Lithium battery internal short circuit diagnosis method and system based on impedance spectroscopy analysis

Abstract

The invention relates to the technical field of lithium battery safety diagnosis, in particular to a lithium battery internal short circuit diagnosis method and system based on impedance spectroscopy analysis, and the system comprises a data collection module, an impedance field reconstruction module, a short circuit diagnosis module, an environment compensation module and a strategy generation module. The system performs three-dimensional impedance distribution reconstruction by acquiring dynamic impedance spectrum data of a battery, and extracts ohmic impedance gradient, charge transfer impedance cumulant and diffusion impedance fluctuation coefficient as internal impedance characteristics. And performing nonlinear anomaly analysis on the features based on a pre-training model to generate a short-circuit risk level and a potential short-circuit region identifier. And after working condition compensation correction is carried out in combination with the temperature monitoring data, a safety coping strategy including current path adjustment and thermal management intervention is generated according to the region identifier. According to the technology, accurate positioning and active directional prevention and control of internal short-circuit early hidden dangers are realized.

Description

Technical Field

[0001] This invention relates to the field of lithium battery safety diagnostic technology, and in particular to a method and system for diagnosing internal short circuits in lithium batteries based on impedance spectroscopy analysis. Background Technology

[0002] Current methods for diagnosing internal short circuits in lithium-ion batteries primarily rely on macroscopic voltage and temperature monitoring by the battery management system, or on equivalent circuit model analysis based on electrochemical impedance spectroscopy. These methods treat the battery as a uniform unit by measuring the overall voltage response at the battery ports, aiming to fit global impedance parameters as an indicator of its health. However, these techniques are slow to respond to early, highly localized anomalies occurring within the battery. Changes in global parameters often lag behind the actual development of local defects, making it difficult to provide effective early warnings before internal short circuits trigger thermal runaway.

[0003] The limitations of existing diagnostic technologies lie in their insufficient spatial resolution. Internal micro-short circuits originate from failures in tiny regions such as electrode interfaces or separators, resulting in small and spatially specific changes in electrochemical signals. Conventional methods cannot resolve the spatial distribution differences of impedance parameters within the battery, leading to insensitivity to early, localized impedance changes and hindering fault localization.

[0004] Another drawback of existing solutions is that diagnostic results are easily affected by environmental and operating conditions, and lack precise control responses. Battery impedance characteristics are highly dependent on temperature and operating status, and existing methods lack effective online dynamic compensation mechanisms, which can easily lead to misjudgments. Furthermore, even if an anomaly is detected, existing systems typically only trigger general protection measures such as power reduction or global alarms, failing to implement differentiated and targeted interventions based on the specific spatial location of the fault, resulting in insufficient control efficiency and accuracy. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and to propose a method and system for diagnosing internal short circuits in lithium batteries based on impedance spectroscopy analysis.

[0006] To achieve the above objectives, the present invention employs the following technical solution: a lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis, comprising:

[0007] The data acquisition module is used to acquire the dynamic impedance spectrum data set of the target lithium battery during the charge and discharge cycle process. The dynamic impedance spectrum data set includes multi-frequency excitation signal sequence, voltage response waveform sequence and temperature distribution monitoring data.

[0008] The impedance field reconstruction module is used to perform three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set to generate the internal impedance characteristics of the target lithium battery. The internal impedance characteristics include the ohmic impedance gradient, the charge transfer impedance accumulation, and the diffusion impedance fluctuation coefficient.

[0009] The short-circuit diagnosis module is used to call a pre-trained short-circuit identification model to perform nonlinear anomaly analysis on the internal impedance characteristics, and generate the short-circuit risk level and potential short-circuit area identification of the target lithium battery.

[0010] An environmental compensation module is used to perform battery condition compensation and correction processing on the short-circuit risk level to generate a corrected short-circuit risk level. The battery condition compensation and correction processing is based on the correlation between the temperature distribution monitoring data and the electrochemical impedance temperature.

[0011] The strategy generation module is used to generate a set of safety response strategies based on the potential short-circuit area identifier. The set of safety response strategies includes current path adjustment schemes and thermal management intervention schemes.

[0012] As a further aspect of the present invention, the impedance field reconstruction module performs three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set to generate the internal impedance characteristics of the target lithium battery, including:

[0013] The multi-frequency excitation signal sequence is divided into multiple excitation sub-bands according to the frequency band range, and each excitation sub-band corresponds to an impedance sampling interval;

[0014] For each of the aforementioned excitation sub-bands, the following processing is performed:

[0015] The three-dimensional potential distribution topology of the target lithium battery is constructed based on the voltage response waveform sequence. The three-dimensional potential distribution topology includes spatial variation data of the anode potential field, cathode potential field, and electrolyte potential field.

