside the target battery is obtained, and the thermal stress distribution is obtained by combining the thermal expansion coefficients of different materials in the target battery. The stress field is obtained by superimposing the thermal stress distribution with the lattice strain, where the lattice strain is the lattice strain inside the target battery caused by lithium ion insertion and extraction.

[0010] In one embodiment, the step of obtaining the battery state estimation result based on the degradation characteristics and the stress field includes: The material structural damage in the degradation feature is discretized into multiple damage nodes, and the local stress concentration points in the stress field are taken as stress nodes. An internal damage network is established based on the damage nodes, the stress nodes, and the material connection relationships between the nodes. Based on the material fracture toughness and interfacial bonding energy, the correlation strength and failure propagation probability between nodes in the internal damage network are obtained. Based on the correlation strength and the failure propagation probability, the cascade diffusion path of damage inside the target battery is simulated, and the amount of reduction in charge storage capacity by each cascade diffusion is obtained. The capacity decay trajectory is obtained by summing up the attenuation caused by each damage propagation event. Based on the capacity decay trajectory and the initial total charge storage capacity, the battery state estimation result is obtained.

[0011] In one embodiment, the step of obtaining the temperature field inside the target battery based on the heat generation rate distribution, and obtaining the thermal stress distribution by combining the thermal expansion coefficients of different materials in the target battery, includes: Based on the heat generation rate distribution, the temperature field inside the target battery is obtained through a three-dimensional unsteady-state heat conduction equation. Based on the temperature field and the thermal expansion coefficients of different materials in the target battery, the thermal stress distribution is obtained through thermoelastic mechanics equations. The different materials include positive electrode material, negative electrode material, and current collector.

[0012] Furthermore, to achieve the above objectives, this application also proposes a battery state estimation device, the device comprising: The acquisition module is used to acquire the raw runtime dataset of the target battery; The first processing module is used to perform spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field; The second processing module is used to perform multi-physics field coupling reconstruction based on the operating state field to obtain the internal state field; The third processing module is used to obtain degradation characteristics and stress field based on the internal state field through a damage evolution model; The state estimation module is used to obtain the battery state estimation result based on the degradation characteristics and the stress field.

[0013] In addition, to achieve the above objectives, this application also proposes a battery state estimation device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the battery state estimation method as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and which, when executed by a processor, implements the steps of the battery state estimation method described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the battery state estimation method described above.

[0016] This application provides a battery state estimation method, which includes acquiring the original operating dataset of the target battery; performing spatiotemporal alignment and data fusion on the original operating dataset to obtain a spatiotemporally synchronized operating state field; performing multiphysics coupling reconstruction based on the operating state field to obtain an internal state field; obtaining degradation features and stress field through a damage evolution model based on the internal state field; and obtaining the battery state estimation result based on the degradation features and stress field.

[0017] This application first obtains the original operational dataset of the target battery. Then, it performs spatiotemporal alignment and data fusion on the original operational dataset to obtain a spatiotemporally synchronized operational state field. Next, it performs multiphysics coupling reconstruction based on the operational state field to obtain the internal state field. Then, based on the internal state field, it obtains degradation characteristics and stress fields through a damage evolution model. Finally, based on the degradation characteristics and stress field, it obtains the battery state estimation result. Compared with existing methods, this application, by constructing an analytical framework combining multiphysics coupling reconstruction and a damage evolution model, maps surface-measurable electrical and temperature data to a state field inside the battery that is difficult to measure directly, and further extracts degradation characteristics and stress evolution laws, thereby achieving estimation of battery health status and remaining lifetime. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an embodiment of the battery state estimation method of this application. Figure 2 This is a flowchart illustrating Embodiment 2 of the battery state estimation method of this application; Figure 3 This is a flowchart illustrating Embodiment 3 of the battery state estimation method of this application; Figure 4 This is a schematic diagram of the module structure of the battery state estimation device according to an embodiment of this application; Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the battery state estimation method in this application embodiment.

[0021] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] The main solution in this application is the estimation of the state of lithium-ion batteries, particularly the estimation of their health state, which is a core function of the battery management system. Existing technologies mainly rely on external macroscopic parameters such as terminal voltage, total current, and a small number of discrete temperature points, combined with experience or simplified electrochemical models. These solutions typically treat the battery interior as a temperature-uniform entity or approximate the overall thermal state using only a single temperature point.

[0025] However, under actual operating conditions with a wide temperature range (e.g., -30℃ to 60℃) and high charge / discharge rates, a significant and dynamically changing temperature gradient will be generated inside the battery, leading to severe inhomogeneity in internal reaction kinetics and ion transport characteristics. Existing methods lack the ability to dynamically perceive the spatial distribution of the internal temperature field and degradation mechanisms of the battery, resulting in severely lagging estimation results and a sharp drop in accuracy, failing to reveal the internal physical causes of capacity decay.

[0026] This application first obtains the original operational dataset of the target battery. Then, it performs spatiotemporal alignment and data fusion on the original operational dataset to obtain a spatiotemporally synchronized operational state field. Next, it performs multiphysics coupling reconstruction based on the operational state field to obtain the internal state field. Then, based on the internal state field, it obtains degradation characteristics and stress fields through a damage evolution model. Finally, based on the degradation characteristics and stress field, it obtains the battery state estimation result. Compared with existing methods, this application, by constructing an analytical framework combining multiphysics coupling reconstruction and a damage evolution model, maps surface-measurable electrical and temperature data to a state field inside the battery that is difficult to measure directly, and further extracts degradation characteristics and stress evolution laws, thereby achieving estimation of battery health status and remaining lifetime.

[0027] It should be noted that the execution subject of the following embodiments can be a battery state estimation device, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or a battery state estimation device capable of performing the above functions. This embodiment does not specifically limit it in this way. The following uses a battery state estimation device (hereinafter referred to as the device) as the execution subject to describe this embodiment and the following embodiments.

[0028] Based on this, this application proposes a battery state estimation method according to the first embodiment, please refer to... Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the battery state estimation method of this application. The battery state estimation method includes steps S10 to S50: Step S10: Obtain the raw running dataset of the target battery.

[0029] It should be noted that the target battery mentioned above can refer to a lithium-ion battery that requires state estimation, such as the power battery pack of an electric vehicle or the battery module in an energy storage power station.

[0030] The aforementioned raw operational dataset can be an unprocessed collection of raw data collected during actual battery operation, including instantaneous voltage records, instantaneous current records, and multi-point asynchronous temperature sampling records. Instantaneous voltage records refer to the sampled values ​​of the voltage change between the positive and negative terminals of the battery over time; instantaneous current records refer to the sampled values ​​of the total current flowing through the battery during charging and discharging over time; multi-point asynchronous temperature sampling records refer to temperature values ​​collected by multiple temperature sensors distributed on or inside the battery, with possible millisecond-level deviations in the sampling time of each sensor. For example, for a square lithium-ion battery, multiple temperature sensors can be arranged at the top, bottom, sides, and center of the battery, with each sensor asynchronously collecting temperature data according to its own clock.

[0031] In the specific implementation, instantaneous voltage records, instantaneous current records, and multi-point asynchronous temperature sampling records are received from the sensor network via a communication interface. Instantaneous voltage records are continuously sampled at millisecond-level frequencies, instantaneous current records are captured synchronously with voltage records, and multi-point asynchronous temperature sampling records are acquired asynchronously by distributed temperature sensors at different time bases. These data are temporarily stored in the internal memory in the order they are received, forming the initial operational dataset.

[0032] To facilitate understanding, the following example is used for explanation, but it does not impose specific limitations on this embodiment. For example, suppose there is a lithium-ion battery for an electric vehicle with a rated capacity of 50 Ah, which is discharged at a rate of 1.5C in a low-temperature environment of -10 degrees Celsius. The processor in the battery management system acquires instantaneous voltage records (sampling interval of 1 ms), instantaneous current records (sampling interval of 1 ms), and asynchronous sampling records from eight temperature sensors (each sensor samples once per second, but there is a random deviation of 2 to 5 ms between the sampling times) during the discharge process through the controller area network bus. The processor temporarily stores these data in memory according to the order of receipt, forming the original running dataset.

[0033] Step S20: Perform spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field.

[0034] It should be explained that the aforementioned spatiotemporal alignment refers to the process of uniformly calibrating data points with different time bases and spatial locations in the original operational dataset across both the temporal and spatial dimensions. In the temporal dimension, the sampling times of voltage records, current records, and multi-point temperature records are aligned to the same system time axis; in the spatial dimension, each temperature record is associated with the coordinates of its corresponding sensor installation location.

[0035] The aforementioned data fusion can refer to integrating aligned multi-source data (voltage, current, temperature) according to a unified spatiotemporal reference to form a data set describing the overall state of the battery.

[0036] The aforementioned spatiotemporally synchronized operating state field can refer to a battery operating state data volume that maintains synchronization in both time and space dimensions, including time-synchronized voltage sequences, current sequences, and spatially continuous temperature field distribution.

[0037] In the specific implementation, firstly, each data point in the original running dataset is labeled with a precise timestamp and sensor spatial coordinates, generating a dataset with spatiotemporal identification. Next, using a unified system timeline as a reference, the timestamped voltage and current records in the dataset are resampled, ensuring that the resampled voltage and current records are fully aligned with the multi-point temperature records in the dataset in terms of time series. Subsequently, the time-aligned multi-point temperature records are mapped to corresponding grid nodes in the battery's 3D geometric model. For the remaining grid nodes without a direct mapping relationship, spatial interpolation is performed using the temperature values ​​of neighboring nodes, ultimately generating a temporally synchronized and spatially continuous temperature field. Finally, the temperature field is combined with the time-aligned voltage and current records to form a spatiotemporally synchronized operating state field.

[0038] To facilitate understanding, the following explanation uses examples, but does not impose specific limitations on this embodiment. For example, continuing from the discharge scenario in the previous example, the processor in the battery management system first marks the sampling time (e.g., the 1st, 000th millisecond, the 2nd, 000th millisecond) and the coordinates of the voltage sensor for each voltage sampling point, marks the sampling time and the coordinates of the current sensor for each current sampling point, and marks the sampling time and the corresponding spatial coordinates of the temperature sensor for each temperature sampling point (e.g., the top sensor is located at coordinates (10, 20, 5) mm). Next, the processor resamples the voltage and current records into data points at uniform intervals of 1000 per second, using the system time axis as a reference, so that these data points are time-aligned with the temperature records (1 sampling point per second). Then, the processor maps the readings of the eight temperature sensors to the corresponding mesh nodes in the battery's three-dimensional geometric model. For other mesh nodes in the model without temperature sensors, kriging interpolation is performed using the readings of the three neighboring temperature sensors to calculate the node temperature value, thereby forming a complete three-dimensional temperature field inside the battery. Finally, the processor packages and stores the three-dimensional temperature field with time-aligned voltage and current sequences as a spatiotemporally synchronized operating state field.

[0039] Step S30: Perform multi-physics coupling reconstruction based on the operating state field to obtain the internal state field.

[0040] It should be explained that the aforementioned multiphysics coupling reconstruction refers to the process of reconstructing or reconstructing the internal state of a battery by inversely inferring from or considering the interactions of multiple physical fields, such as electrochemistry and thermodynamics, based on the battery's operating state field. During the charging and discharging process, multiple physical processes, such as electrochemical reactions, ion transport, and heat generation and conduction, influence and couple with each other. For example, temperature changes affect reaction rates and ion mobility, while electrochemical reactions themselves generate heat.

[0041] The aforementioned internal state field can refer to a dataset describing the spatial distribution of physical quantities within the battery, including potential distribution, ion concentration distribution, and heat generation rate distribution. Potential distribution refers to the distribution of potential values ​​at various spatial points within the battery; ion concentration distribution refers to the spatial variation of lithium ion concentration in the electrolyte; and heat generation rate distribution refers to the rate of heat generation per unit volume at various locations within the battery, including both electrochemical reaction heat and Joule heat.

[0042] In the specific implementation, firstly, based on the temperature field, voltage records, and current records in the operating state field, and combined with the principles of electrochemical reaction kinetics, the reaction current density distribution at the interface between the solid-phase material and the liquid-phase electrolyte inside the battery is calculated. Next, based on the reaction current density distribution and Ohm's law, a potential control equation is established and solved on the global grid nodes of the battery to obtain the potential distribution. Subsequently, based on the reaction current density distribution and the principles of electrochemical thermodynamics, the generation rates of electrochemical reaction heat and Joule heat at various spatial locations inside the battery are calculated, forming a heat generation rate distribution. Then, combining the potential distribution and the mass conservation equation in the electrolyte inside the battery, the spatial variation of lithium ion concentration in the electrolyte is solved to generate the ion concentration distribution. Finally, the potential distribution, ion concentration distribution, and heat generation rate distribution are combined to constitute the internal state field.

[0043] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing from the discharge scenario in the previous example, the processor in the battery management system calculates the reaction current density distribution at the interface between the negative electrode particle surface and the electrolyte based on the three-dimensional temperature field data and time-aligned voltage and current data provided in the operating state field, using the Butler-Volmer electrochemical kinetic equation. The reaction current density in the low-temperature region (such as the area near the battery casing) is significantly lower than that in the high-temperature region (such as the battery center). Next, the processor solves the Poisson equation based on the reaction current density distribution and Ohm's law to obtain the potential distribution in the three-dimensional space inside the battery. The results show that the internal resistance in the low-temperature region is higher, leading to a significant decrease in local potential. Subsequently, the processor calculates the electrochemical reaction heat based on the reaction current density distribution and calculates the Joule heat based on the current density distribution. Adding the two together yields the heat generation rate distribution, revealing that the heat generation rate is highest in the battery center and at the electrode connection. Then, the processor combines the potential distribution with the mass conservation equation in the electrolyte to solve for the lithium ion concentration distribution, showing that the low-temperature region has a lower local lithium ion concentration due to a decrease in ion migration rate. Finally, the processor integrates the potential distribution, ion concentration distribution, and heat generation rate distribution into a single data structure, which serves as the internal state field.