[0016] The three-dimensional potential distribution topology is coupled with the excitation sub-band for analysis to generate the impedance distribution reconstruction result of the current sampling interval. The impedance distribution reconstruction result includes the spatial distribution matrix of the ohmic impedance component, the charge transfer impedance component and the diffusion impedance component.

[0017] The impedance distribution reconstruction results from multiple consecutive sampling intervals are subjected to frequency domain fusion processing to calculate the ohmic impedance gradient, charge transfer impedance accumulation, and diffusion impedance fluctuation coefficient; wherein,

[0018] The ohmic impedance gradient is the rate of change of the ohmic impedance component along the battery thickness direction.

[0019] The cumulative charge transfer impedance is the integral of the charge transfer impedance component along the normal direction of the electrode interface.

[0020] The diffusion impedance fluctuation coefficient is the ratio of the standard deviation to the mean of the diffusion impedance component within a specified frequency range.

[0021] As a further aspect of the present invention, the impedance field reconstruction module performs coupled analysis on the three-dimensional potential distribution topology and the excitation sub-band to generate the impedance distribution reconstruction result for the current sampling interval, including:

[0022] Based on the correspondence between the anode potential field and the high-frequency excitation signal in the excitation sub-band, an anode potential-excitation mapping equation is established, and the distribution function of the ohmic impedance component is obtained by solving the anode potential-excitation mapping equation.

[0023] Based on the correlation characteristics between the cathode potential field and the low-frequency excitation signal, a charge transfer impedance calculation model is constructed. The charge transfer impedance calculation model includes dynamic calibration parameters of the active area of ​​the electrode material and the interface capacitance.

[0024] By combining the spatial gradient of the electrolyte potential field and the ionic conductivity, an iterative calculation process for diffusion impedance is established, which includes a feedback correction mechanism for potential increment and diffusion impedance increment.

[0025] The output results of the distribution function, the charge transfer impedance calculation model, and the diffusion impedance iterative calculation process are subjected to spatial grid fusion processing to generate three-dimensional impedance distribution data containing ohmic impedance, charge transfer impedance, and diffusion impedance.

[0026] As a further aspect of the present invention, the short-circuit diagnosis module calls a pre-trained short-circuit identification model to perform nonlinear anomaly analysis on the internal impedance characteristics, generating the short-circuit risk level and potential short-circuit region identifier of the target lithium battery, including:

[0027] The ohmic impedance gradient is input into the impedance anomaly detection layer of the short-circuit identification model, and the spatial coordinates and impedance change trajectory of the ohmic impedance anomaly region are determined by the gradient mutation identification algorithm.

[0028] The accumulated charge transfer impedance is input into the interface state analysis layer of the short-circuit identification model to perform charge transfer barrier accumulation calculation and generate the interface reactivity degradation probability and reaction rate decay prediction value.

[0029] The diffusion impedance fluctuation coefficient is input into the mass transfer analysis layer of the short-circuit identification model, and the electrolyte ion diffusion coefficient and concentration polarization evolution data are calculated based on the ion migration retardation model.

[0030] By integrating the impedance change trajectory, the interface reaction activity degradation probability, and the ion diffusion coefficient, a comprehensive anomaly index for the target lithium battery is generated, and the short-circuit risk level is determined based on the comparison result between the comprehensive anomaly index and the preset safety threshold.

[0031] Based on the spatial superposition of the spatial coordinates, the predicted reaction rate decay value, and the concentration polarization evolution data, the geometric boundaries of the impedance anomaly region, the interface degradation region, and the mass transfer impeded region are identified.

[0032] As a further aspect of the present invention, the environmental compensation module performs battery condition compensation correction processing on the short-circuit risk level to generate a corrected short-circuit risk level, including:

[0033] Extract the extreme temperatures and temperature change gradients from the temperature distribution monitoring data, and calculate the dynamic compensation amount of the electrochemical impedance temperature coefficient as a function of temperature.

[0034] The temperature drift compensation calculation is performed on the ohmic impedance gradient based on the dynamic compensation amount to generate the corrected ohmic impedance gradient.

[0035] Based on the correlation between the temperature change gradient and electrode reaction kinetics, the cumulative charge transfer impedance is modified by the reaction rate to generate a modified cumulative charge transfer impedance.