[0044] Step S40: Based on the internal state field, obtain the degradation characteristics and stress field through the damage evolution model.

[0045] It should be explained that the aforementioned damage evolution model can refer to a computational model used to describe the gradual development of battery internal materials from an initially intact state to structural damage. This model is based on physical mechanisms rather than empirical fitting and can simulate degradation processes such as material phase interface movement, microcrack initiation and propagation. The aforementioned degradation characteristics can refer to a quantitative description characterizing the degree of structural damage to battery materials, such as the regional distribution of solid electrolyte interface film rupture and the interconnected topology of microcracks appearing in active material particles. The aforementioned stress field can refer to the spatial distribution of internal stress generated by battery internal materials under the influence of factors such as thermal expansion and lattice volume changes, including thermal stress and lattice stress caused by lithium ion insertion and extraction.

[0046] In its implementation, the phase-field method is first used to simulate the movement process of the battery material phase interface based on the ion concentration and potential distribution in the internal state field. The phase-field method distinguishes between undamaged and damaged states of the material by defining order parameters and establishing governing equations including chemical potential gradients and strain energy densities to drive the evolution of these order parameters. By solving the governing equations, the device calculates the connected regions where the solid electrolyte interface film ruptures and microcracks appear in the active material particles, using these connected regions as the degradation characteristics of the material structure damage. Next, the temperature field inside the battery is calculated based on the heat generation rate distribution in the internal state field. Combining the temperature field with the thermal expansion coefficients of different materials in the battery (such as positive electrode materials, negative electrode materials, separators, and current collectors), the device calculates the thermal stress distribution using thermoelastic mechanics equations. Subsequently, the thermal stress distribution is superimposed with lattice strain to obtain the final stress field. Here, lattice strain refers to the strain caused by the change in the material lattice size due to the insertion and extraction of lithium ions into the positive and negative electrode materials.

[0047] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing from the discharge scenario in the previous example, the processor in the battery management system uses the phase field method to simulate the growth and rupture process of the solid electrolyte interface film on the surface of the negative electrode material, based on the lithium-ion concentration distribution and potential distribution in the internal state field. The processor defines an order parameter to represent the degree of material evolution from undamaged to fully damaged, establishes the Ginzburg-Landau equation to control the change of the order parameter, and obtains the connected topology of the ruptured region of the solid electrolyte interface film through numerical solution. This topology is the degradation characteristic. Simultaneously, based on the heat generation rate distribution in the internal state field, the processor obtains the internal temperature field of the battery by solving the three-dimensional unsteady-state heat conduction equation, finding that the temperature in the central region of the battery reaches 35 degrees Celsius, while the temperature in the edge region is only 12 degrees Celsius. The processor combines this with the positive electrode material (thermal expansion coefficient of 1.2 × 10⁻⁶). -5 (per Kelvin), negative electrode material (thermal expansion coefficient of 1.5 × 10⁻⁶) -5 (per Kelvin) and aluminum current collector (coefficient of thermal expansion is 2.3 × 10⁻⁶).-5 The difference in thermal expansion coefficients (per Kelvin) was used to calculate the thermal stress distribution using thermoelasticity equations, revealing a thermal stress concentration of approximately 35 MPa at the interface between the positive and negative electrodes. The processor then superimposed this thermal stress distribution with the stress generated by lattice strain due to lithium-ion insertion and extraction to obtain the final stress field. The results showed that the local stress peak at the electrode particle boundary reached 50 MPa.

[0048] Step S50: Based on the degradation characteristics and the stress field, obtain the battery state estimation result.

[0049] It should be explained that the battery state estimation results mentioned above can refer to quantitative assessment information on the current health status and future usable capacity of the target battery. Health status is used to characterize the degree to which the battery retains its capacity relative to a brand new battery, usually expressed as a percentage; usable capacity decay trajectory is used to describe the trend curve of battery capacity changing with the number of cycles or time during subsequent cycles.

[0050] In its implementation, the device first discretizes the material structural damage in the degradation characteristics into multiple damage nodes, and uses local stress concentration points in the stress field as stress nodes. Next, based on the damage nodes, stress nodes, and the connections established between nodes through material entities, the device constructs an internal damage network to describe the internal damage topology of the battery. Then, using the intrinsic material parameters of fracture toughness and interfacial bonding energy, the device calculates the correlation strength between nodes in the internal damage network and the failure propagation probability of damage from one node to an adjacent node. Following this, based on the correlation strength and failure propagation probability, the device simulates the cascading path of damage propagation within the target battery, determining at each simulation step whether damage propagates from the current node to an adjacent node and quantifying the reduction in overall charge storage capacity of the battery for each cascading diffusion event. The device accumulates the reduction in charge storage capacity caused by each damage diffusion event, forming a capacity loss sequence that varies with the number of cycles or time, and uses this capacity loss sequence as the capacity decay trajectory. Finally, based on the capacity decay trajectory and the initial total charge storage capacity of the target battery, the device calculates the battery state estimation result.

[0051] To facilitate understanding, the following examples are provided for illustration, but do not impose specific limitations on this embodiment. For instance, continuing from the discharge scenario in the previous example, the processor in the battery management system discretizes multiple solid electrolyte interfacial film rupture regions and microcrack regions identified in the degradation features into damage nodes, resulting in 15 discrete nodes; points in the stress field with local stress exceeding 40 MPa are marked as stress nodes, resulting in 8 stress nodes. The processor establishes an internal damage network based on the spatial location of these nodes in the battery's three-dimensional model and the material connectivity between the nodes. The processor calculates the correlation strength and failure propagation probability between each pair of adjacent nodes based on the fracture toughness of the negative electrode material (e.g., the square root of 0.5 MPa·m) and the interfacial binding energy between the solid electrolyte interfacial film and the negative electrode material (e.g., 0.2 joules per square meter), where the failure propagation probability is higher between nodes that are closer together and have higher stress levels. The processor uses the Monte Carlo method to simulate the damage diffusion process, starting from the initial damage node, randomly sampling to determine whether the damage propagates to adjacent nodes. Each propagation event causes the active material of the corresponding node to lose electrochemical activity, resulting in a decrease in charge storage capacity. After 1000 simulation cycles, the processor sums up the capacity loss caused by all damage propagation events, obtaining a total capacity loss of 8.2 Ah. Using the initial total charge storage capacity of the target battery (50 Ah) as a benchmark, and combining it with the capacity decay trajectory, the processor calculates a global health of 83.6%, and outputs both the global health and the capacity decay trajectory as the battery state estimation result.

[0052] This embodiment first acquires the original operational dataset of the target battery. Then, it performs spatiotemporal alignment and data fusion on the original operational dataset to obtain a spatiotemporally synchronized operational state field. Next, it performs multiphysics coupling reconstruction based on the operational state field to obtain the internal state field. Then, based on the internal state field, it obtains degradation characteristics and stress fields through a damage evolution model. Finally, based on the degradation characteristics and stress field, it obtains the battery state estimation result. Compared to existing methods, this embodiment constructs an analytical framework combining multiphysics coupling reconstruction and a damage evolution model, mapping surface-measurable electrical and temperature data to a state field inside the battery that is difficult to measure directly, and further extracts degradation characteristics and stress evolution laws, thereby achieving estimation of battery health status and remaining life.

[0053] Based on the first embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description and will not be repeated hereafter. On this basis, a second embodiment of the battery state estimation method of this application is proposed, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating Embodiment 2 of the battery state estimation method of this application. To obtain a spatiotemporally synchronized operating state field, such as... Figure 2As shown, in this embodiment, the step of performing spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field includes: Step S201: Mark the data points in the original running dataset with timestamps and spatial coordinates to obtain a dataset with spatiotemporal identifiers.

[0054] It should be explained that the timestamp mentioned above refers to a time stamp used to mark the time when data points are collected, with an accuracy of up to the microsecond level, used to solve the problem of asynchronous sampling times between different sensors.

[0055] The aforementioned spatial coordinates may refer to the installation position coordinates of the sensor in the three-dimensional geometric model of the target battery. The three-dimensional model of the battery is predefined and used to mark the spatial acquisition position corresponding to each data point.

[0056] The aforementioned dataset with spatiotemporal identifiers can refer to a data set formed by adding acquisition time information and spatial location information to each data point in the original running dataset. The voltage records, current records, and multi-point temperature records in this data set all carry clear timestamps and spatial coordinates.

[0057] In the specific implementation, each data point is first extracted from the original operational dataset, including instantaneous voltage records, instantaneous current records, and multi-point asynchronous temperature sampling records. For instantaneous voltage records, the device adds a timestamp to each voltage sample value, marking the actual acquisition time of the voltage value. Simultaneously, based on the installation location of the voltage sensor in the battery's 3D model, the device adds corresponding spatial coordinates to each voltage record. For instantaneous current records, the device adds a timestamp of the acquisition time and the spatial coordinates of the current sensor's location to each current sample value. For multi-point asynchronous temperature sampling records, the device adds the actual timestamp of when the temperature sensor acquired the data to each temperature sample value, and adds corresponding spatial coordinates to each temperature record based on the preset coordinates of the temperature sensor in the battery's 3D model. After marking, the device integrates all data points carrying timestamps and spatial coordinates into a dataset with spatiotemporal identification.

[0058] To facilitate understanding, the following examples are provided for illustration, but do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned discharge scenario, the processor in the battery management system reads a voltage instantaneous record from the raw operating dataset. The sampled value of this record is 3.6 volts. The processor timestamps this voltage value as 1000.000 milliseconds and marks the spatial coordinates as the location of the voltage sensor (coordinates of the battery positive terminal tab: (20, 30, 5) mm). The processor reads a current instantaneous record with a sampled value of 75 amperes (corresponding to a 1.5C rate), timestamps it as 1000.000 milliseconds, and marks the spatial coordinates as the location of the current sensor (coordinates of the battery negative terminal tab: (20, 30, 45) mm). The processor reads an asynchronous temperature sampling record with a sampled value of 12.5 degrees Celsius. This temperature data comes from the top sensor, and the processor timestamps this temperature value as 1000.005 milliseconds (lagging behind the voltage and current records by 5 milliseconds), marking the spatial coordinates as the location of the top temperature sensor (10, 20, 5) mm. After the processor marks all data points one by one, it integrates these data points, which carry timestamps and spatial coordinates, into a dataset with spatiotemporal identifiers.

[0059] Step S202: Using a unified time axis as a reference, resample the voltage and current records carrying timestamps in the dataset, and align the resampled voltage and current records with the multi-point temperature record time series in the dataset.

[0060] It should be explained that the aforementioned unified time axis can refer to a predefined standard time reference used to calibrate data points from different sensors with different sampling times to the same time coordinate system, usually with the system clock as the reference.

[0061] The aforementioned resampling can refer to the process of re-extracting values ​​from voltage and current records carrying timestamps according to time points on a unified time axis. When the original sampling time is inconsistent with the target time on the unified time axis, the value corresponding to the target time is calculated by an interpolation algorithm (such as linear interpolation).

[0062] The aforementioned time series alignment can refer to matching the resampled voltage records, current records, and multi-point temperature records on a unified time axis according to the same time points, so that the three data sequences have a corresponding timestamp at each time point.

[0063] In the specific implementation, a unified timeline is first determined, which divides time points at fixed time intervals (e.g., 1 millisecond or 10 milliseconds), with the starting time point aligned with the timestamp of the earliest data point in the dataset. Next, the voltage records in the dataset are traversed. For each target time point on the unified timeline, the two sampling times adjacent to that target time point in the voltage record are found, and the voltage value corresponding to the target time point is calculated using a linear interpolation algorithm, forming a resampled voltage record sequence. Similarly, the same resampling operation is performed on the current records to obtain a resampled current record sequence. Subsequently, the resampled voltage and current record sequences are matched with the multi-point temperature records in the dataset according to the time points on the unified timeline. Since the sampling frequency of temperature records is usually lower than that of voltage and current records, the device copies each temperature sample value to the corresponding time point and subsequent time points on the unified timeline until the next temperature sample value appears, ensuring that the voltage, current, and multi-point temperature records have corresponding timestamps at each time point, achieving complete alignment of the three time series.

[0064] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of marking timestamps, the processor in the battery management system determines a unified time axis, starting at millisecond 1000.000 and setting the time interval to 1 millisecond. The processor iterates through the voltage records, with the original voltage sampling times being milliseconds 1000.000, 1001.000, 1002.000, etc. Each time point has a corresponding voltage sample value, so the resampled voltage sequence is the same as the original sequence. The processor iterates through the current records, similarly resampling at 1-millisecond intervals to obtain the current sequence. For multi-point temperature records, the sampling frequency of the eight temperature sensors is once per second, meaning there is a temperature sample value every 1000 milliseconds. The processor copies the temperature sample value at 1000.000 ms (top 12.5 degrees Celsius, center 15.1 degrees Celsius, bottom 11.8 degrees Celsius, etc.) to every millisecond between 1000.000 ms and 1000.999 ms, and copies the temperature sample value at 1001.000 ms to every millisecond between 1001.000 ms and 1001.999 ms. After the above processing, the processor obtains a time-aligned sequence starting from 1000.000 ms, where each millisecond contains a voltage value, a current value, and a set of temperature values.

[0065] Step S203: Map the time-aligned multi-point temperature records onto the grid nodes in the battery's three-dimensional geometric model that correspond to the spatial coordinates, and perform spatial interpolation on the remaining grid nodes to obtain the temperature field.

[0066] It should be explained that the aforementioned three-dimensional geometric model of the battery can refer to a digital model used to characterize the physical shape and internal structure of the target battery. It typically employs finite element mesh discretization to divide the battery into multiple tiny mesh units, each with specific spatial coordinates and geometric properties. The aforementioned mesh nodes can refer to the vertices of the mesh units in the battery's three-dimensional geometric model. Each node has unique three-dimensional spatial coordinates and is used to represent the physical quantity value (e.g., temperature) at that location.

[0067] The aforementioned spatial interpolation can refer to the process of estimating the temperature value of a grid node that is not directly covered by a temperature sensor by using the temperature records of neighboring grid nodes with known temperature values ​​and employing interpolation algorithms (such as inverse distance weighted interpolation, Kriging interpolation, or linear interpolation).

[0068] In the specific implementation, a pre-built 3D geometric model of the battery is first loaded. This model contains complete grid node coordinate information and grid cell connection relationships. Next, each time-aligned temperature record is traversed. Each temperature record carries a specific temperature value and the spatial coordinates of the sensor that acquired the temperature data within the battery. The device searches for a grid node in the battery's 3D geometric model that matches the spatial coordinate position and directly assigns the temperature value to the found grid node. For any remaining grid nodes in the battery's 3D geometric model not directly covered by the temperature sensor, the spatial coordinates of these remaining grid nodes are obtained. The device then finds the grid nodes with the closest known temperature values ​​(e.g., the three or four closest nodes). Based on the distances between these known nodes and the target node, a spatial interpolation algorithm is used to calculate the estimated temperature value at the target node. Spatial interpolation calculations are performed sequentially for each remaining grid node, ultimately ensuring that all grid nodes in the battery's 3D geometric model obtain their corresponding temperature values. The spatial distribution of these node temperature values ​​collectively constitutes the temperature field.

[0069] To facilitate understanding, the following explanation uses examples, but does not impose specific limitations on this embodiment. For example, continuing with the aforementioned example of time series alignment, the processor in the battery management system loads a three-dimensional geometric model of a square lithium-ion battery containing 100,000 grid nodes. The processor acquires eight temperature records at the 1000,000th millisecond after time alignment, including the top sensor temperature of 12.5 degrees Celsius (coordinates (10, 20, 5) mm), the center sensor temperature of 15.1 degrees Celsius (coordinates (20, 30, 25) mm), and the bottom sensor temperature of 11.8 degrees Celsius (coordinates (10, 40, 45) mm), etc. The processor locates the mesh node at coordinates (10, 20, 5) mm in the 3D geometric model and assigns its temperature a value of 12.5 degrees Celsius; it locates the mesh node at coordinates (20, 30, 25) mm and assigns it a value of 15.1 degrees Celsius; it locates the mesh node at coordinates (10, 40, 45) mm and assigns it a value of 11.8 degrees Celsius; and so on, completing the temperature assignment for the eight sensor nodes. For the remaining 99,992 mesh nodes not covered by temperature sensors, the processor iterates through each node, obtains its coordinates, and then finds the three closest known temperature nodes. The temperature value of the node is calculated by weighting the values ​​based on the reciprocal of the distance. For example, for the mesh node at coordinates (15, 30, 15) mm, the three closest known nodes are the center sensor node (approximately 11.2 mm), the top sensor node (approximately 15.8 mm), and the side sensor node (approximately 18.5 mm). The processor calculates the temperature of this node to be approximately 13.2 degrees Celsius based on the distance weights. After the processor completes the interpolation calculation for all remaining grid nodes, it obtains the temperature value distribution on 100,000 grid nodes of the entire battery three-dimensional geometric model. This distribution is the temperature field at the 1000,000th millisecond.

[0070] Step S204: The operating state field is formed by the voltage record and the current record, which are aligned with the temperature field and time.

[0071] In the specific implementation, after performing spatial interpolation calculations of the temperature field, the three-dimensional temperature field data corresponding to each time point on the unified time axis has been obtained. Simultaneously, after performing resampling and time series alignment, the voltage and current values ​​corresponding to each time point on the unified time axis have been obtained. The voltage, current, and three-dimensional temperature field data at the same time point are correlated and integrated to form a complete data structure. Specifically, for each time point on the unified time axis, the device combines the voltage value, current value, and the calculated three-dimensional temperature field (including the temperature values ​​of all grid nodes) at that time point into an operating state field for a single time slice. The device sequentially performs the above integration operation on all time points on the unified time axis, ultimately obtaining an operating state field sequence covering the entire time range. Each time slice in this operating state field sequence contains the battery's terminal voltage, charging / discharging current, and complete three-dimensional temperature spatial distribution information within the battery at that moment.

[0072] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of temperature field construction, the processor in the battery management system obtains a time-aligned voltage value of 3.6 volts, a time-aligned current value of 75 amps, and three-dimensional temperature field data (containing temperature values ​​of 100,000 grid nodes, with a central region of approximately 15.1 degrees Celsius and an edge region of approximately 11.8 degrees Celsius) at the 1000.000 ms time slice. The processor packages the voltage value of 3.6 volts, the current value of 75 amps, and the three-dimensional temperature field data into a single runtime field data object. Next, the processor performs the same operation at the 1001.000 ms time slice, obtaining a voltage value of 3.59 volts, a current value of 74.8 amps, and the corresponding three-dimensional temperature field data (the temperature distribution may change slightly due to continuous heat generation from electrochemical reactions), and packages this data into another runtime field data object. The processor repeats the above operation for each millisecond time slice during the discharge process, and finally obtains a complete operating state field sequence from the start to the end of the discharge. This sequence can be used for subsequent multiphysics coupling reconstruction and battery state estimation.

[0073] Furthermore, in order to obtain the internal state field of the target battery, in this embodiment, the step of performing multiphysics coupling reconstruction based on the operating state field to obtain the internal state field includes: Step S301: Based on the operating state field and the principle of electrochemical reaction kinetics, obtain the reaction current density distribution at the interface between the solid and liquid phases inside the target battery.

[0074] It should be explained that the above-mentioned electrochemical reaction kinetics principle can refer to the physicochemical principle that describes the quantitative relationship between electrochemical reaction rate and factors such as electrode potential, reactant concentration, and temperature, and is usually characterized by the Butler-Volmer equation.

[0075] The aforementioned solid-liquid interface refers to the boundary surface where the electrode material (solid phase) and the electrolyte (liquid phase) inside the battery come into contact, where lithium-ion insertion and extraction reactions occur. The aforementioned reaction current density distribution refers to the distribution of current generated by the electrochemical reaction at various local locations on the solid-liquid interface per unit area. This distribution reflects the differences in reaction activity in different regions inside the battery and is typically influenced by temperature, lithium-ion concentration, and potential distribution.

[0076] In the specific implementation, the temperature field, voltage record, and current record for the current time slice are first extracted from the operating state field. Next, the solid-liquid phase interface inside the target battery is discretized into multiple micro-surfaces, each corresponding to an interface mesh element in the battery's three-dimensional geometric model. For each micro-surface, the reaction current density at that surface is calculated based on electrochemical reaction kinetics principles (e.g., the Butler-Volmer equation). During the calculation, the device needs to input the local temperature value (interpolated from the temperature field), local lithium-ion concentration (calculated in subsequent steps, using an initial value in the first iteration), local overpotential (related to the battery terminal voltage and local potential), and kinetic parameters of the electrode materials (e.g., exchange current density and charge transfer coefficient). The device iterates through all interface micro-surfaces, calculating the reaction current density for each surface sequentially, and organizes the calculation results for all micro-surfaces according to their spatial location to form a reaction current density distribution.

[0077] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of constructing the operating state field, the processor in the battery management system extracts the operating state field at the 1000,000th millisecond time slice, obtaining data on the battery terminal voltage of 3.6 volts, the discharge current of 75 amperes, and the three-dimensional temperature field. The processor discretizes the interface between the negative electrode material and the electrolyte inside the battery into 5000 micro-surface elements. For a micro-surface element located in the central region of the battery, the processor obtains the local temperature of that location as 15.1 degrees Celsius from the temperature field. According to the Butler-Volmer equation, the input exchange current density at that temperature is 2.5 mA / cm², the charge transfer coefficient is 0.5, and the overpotential is 0.05 volts, calculating the reaction current density of that micro-surface element as 3.2 mA / cm². For a micro-element surface located in the low-temperature region at the edge of the battery, the processor obtains a local temperature of 11.8 degrees Celsius, with the exchange current density decreasing to 1.2 mA / cm², and calculates a reaction current density of 1.5 mA / cm². The processor completes calculations for all 5000 micro-element surfaces, obtaining the spatial distribution of reaction current density across the entire solid-liquid interface. The reaction current density is higher in the central region and lower in the edge region, reflecting the influence of temperature non-uniformity on the activity of the electrochemical reaction.

[0078] Step S302: Based on the reaction current density distribution and Ohm's law, obtain the potential distribution of the target battery over its entire range.

[0079] It should be explained that Ohm's law can refer to the physical law describing the relationship between current density and electric field strength in a conductor. In the conductive medium inside a battery (including electrode materials, electrolyte, and current collector), the potential gradient is proportional to the current density, and the proportionality constant is the conductivity of the material.

[0080] The aforementioned potential distribution over the entire domain can refer to the distribution of potential values ​​at all spatial locations within the target battery in three-dimensional space, including the potential spatial variation patterns of various components such as the positive electrode current collector, positive electrode material, separator region, negative electrode material, and negative electrode current collector.

[0081] In the specific implementation, the calculated reaction current density distribution is first obtained, which describes the local current intensity generated by the electrochemical reaction at the solid-liquid interface. Next, the potential governing equations for the entire target battery are established. These equations are derived based on Ohm's law and take the form of a Poisson or Laplace equation, where the Laplace operator of the potential is related to the divergence of the current density. The equations are discretized at all grid nodes of the battery's three-dimensional geometric model, transforming the continuous partial differential equations into a discrete system of algebraic equations. During the equation solution process, the reaction current density distribution is applied as a source term or boundary condition to the grid nodes corresponding to the solid-liquid interface, while the applied potential or total current at the positive and negative electrode current collectors is set as boundary conditions. Numerical solution methods (such as the finite element method or the finite volume method) are used to solve the discretized system of algebraic equations, obtaining the potential value at each grid node in the battery's three-dimensional geometric model. The device organizes the potential values ​​of all grid nodes according to spatial coordinates to form the potential distribution across the entire target battery.

[0082] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of calculating the reaction current density distribution, the processor in the battery management system has obtained the reaction current density distribution of 5000 micro-surface elements at the interface between the negative electrode and the electrolyte, where the reaction current density in the central region is 3.2 mA / cm² and the edge region is 1.5 mA / cm². The processor establishes a global potential control equation and discretizes the battery's three-dimensional geometric model containing 100,000 grid nodes using the finite element method. The processor applies the reaction current density distribution as a boundary condition to the solid-liquid interface nodes, sets the potential reference zero point at the positive electrode current collector, and sets the total current at the negative electrode current collector to 75 amperes as another boundary condition. The processor solves the discretized algebraic equations to obtain the potential value of each grid node. The results show that the potential at the positive electrode current collector is 3.65 V. The potential inside the positive electrode material gradually decreases, with the potential at the solid-liquid interface dropping to approximately 3.55 V (the central region experiences a more significant potential decrease due to the high reaction current density, reaching approximately 3.54 V; the edge region experiences a smaller potential decrease due to the low reaction current density, reaching approximately 3.56 V). The potential inside the negative electrode material continues to decrease, reaching 3.60 V at the negative electrode current collector. The processor organizes the potential values ​​of all nodes into three-dimensional distribution data according to spatial coordinates, forming the potential distribution across the entire target battery. This distribution clearly demonstrates the spatial variation of potential within the battery caused by electrochemical reactions and ohmic losses.

[0083] Step S303: Based on the reaction current density distribution and the electrochemical thermodynamics principle, the heat generation rate distribution of electrochemical reaction heat and Joule heat is obtained.

[0084] It should be explained that the aforementioned electrochemical thermodynamic principles can refer to the thermodynamic laws describing the relationship between energy conversion and heat generation during electrochemical reactions, including enthalpy change, entropy change, and irreversible thermal effects caused by overpotential.

[0085] The aforementioned electrochemical reaction heat can refer to the heat generated due to the reaction entropy change and overpotential when lithium ions undergo insertion and extraction reactions at the solid-liquid interface. It includes two parts: reversible reaction heat (related to entropy change) and irreversible activation heat (related to overpotential).

[0086] The Joule heat mentioned above can refer to the heat generated by ohmic losses when current passes through the resistive materials inside the battery (including electrolyte, electrode materials, current collectors, and contact interfaces), and its magnitude is equal to the square of the current density multiplied by the resistivity.

[0087] The aforementioned heat generation rate distribution refers to the distribution of the heat generation rate per unit volume of electrochemical reaction heat and Joule heat at various spatial locations inside the battery in three-dimensional space, usually expressed in watts per cubic meter or milliwatts per cubic centimeter.