[0036] Based on the characteristics of electrolyte conductivity change at extreme temperatures, the diffusion impedance fluctuation coefficient is adjusted by ion migration adaptation to generate a corrected diffusion impedance fluctuation coefficient.

[0037] The corrected ohmic impedance gradient, charge transfer impedance accumulation, and diffusion impedance fluctuation coefficient are input into the short-circuit identification model for recalculation, generating the short-circuit risk level after compensation for operating conditions.

[0038] As a further aspect of the present invention, the environmental compensation module performs temperature drift compensation calculations on the ohmic impedance gradient based on the dynamic compensation amount to generate a corrected ohmic impedance gradient, including:

[0039] Obtain the reference impedance temperature coefficient of the target lithium battery at the reference temperature and the dynamic compensation amount, and establish an impedance temperature coefficient-temperature correlation function;

[0040] The impedance temperature drift is calculated based on the impedance temperature coefficient-temperature correlation function, whereby the impedance temperature drift is the product of the temperature change and the impedance temperature coefficient change.

[0041] The impedance temperature drift is superimposed on the calculation of the ohmic impedance gradient to generate an ohmic impedance gradient correction value that includes temperature compensation effect.

[0042] The ohmic impedance gradient correction value is subjected to relaxation effect compensation processing, which is based on the product factor of the electrochemical relaxation time constant and the temperature holding time.

[0043] As a further aspect of the present invention, the strategy generation module generates a set of security response strategies based on the potential short-circuit region identifier, including:

[0044] For the identification of the impedance abnormality region, the optimal current path adjustment scheme is calculated. The optimal current path adjustment scheme is achieved by adjusting the current distribution ratio of the parallel branches inside the battery.

[0045] Based on the identification of the interface degradation area, a thermal management intervention scheme is constructed, which includes the selection of local cooling areas and the optimized configuration of cooling intensity parameters;

[0046] Based on the identification of the mass transfer obstruction region, an electrolyte replenishment strategy is generated. The electrolyte replenishment strategy dynamically adjusts the replenishment dose and replenishment location according to the predicted ion migration rate.

[0047] The optimal current path adjustment scheme, the thermal management intervention scheme, and the electrolyte replenishment strategy are prioritized to generate a set of safety response strategies that include execution order and control parameters.

[0048] As a further aspect of the present invention, the strategy generation module constructs a thermal management intervention scheme, including:

[0049] Extract the thermal characteristic parameters of the interface degradation region, and calculate the heat generation rate and thermal diffusivity.

[0050] The flow rate of the cooling medium is selected based on the heat generation rate, and the flow rate is directly proportional to the heat generation rate.

[0051] The arrangement density of the cooling channels is adjusted based on the thermal diffusivity to achieve a dynamic balance between cooling efficiency and heat accumulation rate.

[0052] The power of the semiconductor cooling chip is dynamically adjusted based on real-time temperature monitoring data to ensure that the interface temperature is maintained below the material phase transition point.

[0053] Generate a configuration table of thermal management parameters, including cooling medium flow rate, channel layout density, and cooling power.

[0054] As a further aspect of the present invention, the system further includes:

[0055] The model optimization module is used to collect actual impedance change data and thermal runaway characteristic parameters of the target lithium battery within a preset verification period.

[0056] The actual impedance change data and the predicted impedance distribution are subjected to deviation analysis to generate the first calibration coefficient.

[0057] The thermal runaway characteristic parameters are compared with the predicted risk level in the time domain to generate a second calibration coefficient.

[0058] The decision threshold of the short-circuit identification model is adjusted according to the first calibration coefficient and the second calibration coefficient to generate an optimized short-circuit identification model.

[0059] The optimized short-circuit identification model was then applied to subsequent batches of lithium battery internal short-circuit diagnosis tasks.

[0060] As a further aspect of the present invention, the present invention also includes a lithium battery internal short-circuit method based on impedance spectroscopy analysis, the method comprising all the modules and method flow of the lithium battery internal short-circuit system based on impedance spectroscopy analysis as described above.

[0061] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0062] By reconstructing the three-dimensional impedance distribution, the dynamic response data under multi-frequency excitation is transformed into a spatial distribution map of the battery's internal impedance parameters. Features such as the ohmic impedance gradient, cumulative charge transfer impedance, and diffusion impedance fluctuation coefficient are quantified and extracted from this three-dimensional field. These features directly reflect the rate of change, cumulative distribution, and fluctuation of impedance parameters in the spatial dimension, and are highly sensitive to minute heterogeneities within the battery caused by dendrite growth, interfacial side reactions, or localized material failures. This enables the system to capture early and spatially localized electrochemical anomalies that are difficult to detect in the overall impedance signal, achieving a shift from overall health assessment to early localized defect localization.