[0088] In the specific implementation, the calculated reaction current density distribution and potential distribution are first obtained. For the calculation of electrochemical reaction heat, each micro-element surface at the solid-liquid interface is traversed. Based on the principles of electrochemical thermodynamics, the local reaction current density of that micro-element surface is obtained from the reaction current density distribution. Combined with the local overpotential of that micro-element surface (obtained from the potential distribution) and the reaction entropy change parameter, the electrochemical reaction heat generation rate at that micro-element surface is calculated. For the calculation of Joule heat, all grid cells in the three-dimensional geometric model of the battery are traversed. Based on Ohm's law, the local potential gradient within each grid cell is calculated from the potential distribution. Combined with the conductivity of the material at that location, the Joule heat generation rate generated when the current density flows inside the material is calculated. For grid cells that are simultaneously at the solid-liquid interface, the device superimposes the electrochemical reaction heat generation rate and the Joule heat generation rate. The device organizes the heat generation rate calculation results of all grid cells according to spatial coordinates to form a heat generation rate distribution.

[0089] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of potential distribution calculation, the processor in the battery management system traverses 5000 micro-surface elements at the interface between the negative electrode and the electrolyte. For a micro-surface element in the central region, the processor obtains a reaction current density of 3.2 mA / cm², an overpotential of 0.05 V, a reversible thermal coefficient corresponding to the reaction entropy change of 0.2 mV / Kelvin, and a local temperature of 15.1 degrees Celsius (288.25 Kelvin), calculating that the electrochemical reaction heat generation rate of this micro-surface element is approximately 0.16 mW / cm². The processor traverses all grid cells in the three-dimensional geometric model of the battery. For a grid cell in the positive electrode material region, the processor obtains a potential gradient of 0.5 V / mm from the potential distribution, and the conductivity of the positive electrode material is 10 Siemens / m, calculating that the Joule heat generation rate is 2.5 mW / cm³. For a single grid cell in the separator region, the separator has a low conductivity (approximately 0.1 Siemens per meter), resulting in a higher Joule heat generation rate of approximately 250 milliwatts per cubic centimeter under the same potential gradient. The processor superimposes the electrochemical reaction heat and Joule heat of all grid cells. For grid cells at the solid-liquid interface, which contain both heat sources, the total heat generation rate is approximately 300 milliwatts per cubic centimeter. For the bulk region far from the interface, which contains only Joule heat, the heat generation rate varies from approximately 2 to 250 milliwatts per cubic centimeter. The processor organizes the heat generation rates of all grid cells into a three-dimensional distribution according to spatial coordinates, resulting in a heat generation rate distribution. This distribution shows that the separator region and the solid-liquid interface region are the main heat-generating sites inside the battery.

[0090] Step S304: Combine the potential distribution with the mass conservation equation in the electrolyte of the target battery to obtain the ion concentration distribution.

[0091] It should be explained that the above-mentioned mass conservation equation can refer to a partial differential equation describing the mass conservation of lithium ions during transport in the electrolyte. It is usually characterized by the Nernst-Planck equation or the simplified Fick's law combined with the electromigration term. This equation takes into account transport mechanisms such as ion diffusion, electromigration, and convection.

[0092] The aforementioned ion concentration distribution refers to the concentration distribution of lithium ions in the electrolyte within the three-dimensional space of the battery, usually expressed in moles per liter or millimoles per cubic centimeter, reflecting the spatial distribution of lithium ions migrating between the positive and negative electrodes during the charging and discharging process.

[0093] In the specific implementation, the calculated potential distribution is first obtained, which provides the potential gradient information at various spatial locations within the electrolyte region. The potential gradient is one of the driving forces for the electromigration of lithium ions. Next, a mass conservation equation is established within the electrolyte region, describing the relationship between the rate of change of lithium ion concentration over time and the ion flux divergence. The ion flux includes a diffusion term (driven by the concentration gradient and following Fick's law) and an electromigration term (driven by the potential gradient and proportional to ion mobility and concentration). The mass conservation equation is discretized on the mesh nodes of the electrolyte region in the three-dimensional geometric model of the battery. The potential gradient data in the potential distribution is used as the input parameter for the electromigration term, and the ion flux at the interface adjacent to the electrolyte in the reaction current density distribution is used as the boundary condition. Numerical solution methods (such as the finite element method or the finite volume method) are then used to solve the discretized algebraic equations to obtain the lithium ion concentration value at each mesh node within the electrolyte region. The equipment organizes the ion concentration values ​​of all grid nodes according to spatial coordinates to form the ion concentration distribution in the electrolyte of the target battery.

[0094] To facilitate understanding, the following examples are provided for illustration, but they do not constitute a specific limitation on this embodiment. For example, continuing with the aforementioned example of calculating the heat generation rate distribution, the processor in the battery management system obtains the calculated potential distribution, which shows a significant potential gradient in the separator region, with the potential decreasing by approximately 0.3 volts from the positive electrode side to the negative electrode side. The processor establishes the mass conservation equation within the electrolyte region, setting the diffusion coefficient to 1.0 × 10⁻⁶. - ¹ 0 The lithium-ion mobility is set to 5.0 × 10⁻⁶ square meters per second. - ¹¹mol·m² / joule·second. The processor applies the reaction current density distribution at the solid-liquid interface as a boundary condition to the interface between the electrolyte and electrode materials. The central region has a higher reaction current density (3.2 mA / cm²), indicating a faster rate of lithium ion embedding from the electrolyte into the negative electrode, leading to a decrease in lithium ion concentration at this interface. The processor uses the finite element method to solve the mass conservation equation, obtaining the lithium ion concentration at each grid node within the electrolyte region. The results show that near the negative electrode, the local lithium ion concentration decreases to 0.8 mol / L due to continuous lithium ion embedding into the negative electrode material; near the positive electrode, the local lithium ion concentration increases to 1.3 mol / L due to lithium ion extraction from the positive electrode material into the electrolyte; and in the central region of the separator, the concentration gradient is relatively gentle, approximately 1.0 mol / L. The processor organizes the ion concentration values ​​of all nodes into a three-dimensional distribution according to spatial coordinates, obtaining an ion concentration distribution that clearly shows the concentration gradient formed in the electrolyte during the migration of lithium ions from the positive to the negative electrode during discharge.

[0095] Step S305: The internal state field is constituted by the potential distribution, the ion concentration distribution, and the heat generation rate distribution.

[0096] It should be explained that the aforementioned internal state field can refer to a multi-dimensional data set describing the spatial distribution of key physical quantities inside the target battery under a unified spatial reference. This data set includes three core components: potential distribution, ion concentration distribution, and heat generation rate distribution, which can comprehensively reflect the spatial characteristics of the battery's internal electrochemical state, mass transport state, and heat source distribution state at a certain moment.

[0097] In the specific implementation, after calculating the potential distribution, ion concentration distribution, and heat generation rate distribution, three spatial distribution datasets describing different physical quantities inside the battery have been obtained. The potential distribution is stored as a three-dimensional array, recording the potential value at each grid node in the battery's three-dimensional geometric model. The ion concentration distribution is stored as a three-dimensional array, recording the lithium-ion concentration value at each grid node in the electrolyte region. The heat generation rate distribution is stored as a three-dimensional array, recording the heat generation rate value at each grid node or grid cell in the battery's three-dimensional geometric model. These three distribution datasets are correlated and integrated to form a unified data structure. Specifically, the device uses the grid node index of the battery's three-dimensional geometric model as a common spatial reference, and correlates and stores the potential value, ion concentration value (if the node is located in the electrolyte region), and heat generation rate value (if the node has a heat source) under the same grid node. For cases where certain physical quantities are not applicable to specific grid nodes (e.g., the ion concentration distribution is not applicable to nodes in the current collector region), the device can store null or default values ​​at that location. The device organizes the integrated data of all grid nodes according to spatial coordinates to form an internal state field.

[0098] To facilitate understanding, the following examples are provided for illustration, but do not impose specific limitations on this embodiment. For example, following the aforementioned examples of calculating the potential distribution, ion concentration distribution, and heat generation rate distribution, the processor in the battery management system has obtained three distribution data: the potential distribution includes the potential values ​​of 100,000 grid nodes, where it is approximately 3.65 volts at the positive electrode current collector and approximately 3.60 volts at the negative electrode current collector; the ion concentration distribution includes the lithium ion concentration values ​​of approximately 30,000 grid nodes in the electrolyte region, where it is approximately 0.8 mol / L near the negative electrode and approximately 1.3 mol / L near the positive electrode; the heat generation rate distribution includes the heat generation rate values ​​of 100,000 grid nodes, where it is approximately 250 mW / cm³ in the separator region, approximately 300 mW / cm³ in the solid-liquid interface region, and approximately 2.5 mW / cm³ in the bulk region of the positive electrode material. The processor uses the grid node index of the battery's 3D geometric model as a reference. For a grid node located at coordinates (20, 30, 25) mm (belonging to the electrolyte region), it associates and stores the node's potential value of 3.62 V, ion concentration value of 1.0 mol / L, and heat generation rate value of 250 mW / cm³. For a grid node located at coordinates (10, 20, 5) mm (belonging to the negative electrode current collector region), it associates and stores the node's potential value of 3.61 V and heat generation rate value of 1.0 mW / cm³, while storing the ion concentration value as empty. After integrating the data from all grid nodes, the processor forms the internal state field for the 1000,000 ms time slice. This state field comprehensively describes the battery's internal potential distribution, lithium ion concentration distribution, and heat source distribution at that moment.

[0099] Based on the first and / or second embodiments of this application, in the third embodiment of this application, the content that is the same as or similar to that in embodiments one and two above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3 , Figure 3 This is a flowchart illustrating Embodiment 3 of the battery state estimation method of this application. To obtain degradation characteristics and stress fields, such as… Figure 3 As shown, in this embodiment, the step of obtaining degradation characteristics and stress field based on the internal state field through a damage evolution model includes: Step S401: Based on the ion concentration distribution and the potential distribution, the material phase interface movement is simulated using the phase field method to obtain the degradation characteristics of material structural damage.

[0100] It should be explained that the aforementioned phase-field method can refer to a numerical calculation method for simulating the evolution of the microstructure of materials. By introducing continuously changing phase-field variables, it describes the interface movement process between different states of the material, thus avoiding the complexity of explicitly tracking the phase interface in traditional methods.

[0101] The aforementioned material phase interface movement can refer to the process by which the interface between the damaged and undamaged areas advances over time as the internal materials of the battery transition from an undamaged state to a damaged state. Examples include the gradual rupture of the solid electrolyte interface film and the gradual expansion of microcracks on the surface of active material particles.

[0102] The degradation characteristics of the aforementioned material structure damage can refer to the spatial distribution information calculated by the phase field method, which is used to quantitatively describe the degree of degradation of the internal material structure of the battery. For example, the connected topology of the ruptured region of the solid electrolyte interface film, and the location and morphology of microcracks in the active material particles.

[0103] In its implementation, the device first obtains the calculated ion concentration and potential distributions from the internal state field, which serve as external driving conditions for the phase-field method. Next, a phase-field model is established in the electrode material region and the solid electrolyte interface film region within the three-dimensional geometric model of the target battery. The device defines phase-field variables, setting the value of the phase-field variable corresponding to undamaged material to 0 and the value of the phase-field variable corresponding to fully damaged material to 1. Continuous changes in the phase-field variable between 0 and 1 represent intermediate states of partial material damage. The device establishes a governing equation containing a chemical potential gradient term and a strain energy density term, where the chemical potential gradient is jointly determined by the ion concentration and potential distributions, and the strain energy density is provided by the stress field. This governing equation drives the evolution of the phase-field variable over time, describing the dynamic process of the material transitioning from an undamaged state to a damaged state. The device uses numerical solution methods (e.g., the finite element method) to discretize and iteratively solve the governing equation, updating the spatial distribution of the phase-field variable at each time step. When the value of the phase-field variable exceeds a preset damage threshold (e.g., 0.8), the device determines that structural damage has occurred at that location. The equipment performs connectivity analysis on all areas identified as damaged according to spatial coordinates, identifies interconnected damaged areas, and extracts the connected areas where the solid electrolyte interface film is ruptured and microcracks appear in the active material particles as degradation characteristics of material structural damage.

[0104] To facilitate understanding, the following examples are provided for illustration, but do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of internal state field construction, the processor in the battery management system acquires the internal state field at the 1000,000th millisecond time slice and extracts the ion concentration distribution and potential distribution from it. The processor establishes a phase field model in the solid electrolyte interface film region on the surface of the negative electrode material, setting the phase field variable value corresponding to the undamaged solid electrolyte interface film to 0 and the phase field variable value corresponding to the completely broken solid electrolyte interface film to 1. The processor establishes the Ginzburg-Landau equation as the governing equation, where the chemical potential gradient term is calculated based on the ion concentration distribution and potential distribution: in regions with low lithium ion concentration and low potential (e.g., the low-temperature region at the edge of the battery, with an ion concentration of 0.85 mol / L and a potential of 3.58 V), the chemical potential gradient is large, driving the phase field variable to evolve rapidly; in regions with high lithium ion concentration and high potential (e.g., the central region of the battery, with an ion concentration of 1.0 mol / L and a potential of 3.62 V), the chemical potential gradient is small, and the phase field variable evolves slowly. The processor uses the finite element method to solve the governing equations, and after iterative calculations, obtains the spatial distribution of phase field variables. The processor marks regions with phase field variable values ​​exceeding 0.8 as damaged areas. The results show that multiple interconnected solid electrolyte interface film rupture regions exist in the low-temperature region at the battery edge, forming a connected damage band extending from coordinates (8, 18, 3) mm to (12, 22, 7) mm. Simultaneously, the processor identifies the initiation locations of multiple microcracks on the surface of the negative electrode active material particles. The processor extracts these connected damaged regions as degradation characteristics of the material structure.