[0063] An environmental compensation correction process based on the temperature correlation of electrochemical impedance is introduced to dynamically remove the background effect of temperature changes on impedance characteristics. This process, based on the inherent laws of electrochemical kinetics, distinguishes between normal operating condition fluctuations and actual fault signals, directly improving the accuracy and reliability of short-circuit risk level assessment under different environments. Based on the identified potential short-circuit area markers, corresponding current path adjustment schemes and thermal management intervention schemes are directly generated. Current path adjustments can isolate or bypass risky units, while thermal management interventions can provide targeted and enhanced cooling to the identified areas. This precise spatial correspondence between diagnostic results and execution strategies constitutes a closed-loop control from anomaly identification to targeted suppression, effectively curbing the spread of localized faults. Attached Figure Description

[0064] Figure 1 This is a timing diagram of the lithium battery internal short circuit diagnosis system based on impedance spectroscopy analysis described in this invention.

[0065] Figure 2 A flowchart for generating impedance distribution reconstruction results in coupled analysis;

[0066] Figure 3 A thermal map showing the spatial distribution of the probability of degradation of interfacial reactivity in lithium batteries;

[0067] Figure 4 A bar chart comparing the short-circuit risk levels of different regions of a lithium battery before and after temperature compensation.

[0068] Figure 5 A bar chart comparing the current distribution ratio before and after the adjustment of the lithium battery safety response strategy. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0070] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0071] See Figure 1 The data acquisition module acquires a dynamic impedance spectrum data set of the target lithium battery during charge-discharge cycles. This set includes multi-frequency excitation signal sequences, voltage response waveform sequences, and temperature distribution monitoring data. The impedance field reconstruction module performs three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set to generate the internal impedance characteristics of the target lithium battery. These characteristics include ohmic impedance gradient, charge transfer impedance accumulation, and diffusion impedance fluctuation coefficient. The short-circuit diagnosis module calls a pre-generated and trained short-circuit identification model to perform nonlinear anomaly analysis on the internal impedance characteristics, thereby generating the short-circuit risk level and potential short-circuit region identification of the target lithium battery. The environmental compensation module uses temperature distribution monitoring data and the correlation between electrochemical impedance and temperature to perform battery condition compensation correction processing on the short-circuit risk level and outputs the corrected short-circuit risk level. The strategy generation module generates a set of safety response strategies based on the content of the potential short-circuit region identification. This set includes current path adjustment schemes and thermal management intervention schemes.

[0072] In one embodiment of the present invention, see [reference] Figure 2 When the lithium battery internal short circuit diagnosis system based on impedance spectroscopy analysis starts running, the data acquisition module acquires the dynamic impedance spectrum data set of the target lithium battery during a complete charge-discharge cycle. The dynamic impedance spectrum data set includes a multi-frequency excitation signal sequence swept from 0.01 Hz to 100 kHz, voltage response waveform sequences acquired from the positive and negative electrodes of the battery, and temperature distribution monitoring data monitored by five temperature sensors arranged on the battery surface. The impedance field reconstruction module performs three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set. In some embodiments, the impedance field reconstruction module divides the multi-frequency excitation signal sequence into ten excitation sub-bands with logarithmically uniform frequency intervals. The ten excitation sub-bands cover, for example, 0.01-0.1 Hz, 0.1-1 Hz, 1-10 Hz up to 10-100 kHz. Each excitation sub-band corresponds to an impedance sampling interval. Processing is performed on each excitation sub-band, and a three-dimensional potential distribution topology of the target lithium battery is constructed based on the voltage response waveform sequence and the battery's geometric model. The three-dimensional potential distribution topology includes the potential values ​​of the anode potential field at the three-dimensional spatial grid nodes, the potential values ​​of the cathode potential field at the three-dimensional spatial grid nodes, and the potential values ​​of the electrolyte potential field at the three-dimensional spatial grid nodes. These potential values ​​constitute spatial variation data.