[0105] Step S402: Based on the heat generation rate distribution, obtain the temperature field inside the target battery, and combine it with the thermal expansion coefficients of different materials in the target battery to obtain the thermal stress distribution.

[0106] It should be explained that the temperature field mentioned above can refer to the distribution of temperature values ​​at various spatial locations inside the target battery in three-dimensional space. It is obtained by solving the heat conduction equation from the heat generation rate distribution, reflecting the spatial distribution law of temperature under the combined effect of heat generation and heat diffusion inside the battery.

[0107] The aforementioned coefficient of thermal expansion refers to the relative length change caused by a unit temperature change in a material. Different materials (such as positive electrode materials, negative electrode materials, separators, and current collectors) typically have different coefficients of thermal expansion, and this difference is a significant cause of thermal stress. The aforementioned thermal stress distribution refers to the distribution of internal mechanical stress in three-dimensional space caused by uneven temperature distribution within the battery and mismatches in the coefficients of thermal expansion between different materials, usually measured in Pascals or Megapascals.

[0108] In the specific implementation, the calculated heat generation rate distribution is first obtained from the internal state field. Next, a three-dimensional unsteady-state heat conduction equation is established inside the target battery. This equation describes the relationship between the rate of temperature change over time, the heat generation rate, material thermal properties (e.g., thermal conductivity, density, specific heat capacity), and the temperature gradient divergence. The device discretizes the heat conduction equation at all grid nodes of the battery's three-dimensional geometric model, applying the heat generation rate distribution as a source term to the corresponding grid nodes, and using the convective heat transfer conditions between the battery surface and the external environment as boundary conditions. The device uses numerical solution methods (e.g., the finite element method) to solve the discretized equations, obtaining the temperature value at each grid node in the battery's three-dimensional geometric model. The temperature values ​​of all nodes are organized according to spatial coordinates to form the temperature field inside the target battery. Subsequently, the thermal stress distribution is calculated based on the temperature field and the thermal expansion coefficients of different materials in the target battery. The device first obtains the temperature value at each grid node and the temperature gradient between adjacent nodes from the temperature field to determine the amount of free expansion caused by temperature changes. The device combines the thermal expansion coefficients of different material regions in the target battery (e.g., positive electrode material, negative electrode material, separator, positive electrode current collector, negative electrode current collector) with mechanical parameters such as the elastic modulus and Poisson's ratio of the materials to establish thermoelastic mechanical equations (e.g., thermoelastic constitutive equations and equilibrium equations). The device discretizes and solves the equations on the three-dimensional geometric model of the battery, considering the constraints and stress concentrations caused by the mismatch of thermal expansion coefficients at the interfaces of different materials. Finally, it obtains the stress tensor components on each grid node and organizes the stress data of all nodes according to spatial coordinates to form a thermal stress distribution.

[0109] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, following the example of calculating the heat generation rate distribution, the processor in the battery management system obtains the heat generation rate distribution at the 1000,000th millisecond time slice. The processor establishes a three-dimensional unsteady-state heat conduction equation, inputting the thermal conductivity of the positive electrode material as 2.5 watts per meter per Kelvin, the thermal conductivity of the negative electrode material as 1.8 watts per meter per Kelvin, and the thermal conductivity of the separator as 0.5 watts per meter per Kelvin, with densities of 2500 kg / m³, 2200 kg / m³, and 1200 kg / m³, respectively, and specific heat capacities of 800 joules per kg per Kelvin, 750 joules per kg per Kelvin, and 1000 joules per kg per Kelvin, respectively. The processor sets the convective heat transfer coefficient between the battery surface and the air to 10 watts per square meter per Kelvin, and the ambient temperature to 25 degrees Celsius. The processor uses the finite element method to solve the heat conduction equation and obtain the temperature field distribution. The results show that the temperature is highest in the central region of the battery, approximately 35.2 degrees Celsius, while the temperature is lower at the edge region, approximately 12.5 degrees Celsius. The processor then calculates the thermal stress distribution based on the temperature field, with the thermal expansion coefficient of the positive electrode material being input as 1.2 × 10⁻⁶.-5 The anode material has a per Kelvin elastic modulus of 100 GPa and a Poisson's ratio of 0.22; its coefficient of thermal expansion is 1.5 × 10⁻⁶. -5 The aluminum current collector has a coefficient of thermal expansion of 2.3 × 10⁻⁶ Kelvin, an elastic modulus of 80 GPa, and a Poisson's ratio of 0.25. -5 The copper current collector has a coefficient of thermal expansion of 1.7 × 10⁻⁶ Kelvin, an elastic modulus of 70 GPa, and a Poisson's ratio of 0.33. -5 The parameters are: per Kelvin, elastic modulus 110 GPa, Poisson's ratio 0.34. The processor solves the thermoelastic equations to obtain the thermal stress distribution. The results show that at the interface between the positive and negative electrode materials, due to the large difference in their thermal expansion coefficients and a significant temperature gradient, the local thermal stress peak reaches 35 MPa; at the interface between the negative electrode material and the copper current collector, the thermal stress is approximately 15 MPa; in the low-temperature region at the edge of the battery, the contraction caused by the sudden temperature drop is constrained by the adjacent high-temperature region, generating a tensile stress of approximately 10 MPa. The processor organizes the thermal stress data of all nodes into a three-dimensional distribution according to spatial coordinates, forming the thermal stress distribution.

[0110] Step S403: The thermal stress distribution is superimposed with the lattice strain to obtain the stress field, wherein the lattice strain is the lattice strain inside the target battery caused by lithium ion insertion and extraction.

[0111] It should be explained that the aforementioned lattice strain refers to the strain caused by the change in the lattice constant due to the size effect of lithium ions when they are inserted into or extracted from the electrode material lattice. When lithium ions are inserted into the lattice structure of the positive or negative electrode material, the lattice spacing increases; when lithium ions are extracted, the lattice spacing decreases. This change in lattice size generates mechanical stress within the material.

[0112] The aforementioned stress field can refer to the comprehensive spatial distribution of all mechanical stresses inside the target battery, including thermal stress distribution and stress distribution caused by lattice strain. The two stress sources are superimposed in space to form the final stress field, which is used to assess the risk and extent of structural damage to the battery's internal materials.

[0113] In the specific implementation, the calculated thermal stress distribution is first obtained, which describes the stress values ​​at various spatial locations caused by temperature inhomogeneity and differences in thermal expansion coefficients. Next, the stress components corresponding to the lattice strain caused by lithium-ion insertion and extraction are calculated. The device obtains the ion concentration distribution from the internal state field and determines the local lithium-ion insertion amount in the electrode material based on this distribution. For the positive electrode material, the lattice constant decreases during lithium-ion extraction, resulting in contractile strain; for the negative electrode material, the lattice constant increases during lithium-ion insertion, resulting in expansion strain. The device calculates the lattice strain value at each spatial location based on the chemical expansion coefficient of the electrode material (i.e., the proportion of lattice strain caused by a unit change in lithium-ion concentration) and the change in local lithium-ion concentration relative to a reference concentration. The device multiplies the lattice strain value by the elastic modulus of the material to obtain the stress components caused by the lattice strain. Subsequently, at each mesh node of the battery's three-dimensional geometric model, the stress tensor of the thermal stress distribution is vector-superimposed with the stress components caused by the lattice strain. For regions simultaneously subjected to thermal stress and lattice strain (e.g., the bulk phase regions of cathode and anode materials), the device adds the corresponding components of the two stresses; for regions only affected by lattice strain (e.g., uniform temperature regions unaffected by temperature gradients), the device directly uses the stress component caused by lattice strain. The device organizes the stress data from all superimposed mesh nodes according to spatial coordinates to form the final stress field.

[0114] To facilitate understanding, the following examples are provided for illustration, but do not constitute a specific limitation on this embodiment. For instance, continuing with the aforementioned example of thermal stress distribution calculation, the processor in the battery management system obtains the thermal stress distribution, which shows a peak thermal stress of 35 MPa at the interface between the positive and negative electrode materials. The processor obtains the ion concentration distribution from the internal state field; in the negative electrode material region, the lithium-ion concentration increases from an initial 0.1 mol / L to 0.9 mol / L during discharge, representing an increase of 0.8 mol / L in lithium-ion intercalation. The processor then calculates the ion concentration distribution based on the chemical expansion coefficient of the negative electrode material (e.g., 2.0 × 10⁻⁶). -6 The lattice strain was calculated to be 1.6 × 10⁻⁶ per mole per liter. -³ (i.e., 0.16% expansion strain). The processor multiplies the lattice strain by the elastic modulus of the negative electrode material, 80 GPa, to obtain a compressive stress of approximately 128 MPa caused by the lattice strain. At each grid node in the negative electrode material region, the processor superimposes the thermal stress value of that node in the thermal stress distribution (e.g., the thermal stress at the interface between the negative electrode material and the separator is approximately 10 MPa) with the compressive stress of 128 MPa caused by the lattice strain, obtaining a total stress of approximately 138 MPa (compressive stress) for that node. In the positive electrode material region, the lithium ion concentration decreases from an initial 1.0 mol / L to 0.3 mol / L during the discharge process, with a lithium ion extraction rate of 0.7 mol / L. The lattice strain is a contraction strain. Based on the chemical expansion coefficient and elastic modulus, the tensile stress caused by the lattice strain is calculated to be approximately 70 MPa. After superimposing the thermal stress (approximately 25 MPa) in this region, the total stress is approximately 95 MPa (tensile stress). The processor organizes the superimposed stress of all grid nodes into a three-dimensional distribution according to spatial coordinates to obtain the final stress field. This stress field shows that the expansion strain caused by lithium ion insertion in the negative electrode material region is the main source of stress, with the local total stress peak reaching 138 MPa, located near the interface between the negative electrode material and the separator.

[0115] Furthermore, in order to obtain the battery state estimation result, in this embodiment, the step of obtaining the battery state estimation result based on the degradation characteristics and the stress field includes: Step S501: Discretize the material structure damage in the degradation feature into multiple damage nodes, and take the local stress concentration points in the stress field as stress nodes.

[0116] It should be explained that the above discretization can refer to the process of transforming continuously distributed material structural damage regions (such as connected microcrack bands or solid electrolyte interface film rupture regions) into a series of discrete points with spatial coordinates.

[0117] The aforementioned damage nodes can refer to discrete points extracted from the material structural damage region in the degradation characteristics. Each damage node represents a local damage location and has a clear three-dimensional spatial coordinate attribute. The aforementioned local stress concentration points can refer to spatial locations in the stress field where the stress value is significantly higher than the average level of the surrounding area. They are usually located at material geometric discontinuities, interfaces between different materials, or the tips of the damage region. These locations are high-risk areas for damage initiation and propagation.

[0118] The aforementioned stress nodes can refer to discrete points formed by extracting local stress concentration points in the stress field. Each stress node also has clear three-dimensional spatial coordinates and stress values ​​at that location.

[0119] In the specific implementation, the calculated degradation features are first acquired. These features contain information about the connected regions of material structural damage, such as the connected topology of the solid electrolyte interface membrane rupture region or the spatial distribution of microcracks in active material particles. The damaged regions in the degradation features are then spatially discretized, extracting discrete spatial points from continuous damaged regions at preset sampling intervals (e.g., 0.5 mm or 1 mm). For connected damage bands, the device samples uniformly along the direction of the damage band, generating multiple damage nodes with spatial coordinates. For isolated damage points, the device directly uses the coordinate position of the damage point as a damage node. The device records the spatial coordinates and damage degree parameters (e.g., phase field variable values ​​or crack width) at each damage node. Next, the calculated stress field is acquired, containing stress tensor data at each grid node in the battery's three-dimensional geometric model. All grid nodes in the stress field are traversed. For each grid node, the principal stress or von Mises equivalent stress at that node is calculated, and this stress value is compared with a preset stress threshold (e.g., 30 MPa or 50% of the material's yield strength). When the stress value of a mesh node exceeds a preset stress threshold, the device marks that mesh node as a local stress concentration point. The device extracts the spatial coordinates of all mesh nodes marked as local stress concentration points, using them as stress nodes. For each stress node, the device records its spatial coordinates and the stress value at that node (including the principal stress direction and equivalent stress magnitude). Ultimately, the device obtains a set of damage nodes and a set of stress nodes, which are used together to subsequently construct the internal damage network.

[0120] To facilitate understanding, the following explanation uses examples, but does not limit the scope of this embodiment. For example, following the aforementioned example of degradation features and stress field calculation, the processor in the battery management system acquires a degradation feature containing a connected damage band extending from coordinates (8, 18, 3) mm to (12, 22, 7) mm, with a length of approximately 6 mm and a width of approximately 0.5 mm. The processor extracts a discrete point every 0.5 mm along the direction of the damage band, extracting a total of 12 damage nodes. The coordinates of each damage node are (8.5, 18.5, 3.2) mm, (9.0, 19.0, 3.5) mm, etc., and the phase field variable value at each node is recorded as approximately 0.85. The processor also acquires two other isolated microcrack tip locations in the degradation feature, located at coordinates (15, 30, 20) mm and (25, 20, 35) mm, respectively, and these two locations are also taken as damage nodes, resulting in a total of 14 damage nodes. Next, the processor acquires the stress field, iterating through the von Mises equivalent stress value of each of the 100,000 grid nodes. The processor sets the stress threshold to 30 MPa and filters out grid nodes with equivalent stresses exceeding 30 MPa. The results show that near the interface between the positive and negative electrode materials, 8 grid nodes have equivalent stresses exceeding 35 MPa, with a peak value reaching 50 MPa; near the interface between the negative electrode material and the copper current collector, 5 grid nodes have equivalent stresses exceeding 32 MPa; and near the microcrack tip, 3 grid nodes have equivalent stresses exceeding 45 MPa. The processor extracts the spatial coordinates of these grid nodes, obtaining 16 stress nodes, each recording its coordinates and corresponding equivalent stress value. Ultimately, the processor obtains 14 damage nodes and 16 stress nodes for subsequent construction of the internal damage network.