[0073] Optionally, the impedance field reconstruction module couples the three-dimensional potential distribution topology with the currently processed excitation subband. Based on the correspondence between the anode potential field and the high-frequency excitation signal in the excitation subband, an anode potential-excitation mapping equation is established. The distribution function of the ohmic impedance component is obtained by solving the anode potential-excitation mapping equation. A charge transfer impedance calculation model is constructed based on the correlation characteristics between the cathode potential field and the low-frequency excitation signal. This model includes dynamic calibration parameters for the active area of ​​the electrode material and the interface capacitance. A mathematical expression of a charge transfer impedance calculation model reflecting these dynamic calibration parameters is as follows:

[0074]

[0075] in: This represents the charge transfer impedance at spatial coordinates (x, y, z). It is the gas constant. It is absolute temperature. It is the number of electrons in the reaction. It is Faraday's constant. It is the dynamic calibration parameter of the active area of ​​the electrode material at coordinates (x, y, z). These are the dynamic calibration parameters for the interface capacitance at coordinates (x, y, z). It is the exchange current density.

[0076] In practical implementation, the impedance field reconstruction module combines the spatial gradient of the electrolyte potential field and the ionic conductivity to establish an iterative calculation process for diffusion impedance. The iterative calculation process for diffusion impedance includes a feedback correction mechanism for potential increment and diffusion impedance increment. The distribution function of ohmic impedance component, the output results of the charge transfer impedance calculation model, and the output results of the iterative calculation process for diffusion impedance are fused into a spatial grid to generate an impedance distribution reconstruction result for the current sampling interval, which includes the spatial distribution matrix of ohmic impedance component, the spatial distribution matrix of charge transfer impedance component, and the spatial distribution matrix of diffusion impedance component.

[0077] In some embodiments, the reconstructed impedance distribution results of ten consecutive sampling intervals are fused in the frequency domain to calculate the final internal impedance characteristics. The calculation process includes calculating the rate of change of the spatial distribution matrix of the ten ohmic impedance components along the battery thickness direction to obtain the ohmic impedance gradient matrix, performing numerical integration of the spatial distribution matrix of the ten charge transfer impedance components in the electrode interface normal direction to obtain the charge transfer impedance cumulative scalar field, and calculating the ratio of the standard deviation to the mean of each spatial point in the spatial distribution matrix of the ten diffusion impedance components in the frequency range of 1 Hz to 10 Hz to obtain the diffusion impedance fluctuation coefficient matrix.

[0078] It can be understood that the ohmic impedance gradient is defined as the rate of change of the ohmic impedance component along the thickness direction of the battery. In a three-dimensional mesh, this is specifically represented by calculating the difference between the impedance values ​​of adjacent mesh points in the thickness direction. The charge transfer impedance accumulation is the integral of the charge transfer impedance component along the normal direction of the electrode interface, achieved by summing the discrete charge transfer impedance values ​​along the normal direction. The diffusion impedance fluctuation coefficient is the ratio of the standard deviation to the mean of the diffusion impedance component within a specified frequency range, which is set to 1 Hz to 10 Hz in specific implementations. Optionally, when constructing the three-dimensional potential distribution topology, the internal space of the battery is discretized into a mesh composed of cubic units. Each mesh node is assigned a potential value calculated from the voltage response waveform sequence. The anode potential field consists of the potential values ​​of the mesh nodes belonging to the anode region, the cathode potential field consists of the potential values ​​of the mesh nodes belonging to the cathode region, and the electrolyte potential field consists of the potential values ​​of the mesh nodes belonging to the electrolyte region. Spatial variation data records the potential values ​​at each mesh node.

[0079] In one embodiment of the present invention, the short-circuit diagnosis module receives internal impedance features generated by the impedance field reconstruction module. The internal impedance features include the Ohmic impedance gradient matrix, the charge transfer impedance cumulative scalar field, and the diffusion impedance fluctuation coefficient matrix. The short-circuit diagnosis module calls a pre-generated and trained short-circuit identification model to perform nonlinear anomaly analysis on the internal impedance features. The pre-trained short-circuit identification model is trained based on a large amount of historical battery failure data, and its structure includes an impedance anomaly detection layer, an interface state analysis layer, and a mass transfer analysis layer. In some embodiments, the short-circuit diagnostic module inputs the ohmic impedance gradient matrix into the impedance anomaly detection layer of the short-circuit identification model. The impedance anomaly detection layer has a built-in gradient mutation identification algorithm. The gradient mutation identification algorithm traverses all data points in the ohmic impedance gradient matrix, identifies spatial locations where the gradient value exceeds a preset mutation threshold, and records these spatial locations as spatial coordinates of the ohmic impedance anomaly region. At the same time, it records the change sequence of the gradient value of each abnormal coordinate point with the time step. This change sequence constitutes the impedance change trajectory. For a 50 Ah automotive ternary lithium-ion battery, in the diagnostic triggered after 300 charge-discharge cycles, the gradient mutation identification algorithm identifies three spatial coordinates in the region near the geometric center of the battery. The ohmic impedance gradient values ​​of the three spatial coordinates are on average 15 times higher than those of the surrounding area, and the impedance change trajectory of the three spatial coordinates shows a rapidly rising peak shape.

[0080] It can be understood that the scalar field of charge transfer impedance accumulation is input into the interface state analysis layer of the short-circuit identification model. The interface state analysis layer performs charge transfer barrier accumulation calculation, which performs spatial differentiation and statistical analysis on the scalar field of charge transfer impedance accumulation to generate the interface reactivity degradation probability and reaction rate decay prediction value corresponding to each grid cell. The interface reactivity degradation probability is a value between 0 and 1, and the reaction rate decay prediction value is expressed as a percentage of the expected decrease in reaction rate.

[0081] Optionally, the diffusion impedance fluctuation coefficient matrix is ​​input into the mass transfer analysis layer of the short-circuit identification model. The mass transfer analysis layer performs calculations based on the ion migration retardation model. The ion migration retardation model couples the diffusion impedance fluctuation coefficient matrix with the ion concentration field to calculate the electrolyte ion diffusion coefficient tensor field and concentration polarization evolution data. The ion diffusion coefficient tensor field describes the anisotropic migration ability of lithium ions in the electrolyte, while the concentration polarization evolution data records the predicted curve of lithium ion concentration change over time near the electrode surface.

[0082] In practical implementation, the short-circuit diagnostic module integrates impedance change trajectory, interface reaction activity degradation probability, and ion diffusion coefficient tensor field to generate a comprehensive anomaly index for the target lithium battery. The integration process involves weighted normalization and nonlinear mapping of multidimensional data. The calculation formula for the comprehensive anomaly index is as follows:

[0083]

[0084] in: Indicates comprehensive abnormal indicators, , , These are preset weighting coefficients for ohmic impedance gradient, interface degradation, and ion diffusion, respectively. It is the mean of the gradient norm of the anomalous region in the Ohm impedance gradient matrix. It is the spatial average of the probability of degradation of interfacial reactivity. It is the nominal lithium-ion diffusion coefficient at the reference temperature. It is the mean of the principal components of the lithium-ion diffusion coefficient tensor field obtained from the current calculation, based on the comprehensive anomaly index obtained from the calculation. The short-circuit risk level is determined by comparing the results with a preset set of safety thresholds, which defines the numerical ranges corresponding to four levels: "normal", "low risk", "medium risk" to "high risk".

[0085] In some embodiments, the short-circuit diagnosis module performs spatial overlay analysis based on the spatial coordinates of the ohmic impedance anomaly region, the spatial distribution map of the predicted reaction rate decay value output by the interface state analysis layer, and the spatial distribution map of the concentration polarization evolution data output by the mass transfer analysis layer. This spatial overlay analysis maps the three types of data onto a unified three-dimensional battery mesh model. By setting boundary judgment rules, the geometric boundaries of the impedance anomaly region, the interface degradation region, and the mass transfer impeded region are identified. These geometric boundaries are defined by a series of closed spatial polygonal surfaces. It can be understood that the outputs of the impedance anomaly detection layer, the interface state analysis layer, and the mass transfer analysis layer are processed in parallel. The core of the nonlinear anomaly analysis lies in the extraction and judgment of the deep nonlinear relationships of the input features by the multi-layer neural network within the short-circuit identification model. The determination of the short-circuit risk level depends on the comparison of comprehensive anomaly indicators and multi-level thresholds.

[0086] See Figure 3This is a thermal map showing the spatial distribution of the degradation probability of interface reactivity in a lithium battery. The colors in the map represent the degradation probability of interface reactivity at different locations within the battery, with the red area (center) having a probability close to 1.0 and the blue area (edge) having a probability close to 0. The degradation probability exhibits a symmetrical distribution, high at the center and low at the edges, indicating that the electrode interface reactivity in the geometric center of the battery decreases most severely, making it a high-risk area for potential short circuits. This map visually presents the spatial clustering of interface degradation and can be used to locate high-risk areas within the battery, providing a spatial basis for subsequent thermal management interventions and current path adjustments. The spatial distribution data output by this map can directly provide parameters for thermal management solutions, such as configuring higher-density cooling channels in high-degradation-probability areas or adjusting local current distribution to reduce interface load.