[0121] Step S502: Establish an internal damage network based on the damaged nodes, the stress nodes, and the material connection relationships between the nodes.

[0122] It should be explained that the material connection relationship between the above nodes can refer to the spatial adjacency or connectivity between damaged nodes and stress nodes, or between nodes of the same type, established through continuous material entities inside the battery. For example, two nodes may be located inside the same electrode material particle and there may be a continuous material path between them, or one node may be located on the solid electrolyte interface film while the other node is located in an adjacent active material particle, and the two are connected through the material interface.

[0123] The aforementioned internal damage network can refer to a graph data structure constructed using damage nodes and stress nodes as network nodes and the material connection relationships between nodes as network edges, to describe the internal damage topology and damage propagation path of a battery. This network can reflect the spatial distribution of damage inside the battery and the potential path of damage spreading from one location to an adjacent location.

[0124] In the specific implementation, the first step is to acquire the existing set of damaged nodes and stress nodes, each with clearly defined three-dimensional spatial coordinates. Next, the three-dimensional geometric model of the target battery is loaded. This model includes the spatial coordinate information of the mesh nodes, the connection relationships of the mesh cells, and the division information of different material regions (e.g., positive electrode material region, negative electrode material region, separator region, solid electrolyte interface film region, and current collector region). For each damaged node and stress node, the device locates the mesh cell to which the node belongs and the material region it is located in the three-dimensional geometric model. Subsequently, the material connection relationships between nodes are established. The device traverses all node pairs, and for each pair, it determines whether a continuous material path exists between the two nodes. Specifically, the device checks whether the two nodes are located within the same material region or in adjacent material regions (e.g., one in the solid electrolyte interface film and the other in the adjacent negative electrode material), and whether the distance between the two nodes is less than a preset connection threshold (e.g., 2 mm or 5 mm). When the two nodes meet the above conditions, the device establishes a network edge between the two nodes and records the attributes of the edge, including the spatial distance between the two nodes, the material type traversed by the connection path, and the average stress level along the path. After traversing and evaluating all node pairs, the device organizes all nodes and the network edges between them into a graph data structure, which is the internal damage network. The device can also label nodes according to their type, distinguishing between damaged nodes, stress nodes, and hybrid nodes that belong to both categories.

[0125] For ease of understanding, the following examples illustrate the concept, but do not limit the scope of this embodiment. For instance, following the example of damage node and stress node extraction described above, the processor in the battery management system acquires the coordinate sets of 14 damage nodes and 16 stress nodes, totaling 30 nodes. The processor loads a three-dimensional geometric model of the battery, containing 100,000 mesh cells, each labeled with its material type (e.g., positive electrode material, negative electrode material, solid electrolyte interface film, separator, aluminum current collector, copper current collector). The processor locates the mesh cell and material region to which each node belongs: 12 damage nodes distributed along the damage band are located in the solid electrolyte interface film region; 2 microcrack tip damage nodes are located in the negative electrode material region; 8 stress nodes are located on the negative electrode material side near the interface between the positive and negative electrode materials; 5 stress nodes are located at the interface between the negative electrode material and the copper current collector; and 3 stress nodes are located near the microcrack tip. The processor sets a connection threshold of 1 mm and traverses all 450 node combinations. For two adjacent damaged nodes located on the solid electrolyte interface membrane damage zone, with a distance of approximately 0.5 mm and both within the solid electrolyte interface membrane region, the processor establishes a network edge between them, recording a distance of 0.5 mm, the material type as solid electrolyte interface membrane, and an average stress of approximately 20 MPa along the path. For a damaged node located on the solid electrolyte interface membrane damage zone and a stress node located in an adjacent negative electrode material region, with a distance of approximately 0.8 mm and connected to the negative electrode material interface through the solid electrolyte interface membrane, the processor establishes a network edge between them, recording a distance of 0.8 mm, the path passing through both the solid electrolyte interface membrane and the negative electrode material, and an average stress of approximately 35 MPa along the path. For two nodes located in the positive electrode material region and the negative electrode material region respectively, separated by a separator region, with a distance exceeding the connection threshold and no direct material connection path, the processor does not establish a network edge between them. After the processor completes all traversals, it establishes approximately 85 network edges, connecting all nodes into an internal damage network. This network clearly demonstrates the diffusion of damage from the solid electrolyte interface film into the interior of the negative electrode material, as well as the spatial correlation between stress concentration areas and damage areas.

[0126] Step S503: Based on the material fracture toughness and interfacial bonding energy, obtain the correlation strength and failure propagation probability between nodes in the internal damage network.

[0127] It should be explained that the fracture toughness of the aforementioned material refers to the material's ability to resist crack propagation, that is, the amount of energy that the material can absorb before fracture. It is usually expressed as the square root of megapascals per meter. The higher the value, the less likely the material is to fracture.

[0128] The aforementioned interfacial bonding energy refers to the bonding strength between two different materials, that is, the energy required to separate the two materials per unit area. It is usually expressed in joules per square meter. The higher the value, the stronger the interfacial bonding and the less likely it is to occur interfacial delamination.

[0129] The aforementioned correlation strength can refer to the degree of interrelation between two nodes in an internal damage network. The higher the correlation strength, the stronger the mechanical and structural interdependence between the two nodes, and the easier it is for damage to one node to affect the other node.

[0130] The failure propagation probability mentioned above refers to the likelihood of damage spreading from one node to an adjacent node. A higher failure propagation probability means that the damage is more likely to propagate from the source node to the target node along the network edge.

[0131] In the specific implementation, the established internal damage network is first acquired. This network contains multiple nodes and network edges between the nodes. Each network edge records the spatial distance connecting two nodes, the material type traversed by the path, and the average stress level along the path. Next, a pre-stored material parameter database is loaded, and the fracture toughness values ​​of different materials (e.g., cathode materials, anode materials, solid electrolyte interfacial membranes, separators, current collectors) and the interfacial binding energy values ​​between different material combinations (e.g., the interfacial binding energy between the solid electrolyte interfacial membrane and the anode material, and the interfacial binding energy between the cathode material and the electrolyte) are retrieved from the database. For each network edge in the internal damage network, the following calculations are performed: The device obtains the type (damaged node or stress node) of the two nodes connected by the network edge and the spatial coordinates of the two nodes. The device determines the critical material on the path based on the material type traversed by the network edge and obtains the fracture toughness value of that material. The device obtains the interfacial binding energy value of the corresponding material combination based on whether the network edge crosses the interface of different materials. The device calculates the association strength value of a network edge using a preset weighted calculation formula, based on the spatial distance between two nodes (closer distance means stronger association), the average stress level on the path (higher stress means stronger association), the fracture toughness of the material (lower fracture toughness means stronger association), and the interfacial bonding energy (lower interfacial bonding energy means stronger association). Then, based on the calculated association strength value and the damage level or stress level at the current node, the device converts the association strength into a failure propagation probability using a preset probability mapping function (e.g., the sigmoid function or a linear mapping function). Higher association strength corresponds to a higher failure propagation probability; when the association strength exceeds a certain threshold, the failure propagation probability approaches 1. The device sequentially calculates the association strength and failure propagation probability for all network edges in the internal damage network, storing the results in the attributes of each network edge.

[0132] To facilitate understanding, the following explanation uses examples, but does not impose specific limitations on this embodiment. For example, continuing with the aforementioned example of establishing an internal damage network, the processor in the battery management system obtains an internal damage network containing 30 nodes and 85 network edges. The processor loads a material parameter database and obtains the fracture toughness of the negative electrode material (graphite) as 0.8 MPa × m², the fracture toughness of the solid electrolyte interfacial membrane as 0.2 MPa × m², and the interfacial bonding energy between the solid electrolyte interfacial membrane and the negative electrode material as 0.5 joules per square meter. For a certain network edge connecting two solid electrolyte interfacial membrane damage nodes, the distance between the two nodes is 0.5 mm, the average stress on the path is 20 MPa, and the path is entirely within the solid electrolyte interfacial membrane. Based on the solid electrolyte interfacial membrane's fracture toughness of 0.2 MPa × m² (low, indicating easy fracture), the node distance of 0.5 mm (short), and the average stress of 20 MPa (medium), the processor calculates the correlation strength of this network edge to be 0.75. The processor maps the correlation strength of 0.75 to a failure propagation probability of 0.85 using the Sigmoid function, indicating that there is an 85% probability that damage will propagate from one damage node to another. For a network edge connecting a solid electrolyte interfacial membrane damage node and a negative electrode material stress node, the distance between the two nodes is 0.8 mm, the path crosses the interface between the solid electrolyte interfacial membrane and the negative electrode material, and the average stress is 35 MPa. Based on the interfacial bonding energy between the solid electrolyte interfacial membrane and the negative electrode material (0.5 joules per square meter, medium), the fracture toughness of the negative electrode material (0.8 MPa multiplied by the square root of meters, high), and the stress level (35 MPa, high), the processor calculates a correlation strength of 0.82, which is mapped to a failure propagation probability of 0.90. For a network edge connecting two negative electrode material stress nodes, the distance between the two nodes is 1.2 mm, the path lies within the negative electrode material, and the average stress is 25 MPa. Based on the square root of the fracture toughness of the negative electrode material (0.8 MPa multiplied by the square root of the meter), a distance of 1.2 mm, and an average stress of 25 MPa, the processor calculates an association strength of 0.35, which maps to a failure propagation probability of 0.30, indicating a low likelihood of damage propagating from one stress node to another. The processor performs calculations on all 85 network edges, storing the association strength and failure propagation probability of each edge in the internal damage network.

[0133] Step S504: Based on the correlation strength and the failure propagation probability, simulate the cascade diffusion path of damage inside the target battery, and obtain the amount of reduction in charge storage capacity for each cascade diffusion.

[0134] It should be explained that the aforementioned cascaded diffusion path can refer to the diffusion trajectory formed by damage spreading from the initial damage location inside the battery along the network edges in the internal damage network to adjacent nodes. This trajectory describes the evolution of damage in time and space, similar to the propagation chain of dominoes falling one after another in the domino effect.

[0135] The aforementioned reduction in charge storage capacity can refer to the decrease in the total amount of charge that the battery can store due to damage to the newly added material structure during each damage propagation event. It is usually expressed in ampere-hours or milliampere-hours and reflects the degree of loss in battery capacity caused by a single damage event.

[0136] In the specific implementation, the device first acquires the established and assigned internal damage network, which includes multiple nodes, network edges between nodes, and the association strength and failure propagation probability of each edge. The device identifies the initial damage node in the internal damage network. This initial damage node can be the node with the most severe material structural damage in the degradation characteristics (e.g., the node with the largest phase field variable value) or the stress node with the highest stress level in the stress field. The device marks the initial damage node as a failed node and uses it as the starting point for cascade diffusion. Next, the iterative process of cascade diffusion simulation is executed. In the current iteration, the device traverses all failed nodes. For each failed node, it finds all adjacent nodes connected to it. For each adjacent node, the device obtains the failure propagation probability of the network edge connecting the failed node and the adjacent node. The device uses a random sampling method (e.g., Monte Carlo method) to generate a random number between 0 and 1 and compares this random number with the failure propagation probability. When the random number is less than or equal to the failure propagation probability, the device determines that the damage has successfully propagated from the failed node to the adjacent node and marks the adjacent node as a newly added failed node. When the random number is greater than the failure propagation probability, the damage has not been propagated to the adjacent node. After the device completes the judgment of all adjacent nodes, it enters the next iteration, using the newly added failed node as the starting point to continue propagation. When no new failed nodes are added in an iteration, the cascade diffusion process stops, and all nodes marked as failed and the network edges between them together constitute the cascade diffusion path. Each time the damage is successfully propagated to an adjacent node, the reduction in charge storage capacity caused by the diffusion event is calculated. The device obtains the volume of the material region represented by the newly failed node, and calculates the volume loss of the active material corresponding to the node based on the type of the material region (e.g., positive electrode material, negative electrode material, solid electrolyte interface film) and the degree of material damage at that location (e.g., phase field variable value). The device multiplies the volume loss of the active material by the theoretical volumetric capacity of the material (the amount of charge that can be stored per unit volume) to obtain the reduction in charge storage capacity caused by the diffusion event. The device records the time sequence and corresponding reduction amount of each diffusion event.