[0087] In one embodiment of the present invention, the environmental compensation module receives the short-circuit risk level generated by the short-circuit diagnosis module and the temperature distribution monitoring data continuously collected by the data acquisition module. The environmental compensation module performs battery condition compensation correction processing on the short-circuit risk level. In a specific example scenario, the target lithium battery is operating in a high-temperature environment. The temperature distribution monitoring data shows that the temperature at the tab connection reaches 45 degrees Celsius while the temperature at the bottom of the battery is 32 degrees Celsius. The temperature distribution monitoring data also records that the temperature in the tab area rises by 8 degrees Celsius within ten minutes.

[0088] In some embodiments, the environmental compensation module extracts extreme temperatures and temperature change gradients from the temperature distribution monitoring data. Extreme temperatures refer to the highest and lowest temperature values ​​within the monitoring period, and temperature change gradients are the rate of temperature change per unit time. Based on the electrochemical impedance spectroscopy database of the target lithium battery, the environmental compensation module calculates the dynamic compensation amount of the electrochemical impedance temperature coefficient as a function of temperature. The electrochemical impedance temperature coefficient describes the rate at which the impedance value decreases as the temperature increases, and the dynamic compensation amount is a function output value related to the difference between the real-time temperature and the reference temperature.

[0089] Optionally, the environmental compensation module performs temperature drift compensation calculations on the ohmic impedance gradient based on the dynamic compensation amount. It obtains the reference impedance temperature coefficient of the target lithium battery at a reference temperature of 25 degrees Celsius and the aforementioned calculated dynamic compensation amount, and establishes an impedance temperature coefficient-temperature correlation function. The impedance temperature coefficient-temperature correlation function expresses the specific value of the impedance temperature coefficient at different temperature points. The impedance temperature drift is calculated based on the impedance temperature coefficient-temperature correlation function. The impedance temperature drift is the product of the temperature change and the impedance temperature coefficient change. The calculated impedance temperature drift is superimposed on the original calculation process of the ohmic impedance gradient to generate an ohmic impedance gradient correction value that includes the temperature compensation effect. The ohmic impedance gradient correction value is further subjected to relaxation effect compensation processing. The relaxation effect compensation processing is based on the product factor of the electrochemical relaxation time constant and the temperature holding time. The electrochemical relaxation time constant describes the time scale required for the internal electrochemical process of the battery to reach equilibrium.

[0090] In practical implementation, based on the Arrhenius relationship between temperature change gradient and electrode reaction kinetics, the environmental compensation module performs reaction rate correction processing on the charge transfer impedance accumulation. The reaction rate correction processing recalculates the reaction rate constant based on the real-time temperature and updates the calculation formula of charge transfer impedance accumulation using the corrected reaction rate constant, thereby generating the corrected charge transfer impedance accumulation. According to the electrolyte conductivity change characteristics at extreme temperatures, the electrolyte conductivity exhibits an exponential growth relationship with increasing temperature. The environmental compensation module performs ion migration adaptation adjustment processing on the diffusion impedance fluctuation coefficient. The ion migration adaptation adjustment processing adjusts the ion mobility parameters in the diffusion model based on the conductivity-temperature relationship, generating the corrected diffusion impedance fluctuation coefficient.

[0091] Understandably, after the above corrections are completed, the environmental compensation module re-inputs the corrected ohmic impedance gradient, the corrected charge transfer impedance accumulation, and the corrected diffusion impedance fluctuation coefficient into the short-circuit identification model for calculation. The short-circuit identification model uses the same process as the initial diagnosis, but performs nonlinear anomaly analysis based on the corrected impedance characteristic value to generate a short-circuit risk level after compensating for the influence of operating conditions. In a data comparison example, before compensation, the impedance anomaly caused by local high temperature made the system determine the short-circuit risk level as "medium risk". After the environmental compensation module corrects the ohmic impedance gradient and charge transfer impedance accumulation, the recalculated comprehensive anomaly index decreases, and the final output short-circuit risk level is corrected to "low risk".

[0092] See Figure 4This is a bar chart comparing the short-circuit risk levels of different regions of a lithium battery before and after temperature compensation, with overlaid curves showing the changes in temperature impact. It has the following core values ​​in the lithium battery short-circuit diagnostic system: It clearly shows the changes in short-circuit risk levels of each region before compensation (red bars) and after compensation (blue bars). The risk in all regions is significantly reduced after temperature compensation, verifying the effectiveness of the environmental compensation module. The orange curve quantifies the temperature impact in different regions, with the tab region having the highest temperature impact (approximately 1.2) and the electrolyte region the lowest (approximately 0.2), indicating that temperature has the greatest impact on short-circuit risk in the tab region. The temperature impact data for each region can be used to optimize the temperature compensation algorithm of the short-circuit identification model, improving the accuracy of risk level determination under different operating conditions. This chart can directly guide the formulation of differentiated thermal management strategies, such as configuring a stronger cooling scheme in the tab region and appropriately reducing the thermal management intensity in the electrolyte region, achieving efficient resource allocation.