[0137] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, following the example of calculating the correlation strength and failure propagation probability mentioned above, the processor in the battery management system acquires an internal damage network containing 30 nodes and 85 edges, each edge already assigned a failure propagation probability. The processor identifies the initial damage node as a node located at one end of the damage zone of the solid electrolyte interface film, with coordinates (8.5, 18.5, 3.2) mm. The phase field variable value of this node is 0.92, indicating the most severe damage. The processor marks the initial damage node as a failed node and begins a cascaded diffusion simulation. In the first iteration, the processor traverses all neighboring nodes of the initial damage node, finding three neighboring nodes with corresponding failure propagation probabilities of 0.85, 0.90, and 0.30, respectively. The processor generates three random numbers: 0.23, 0.87, and 0.65. For the first adjacent node, the random number 0.23 is less than 0.85, indicating successful damage propagation. For the second adjacent node, the random number 0.87 is less than 0.90, indicating successful damage propagation. For the third adjacent node, the random number 0.65 is greater than 0.30, indicating failed damage propagation. The processor marks the two successfully propagated adjacent nodes as newly added failure nodes and records the first cascade diffusion event. For each newly added failure node, the processor calculates the reduction in charge storage capacity. The first newly added failure node is located at the solid electrolyte interface membrane, with a corresponding material volume of 0.02 cubic millimeters, a phase field variable value of 0.85, and an active material volume loss of 0.017 cubic millimeters. Multiplying this by the theoretical volumetric specific capacity of the solid electrolyte interface membrane, 0.5 ampere-hours per cubic millimeter, yields a reduction of 0.0085 ampere-hours. The second newly added failure node is also located at the solid electrolyte interface membrane, with a reduction of 0.0078 ampere-hours. The first iteration resulted in a total capacity loss of 0.0163 ampere-hours. In the second iteration, the processor started with two newly added failed nodes and continued to propagate to adjacent nodes, adding three more failed nodes, resulting in a total reduction of 0.0245 amp-hours. The simulation continued until the 5th iteration when no new failed nodes were added, at which point the cascading diffusion process stopped. Ultimately, a total of 15 nodes (including the initial damaged node) were marked as failed nodes, forming a cascading diffusion path extending from coordinates (8.5, 18.5, 3.2) mm to (12, 22, 7) mm. The reduction amount corresponding to each diffusion event was recorded as a sequence: [0.0163, 0.0245, 0.0187, 0.0122, 0.0098] amp-hours, with a cumulative total reduction of 0.0815 amp-hours.

[0138] Step S505: Accumulate the attenuation caused by each damage propagation event to obtain the capacity decay trajectory.

[0139] It should be explained that the above-mentioned capacity decay trajectory can refer to the curve or data sequence of the change in the usable capacity of the battery as the number of damage diffusion events or the number of cycles increases after the battery has experienced multiple damage diffusion events. This trajectory describes the dynamic process of battery capacity decay and reflects the quantitative relationship between the accumulation of internal structural damage and the loss of external usable capacity.

[0140] In the specific implementation, the device first acquires all damage diffusion events recorded during the cascade diffusion simulation and their corresponding reductions in charge storage capacity. Each damage diffusion event corresponds to a reduction value, representing the immediate loss of battery capacity caused by that event. The reduction values ​​are then accumulated sequentially according to the chronological order of the damage diffusion events. Specifically, the device uses the reduction value of the first damage diffusion event as the first accumulated value, representing the cumulative capacity loss of the battery after the first event. The device adds the first accumulated value to the reduction value of the second damage diffusion event to obtain the second accumulated value, representing the cumulative capacity loss of the battery after the first two events. The device repeats the above accumulation operation, adding the previous accumulated value to the reduction value of the next damage diffusion event, until all damage diffusion events have been traversed. The device organizes all accumulated values ​​into a sequence according to the order of the events, where each value in the sequence corresponds to the cumulative capacity loss at a specific time point (or a specific event number). The device uses the sequence number of the damage diffusion event or the corresponding cycle number as the x-axis and the cumulative capacity loss as the y-axis to form a capacity decay trajectory. Capacity decay trajectories can be stored as a data list or represented as a curve function.

[0141] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of cascaded diffusion simulation, the processor in the battery management system records the reduction sequence corresponding to five damage diffusion events, as follows: the first event's reduction is 0.0163 AH, the second event's reduction is 0.0245 AH, the third event's reduction is 0.0187 AH, the fourth event's reduction is 0.0122 AH, and the fifth event's reduction is 0.0098 AH. The processor performs cumulative calculations according to the chronological order of the events. The first cumulative value is 0.0163 AH (i.e., 0.0163). The second cumulative value is the first cumulative value of 0.0163 AH plus the second event's reduction of 0.0245 AH, resulting in 0.0408 AH. The third cumulative value is 0.0408 AH plus 0.0187 AH, resulting in 0.0595 AH. The fourth accumulated value is 0.0595 AH plus 0.0122 AH, resulting in 0.0717 AH. The fifth accumulated value is 0.0717 AH plus 0.0098 AH, resulting in 0.0815 AH. The processor organizes these accumulated values ​​into a capacity decay trajectory data sequence according to the event sequence number: [0.0163, 0.0408, 0.0595, 0.0717, 0.0815] AH. This sequence indicates that after the first damage propagation event, the battery has accumulated a capacity loss of 0.0163 AH; after the second event, the cumulative loss is 0.0408 AH; after the third event, the cumulative loss is 0.0595 AH; after the fourth event, the cumulative loss is 0.0717 AH; and after the fifth event, the cumulative loss is 0.0815 AH. The processor stores this capacity decay trajectory in memory for subsequent calculation of battery state estimation results.

[0142] Step S506: Based on the capacity decay trajectory and the initial total charge storage capacity, obtain the battery state estimation result.

[0143] It should be explained that the aforementioned initial total charge storage capacity can refer to the maximum amount of charge that the target battery can store in a brand new state (i.e., without any cycle aging or structural damage), usually measured in ampere-hours (AHs). It is the rated capacity value of the battery at the time of manufacture and serves as a benchmark reference value for assessing battery capacity degradation.

[0144] The battery state estimation results mentioned above can refer to the quantitative assessment information of the current health state of the target battery, which is usually expressed in the form of global health. Global health is the ratio of the current available capacity to the initial total charge storage capacity, reflecting the degree of capacity retention of the battery relative to its brand-new state, and is presented as a percentage.

[0145] In the specific implementation, the calculated capacity decay trajectory is first obtained. This trajectory records the total charge lost by the battery due to accumulated internal structural damage from the initial state to the current simulation moment. Simultaneously, the initial total charge storage capacity of the target battery is acquired. This initial total charge storage capacity can be read from the battery's factory specifications or obtained through capacity calibration tests under standard operating conditions. The cumulative capacity loss value at the current moment is extracted from the capacity decay trajectory; this is the last value in the capacity decay trajectory sequence, representing the total charge lost by the battery up to the current simulation moment. Next, the initial total charge storage capacity is subtracted from the current cumulative capacity loss value to obtain the current available capacity, which represents the maximum charge the battery can store in the current state. Subsequently, the current available capacity is divided by the initial total charge storage capacity to calculate the global health. The global health is a decimal between 0 and 1, which can be multiplied by 100% to convert it to a percentage. The global health is output as the battery state estimation result. The device can also output both the capacity decay trajectory and the global health together as a complete battery state estimation result to guide subsequent decisions of the battery management system.

[0146] To facilitate understanding, the following examples are provided for illustration, but they do not impose specific limitations on this embodiment. For instance, continuing with the aforementioned example of capacity decay trajectory calculation, the processor in the battery management system obtains a capacity decay trajectory sequence of [0.0163, 0.0408, 0.0595, 0.0717, 0.0815] ampere-hours (Ahs), and the cumulative capacity loss at the current moment (after the fifth damage propagation event) is 0.0815 Ahs. The processor obtains the initial total charge storage capacity of the target battery, whose factory rated capacity is 50 Ahs. The processor subtracts the current cumulative capacity loss of 0.0815 Ahs from the initial total charge storage capacity of 50 Ahs, obtaining a current available capacity of 49.9185 Ahs. Then, the processor divides the current available capacity of 49.9185 Ahs by the initial total charge storage capacity of 50 Ahs, calculating a global health of 0.99837, or 99.837%. Since the damage propagation event in the example corresponds to only a small amount of damage in the early stages of the battery's lifespan, the global health is still close to 100%. The processor outputs this global health of 99.837% as the battery state estimation result. If the simulation continues for more cycles, assuming that after 1000 simulation cycles, the capacity decay trajectory shows a cumulative capacity loss of 8.2 Ah, then the current available capacity is 50 Ah minus 8.2 Ah, which equals 41.8 Ah. The global health of 41.8 Ah divided by 50 Ah equals 0.836, or 83.6%. The processor outputs both the global health of 83.6% and the capacity decay trajectory as the battery state estimation result, indicating to the battery management system that the battery has aged to 83.6% of its health level.

[0147] Furthermore, in order to obtain the thermal stress distribution, in this embodiment, the step of obtaining the temperature field inside the target battery based on the heat generation rate distribution, and obtaining the thermal stress distribution by combining the thermal expansion coefficients of different materials in the target battery, includes: Step S4021: Based on the heat generation rate distribution, obtain the temperature field inside the target battery through the three-dimensional unsteady-state heat conduction equation.

[0148] It should be explained that the aforementioned three-dimensional unsteady-state heat conduction equation can refer to a partial differential equation describing the dynamic propagation of heat in three-dimensional space over time. This equation expresses the quantitative relationship between the rate of temperature change with time, the rate of heat generation, the material's thermal properties (thermal conductivity, density, specific heat capacity), and the temperature gradient divergence. Unlike the steady-state heat conduction equation, the unsteady-state equation considers the process of temperature change over time and is suitable for describing the dynamic temperature change of a battery during charging and discharging.

[0149] In the specific implementation, the calculated heat generation rate distribution is first obtained. This distribution records the rate at which heat is generated per unit volume at each spatial location in the battery's three-dimensional geometric model. The three-dimensional geometric model of the target battery is loaded, and corresponding thermal property parameters are set for different material regions in the model (e.g., positive electrode material, negative electrode material, separator, positive electrode current collector, negative electrode current collector, and battery casing), including thermal conductivity (unit: watts per meter per Kelvin), density (unit: kilograms per cubic meter), and specific heat capacity (unit: joules per kilogram per Kelvin). A three-dimensional unsteady-state heat conduction equation is established. The general form of this equation is: density multiplied by specific heat capacity multiplied by the partial derivative of temperature with respect to time equals thermal conductivity multiplied by the divergence of the temperature gradient plus the heat generation rate. This equation is discretized at all mesh nodes of the battery's three-dimensional geometric model, and the continuous partial differential equation is transformed into a discrete system of algebraic equations using the finite element method or finite volume method. The heat generation rate distribution is applied as a source term to the corresponding grid nodes. Convective heat transfer conditions between the battery surface and the external environment are used as boundary conditions (e.g., setting the convective heat transfer coefficient between the battery surface and air and the ambient temperature). The initial temperature distribution of the battery is used as the initial condition (e.g., the initial temperature equals the ambient temperature). A numerical time integration method (e.g., implicit Euler method or Runge-Kutta method) is employed to progressively solve the problem along the time axis. Within each time step, the discretized algebraic equations are solved to obtain the temperature value at each grid node in the current time step of the battery's three-dimensional geometric model. The temperature values ​​of all grid nodes are organized according to spatial coordinates to form the three-dimensional temperature field inside the target battery at the current time step. If it is necessary to describe the temperature change process over time, the device can store temperature field sequences under multiple time slices.

[0150] To facilitate understanding, the following examples are provided for illustration, but they do not constitute a specific limitation on this embodiment. For example, continuing with the aforementioned example of heat generation rate distribution calculation, the processor in the battery management system obtains the heat generation rate distribution at the 1000.000th millisecond time slice. This distribution shows that the heat generation rate in the separator region is approximately 250 milliwatts per cubic centimeter, and in the solid-liquid interface region, it is approximately 300 milliwatts per cubic centimeter. The processor loads the three-dimensional geometric model of the battery, setting the thermal conductivity of the positive electrode material to 2.5 watts per meter per Kelvin, the density to 2500 kJ per cubic meter, and the specific heat capacity to 800 joules per kilogram per Kelvin; setting the thermal conductivity of the negative electrode material to 1.8 watts per meter per Kelvin, the density to 2200 kJ per cubic meter, and the specific heat capacity to 750 joules per kilogram per Kelvin; and setting the thermal conductivity of the separator to 0.5 watts per meter per Kelvin. The thermal conductivity of the current collector is set to 237 watts per meter per Kelvin, density to 1200 kg / m³, and specific heat capacity to 1000 joules per kg / Kelvin. For the aluminum current collector, the thermal conductivity is set to 237 watts per meter per Kelvin, density to 2700 kg / m³, and specific heat capacity to 900 joules per kg / Kelvin. For the copper current collector, the thermal conductivity is set to 401 watts per meter per Kelvin, density to 8960 kg / m³, and specific heat capacity to 385 joules per kg / Kelvin. The processor establishes a three-dimensional unsteady-state heat conduction equation, setting the convective heat transfer coefficient between the battery surface and air to 10 watts per square meter per Kelvin, the ambient temperature to 25 degrees Celsius, and the initial battery temperature to 25 degrees Celsius. The processor discretizes the equations using the finite element method, establishing a system of algebraic equations on 100,000 grid nodes, and solves them using the implicit Euler method with a time step of 1 millisecond. The results show that at 1000.000 milliseconds, the temperature in the central region of the battery rises to 35.2 degrees Celsius. The separator region, due to its high heat generation rate and low thermal conductivity, reaches a local temperature of 36.5 degrees Celsius. The battery edge region, being close to the outer casing, has better heat dissipation conditions, resulting in a temperature of only 28.3 degrees Celsius. The electrode connection area, due to the high thermal conductivity of the metal current collector, has a relatively uniform temperature of approximately 32.0 degrees Celsius. The processor organizes the temperature values ​​of all nodes into a three-dimensional temperature field according to spatial coordinates. This temperature field clearly demonstrates the temperature gradient inside the battery caused by the uneven distribution of heat generation rate and differences in heat dissipation conditions.

[0151] Step S4022: Based on the temperature field and the thermal expansion coefficients of different materials in the target battery, the thermal stress distribution is obtained through the thermoelastic mechanics equation. The different materials include the positive electrode material, the negative electrode material, and the current collector.

[0152] It should be explained that the above thermoelastic equations can refer to a combination of constitutive equations and equilibrium equations that describe the relationship between strain and stress when an elastic body generates thermal stress under the action of temperature changes. They usually include the generalized Hooke's law (where strain includes both thermal strain and elastic strain) and stress equilibrium equations, which are used to calculate the internal stress distribution caused by temperature changes and non-uniform thermal expansion.

[0153] The aforementioned coefficient of thermal expansion refers to the relative length change caused by a unit temperature change in a material. Different materials have different coefficients of thermal expansion, and this difference is an important cause of thermal stress. The aforementioned thermal stress distribution refers to the distribution of internal mechanical stress in three-dimensional space caused by uneven temperature distribution inside the battery and mismatch in the coefficients of thermal expansion between different materials. It is usually expressed in Pascals or Megapascals and includes normal stress and shear stress components in all directions.

[0154] In the specific implementation, the calculated temperature field is first acquired, which records the temperature value of each grid node in the battery's three-dimensional geometric model at the current moment. The three-dimensional geometric model of the target battery is loaded, and corresponding mechanical parameters are set for different material regions in the model, including the coefficient of thermal expansion (unit: per Kelvin), the elastic modulus (unit: Pascal), and Poisson's ratio (dimensionless). For the positive electrode material, negative electrode material, and current collectors (including aluminum and copper current collectors), the device sets their respective coefficients of thermal expansion and mechanical parameters. Thermoelastic mechanical equations are then established. First, the device calculates the thermal strain at each spatial location based on the temperature field; the thermal strain equals the coefficient of thermal expansion multiplied by the temperature change (current temperature minus reference temperature). For isotropic materials, the normal strain components in the three directions equal the thermal strain, and the shear strain component is zero. Next, based on the relationship between total strain and elastic strain, the device separates the thermal strain from the total strain to obtain the elastic strain. Then, based on the generalized Hooke's law, the device calculates the stress components from the elastic strain. Finally, the device establishes stress balance equations to ensure that stresses satisfy mechanical equilibrium conditions within the material, and applies appropriate boundary conditions (e.g., the battery casing surface is a free boundary with no external forces). The device uses the finite element method to discretize and solve the thermoelastic equations at all mesh nodes of the battery's three-dimensional geometric model, obtaining the stress tensor components at each mesh node, including three normal stress components and three shear stress components. The device organizes the stress data from all nodes according to spatial coordinates to form a thermal stress distribution. The device can also calculate equivalent stresses (e.g., von Mises stress) based on the stress tensor to assess the risk of material failure.

[0155] To facilitate understanding, the following examples are provided for illustration, but they do not constitute a specific limitation on this embodiment. For instance, continuing with the aforementioned example of temperature field calculation, the processor in the battery management system acquires the three-dimensional temperature field at the 1000.000th millisecond. This temperature field shows that the temperature in the central region of the battery is approximately 35.2 degrees Celsius, and the temperature in the edge region is approximately 28.3 degrees Celsius. The reference temperature is set to 25 degrees Celsius. The processor loads a material mechanical parameter database and sets the coefficient of thermal expansion for the cathode material (e.g., lithium cobalt oxide) to 1.2 × 10⁻⁶. -5The parameters are: per Kelvin, elastic modulus of 100 GPa, Poisson's ratio of 0.22; and a coefficient of thermal expansion of 1.5 × 10⁻⁶ for the negative electrode material (e.g., graphite). -5 The coefficient of thermal expansion for the aluminum current collector is set at 2.3 × 10⁻⁶ Kelvin, with an elastic modulus of 80 GPa and a Poisson's ratio of 0.25. -5 The coefficient of thermal expansion for the copper current collector is set at 1.7 × 10⁻⁶ per Kelvin, with an elastic modulus of 70 GPa and a Poisson's ratio of 0.33. -5 The diaphragm has a coefficient of thermal expansion of 5.0 × 10⁻⁶ Kelvin, an elastic modulus of 110 GPa, and a Poisson's ratio of 0.34. -5 The parameters are: Kelvin, elastic modulus 0.2 GPa, Poisson's ratio 0.40. The processor establishes the thermoelastic equations to calculate the thermal strain at each grid node. For a grid node located in the central region of the cathode material, the temperature change is 10.2 degrees Celsius (35.2 minus 25), and the thermal strain is 1.2 × 10⁻⁶. -5 Multiplying by 10.2 equals 1.22 × 10 -4 For a grid node located at the edge of the negative electrode material, the temperature change is 3.3 degrees Celsius (28.3 minus 25), and the thermal strain is 1.5 × 10⁻⁶. -5 Multiplying by 3.3 equals 4.95 × 10 -5 The processor uses the finite element method to solve the stress components at each mesh node based on the generalized Hooke's law and stress balance equations. The results show that at the interface between the positive and negative electrode materials, due to the difference in their coefficients of thermal expansion (positive electrode 1.2 × 10⁻⁶), stress components are generated. -5 negative electrode 1.5×10 -5 The temperature gradient (higher temperature in the central region and lower temperature in the peripheral region) generates significant compressive stress, with the local von Mises stress peak reaching 35 MPa. At the interface between the aluminum current collector and the cathode material, the thermal expansion coefficient of aluminum (2.3 × 10⁻⁶) is high. -5 The efficiency is significantly higher than that of the cathode material (1.2 × 10⁻⁶). -5 As the temperature rises, the aluminum current collector expands more, generating compressive stress due to the constraint of the cathode material, with a stress of approximately 18 MPa at the interface. In the battery edge region, due to the lower temperature and constraint from the high-temperature central region, a tensile stress of approximately 10 MPa is generated. The processor organizes the stress data of all nodes into a three-dimensional thermal stress distribution according to spatial coordinates. This distribution clearly demonstrates the spatial distribution law of thermal stress inside the battery caused by the combined effects of temperature non-uniformity and the mismatch of material thermal expansion coefficients.

[0156] This application also provides a battery state estimation device, please refer to... Figure 4 The device includes: Module 10 is used to acquire the raw running dataset of the target battery; The first processing module 20 is used to perform spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field; The second processing module 30 is used to perform multi-physics field coupling reconstruction based on the running state field to obtain the internal state field. The third processing module 40 is used to obtain degradation characteristics and stress field based on the internal state field through a damage evolution model; The state estimation module 50 is used to obtain the battery state estimation result based on the degradation characteristics and the stress field.

[0157] The battery state estimation device provided in this application, employing the battery state estimation method in the above embodiments, can solve the technical problem of how to sense wide-temperature-range interference and accurately estimate battery state. Compared with the prior art, the beneficial effects of the battery state estimation device provided in this application are the same as those of the battery state estimation method provided in the above embodiments, and other technical features in the battery state estimation device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0158] This application provides a battery state estimation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the battery state estimation method in the above embodiment 1.

[0159] The following is for reference. Figure 5 This document illustrates a structural schematic diagram of a battery state estimation device suitable for implementing embodiments of this application. The battery state estimation device in these embodiments may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 5 The battery state estimation device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0160] like Figure 5As shown, the battery state estimation device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the battery state estimation device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. The communication device 1009 allows the battery state estimation device to communicate wirelessly or wiredly with other devices to exchange data. Although the figures show battery state estimation devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0161] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0162] The battery state estimation device provided in this application, employing the battery state estimation method in the above embodiments, can solve the technical problem of how to sense wide-temperature-range interference and accurately estimate battery state. Compared with the prior art, the beneficial effects of the battery state estimation device provided in this application are the same as those of the battery state estimation method provided in the above embodiments, and other technical features in this battery state estimation device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0163] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0164] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0165] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the battery state estimation method in the above embodiments.

[0166] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0167] The aforementioned computer-readable storage medium may be included in the battery state estimation device; or it may exist independently and not assembled into the battery state estimation device.

[0168] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the battery state estimation device, the battery state estimation device performs the following: acquires the original operating dataset of the target battery; performs spatiotemporal alignment and data fusion on the original operating dataset to obtain a spatiotemporally synchronized operating state field; performs multiphysics coupling reconstruction based on the operating state field to obtain an internal state field; obtains degradation features and stress field through a damage evolution model based on the internal state field; and obtains the battery state estimation result based on the degradation features and stress field.

[0169] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0170] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0171] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0172] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described battery state estimation method, which can solve the technical problem of how to sense wide-temperature-range interference and accurately estimate battery state. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the battery state estimation method provided in the above embodiments, and will not be repeated here.

[0173] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the battery state estimation method described above.

[0174] The computer program product provided in this application can solve the technical problem of how to sense wide-temperature-range interference and accurately estimate battery state. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the battery state estimation method provided in the above embodiments, and will not be repeated here.

[0175] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A battery state estimation method, characterized in that, The method is applied to a battery management system, and the method includes: Obtain the raw runtime dataset for the target battery; The original running dataset is spatiotemporally aligned and fused to obtain a spatiotemporally synchronized running state field; Based on the aforementioned operating state field, multi-physics field coupling reconstruction is performed to obtain the internal state field; Based on the internal state field, degradation characteristics and stress field are obtained through a damage evolution model; Based on the degradation characteristics and the stress field, the battery state estimation results are obtained; The steps of performing spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field include: The data points in the original running dataset are labeled with timestamps and spatial coordinates to obtain a dataset with spatiotemporal identifiers; Using a unified time axis as a reference, the voltage and current records carrying timestamps in the dataset are resampled, and the resampled voltage and current records are aligned with the multi-point temperature record time series in the dataset. The time-aligned multi-point temperature records are mapped onto the grid nodes in the battery's three-dimensional geometric model that correspond to the spatial coordinates, and spatial interpolation is performed on the remaining grid nodes to obtain the temperature field. The operating state field is composed of the voltage record and the current record, which are aligned with the temperature field and time.

2. The method as described in claim 1, characterized in that, The step of performing multiphysics coupling reconstruction based on the operating state field to obtain the internal state field includes: Based on the operating state field and the principle of electrochemical reaction kinetics, the reaction current density distribution at the solid-liquid interface inside the target battery is obtained. Based on the reaction current density distribution and Ohm's law, the potential distribution over the entire target battery is obtained; Based on the aforementioned reaction current density distribution and the principles of electrochemical thermodynamics, the heat generation rate distribution of electrochemical reaction heat and Joule heat is obtained. By combining the potential distribution with the mass conservation equation in the electrolyte of the target battery, the ion concentration distribution is obtained; The internal state field is constituted by the potential distribution, the ion concentration distribution, and the heat generation rate distribution.

3. The method as described in claim 2, characterized in that, The steps of obtaining degradation characteristics and stress field based on the internal state field through the damage evolution model include: Based on the ion concentration distribution and the potential distribution, the material phase interface movement is simulated using the phase field method to obtain the degradation characteristics of material structural damage. Based on the heat generation rate distribution, the temperature field inside the target battery is obtained, and the thermal stress distribution is obtained by combining the thermal expansion coefficients of different materials in the target battery. The stress field is obtained by superimposing the thermal stress distribution with the lattice strain, where the lattice strain is the lattice strain inside the target battery caused by lithium ion insertion and extraction.

4. The method as described in claim 3, characterized in that, The step of obtaining the battery state estimation result based on the degradation characteristics and the stress field includes: The material structural damage in the degradation feature is discretized into multiple damage nodes, and the local stress concentration points in the stress field are taken as stress nodes. An internal damage network is established based on the damage nodes, the stress nodes, and the material connection relationships between the nodes. Based on the material fracture toughness and interfacial bonding energy, the correlation strength and failure propagation probability between nodes in the internal damage network are obtained. Based on the correlation strength and the failure propagation probability, the cascade diffusion path of damage inside the target battery is simulated, and the amount of reduction in charge storage capacity by each cascade diffusion is obtained. The capacity decay trajectory is obtained by summing up the attenuation caused by each damage propagation event. Based on the capacity decay trajectory and the initial total charge storage capacity, the battery state estimation result is obtained.

5. The method as described in claim 3, characterized in that, The steps of obtaining the temperature field inside the target battery based on the heat generation rate distribution, and obtaining the thermal stress distribution by combining the thermal expansion coefficients of different materials in the target battery, include: Based on the heat generation rate distribution, the temperature field inside the target battery is obtained through a three-dimensional unsteady-state heat conduction equation. Based on the temperature field and the thermal expansion coefficients of different materials in the target battery, the thermal stress distribution is obtained through the thermoelastic mechanics equation. The different materials include positive electrode material, negative electrode material, and current collector.

6. A battery state estimation device, characterized in that, The device includes: The acquisition module is used to acquire the raw runtime dataset of the target battery; The first processing module is used to perform spatiotemporal alignment and data fusion on the original running dataset to obtain a spatiotemporally synchronized running state field; The second processing module is used to perform multi-physics field coupling reconstruction based on the operating state field to obtain the internal state field; The third processing module is used to obtain degradation characteristics and stress field based on the internal state field through a damage evolution model; The state estimation module is used to obtain the battery state estimation result based on the degradation characteristics and the stress field; The second processing module marks the data points in the original running dataset with timestamps and spatial coordinates to obtain a dataset with spatiotemporal identifiers; using a unified time axis as a reference, it resamples the voltage and current records carrying timestamps in the dataset, and aligns the resampled voltage and current records with the time series of multi-point temperature records in the dataset; it maps the time-aligned multi-point temperature records onto the grid nodes in the battery's three-dimensional geometric model corresponding to the spatial coordinates, and performs spatial interpolation on the remaining grid nodes to obtain a temperature field; the temperature field, together with the time-aligned voltage and current records, constitutes the running state field.

7. A battery state estimation device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the battery state estimation method as described in any one of claims 1 to 5.

8. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the battery state estimation method as described in any one of claims 1 to 5.

9. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the battery state estimation method as described in any one of claims 1 to 5.

Images