[0093] In one embodiment of the present invention, the strategy generation module receives potential short-circuit region identifiers output by the short-circuit diagnosis module. The potential short-circuit region identifiers include the geometric boundaries of the impedance anomaly region, the geometric boundaries of the interface degradation region, and the geometric boundaries of the mass transfer obstruction region. The strategy generation module generates a set of safety response strategies based on these identifiers. In an example scenario, the target lithium battery is identified as having an impedance anomaly region located in the northeast corner of the battery cell, a strip-shaped interface degradation region distributed along the positive electrode tab, and a circular mass transfer obstruction region in the center of the battery. In some embodiments, the strategy generation module calculates the optimal current path adjustment scheme for the identification of impedance abnormal regions. The optimal current path adjustment scheme is achieved by adjusting the current distribution ratio of the parallel branches inside the battery. The adjustment process is based on the equivalent circuit model of the battery and finite element simulation. In a battery module with multiple parallel cells, if an impedance abnormal region is identified inside a cell, the strategy generation module will calculate an instruction to reduce the current distribution ratio of the parallel branch where the cell is located. The instruction is specifically manifested in adjusting the switching duty cycle of the equalization circuit inside the battery management system. A data comparison example shows that before the adjustment, the abnormal branch carried 22% of the total current. After the strategy generation module calculates and implements the current path adjustment scheme, the current distribution ratio of the abnormal branch is controlled to be below 15% of the total current.

[0094] Optionally, the strategy generation module constructs a thermal management intervention scheme based on the identification of the interface degradation region. The thermal management intervention scheme includes the selection of local cooling regions and the optimized configuration of cooling intensity parameters. The local cooling region directly corresponds to the projection area of ​​the geometric boundary of the interface degradation region on the outer surface of the battery. The strategy generation module extracts the thermal characteristic parameters of the interface degradation region. The thermal characteristic parameters include the specific heat capacity of the material retrieved from the material database, the heat generation rate obtained by fitting historical thermal imaging data, and the thermal diffusivity obtained by inversion calculation through the thermal diffusivity equation.

[0095] In practice, the flow rate of the cooling medium is selected based on the heat generation rate, which is directly proportional to the heat generation rate. The proportionality coefficient is determined by the heat exchange efficiency of the cooling system. The arrangement density of the cooling channels is adjusted based on the thermal diffusivity. The arrangement density refers to the channel length per unit area. The adjustment goal is to achieve a dynamic balance between cooling efficiency and heat accumulation rate. The power of the semiconductor refrigeration chip is dynamically adjusted based on real-time temperature monitoring data. The installation position of the semiconductor refrigeration chip is aligned with the local cooling area. The power adjustment logic is to ensure that the interface temperature is maintained below the material phase transition point. The material phase transition point refers to the temperature threshold at which the battery electrode material undergoes a structural phase transition. Finally, a thermal management parameter configuration table containing cooling medium flow rate, channel arrangement density, and cooling power parameters is generated. See Table 1 for the thermal management parameter configuration table.

[0096] Table 1: Thermal Management Parameter Configuration Table

[0097] Parameter categories Calculated value unit Cooling medium flow rate 2.5 L / min Flow channel arrangement density 15 cm / cm² Cooler power setting 45 W

[0098] In some embodiments, the strategy generation module prioritizes the optimal current path adjustment scheme, thermal management intervention scheme, and electrolyte replenishment strategy. The prioritization is based on the short-circuit risk sub-level and risk evolution rate corresponding to each potential short-circuit region identifier. The short-circuit risk sub-level is generated synchronously by the short-circuit diagnosis module when generating comprehensive anomaly indicators. The risk evolution rate is obtained by comparing the change rate of the identified regions in the current diagnosis with that in historical diagnoses. After prioritization, a set of safety response strategies containing execution order and control parameters is generated. In one execution order example, the system determines to first execute the current path adjustment for high-risk impedance anomaly regions, then execute the thermal management intervention for medium-risk interface degradation regions, and finally execute the electrolyte replenishment strategy for low-risk mass transfer obstruction regions.

[0099] It is understandable that constructing a thermal management intervention scheme is a multi-parameter optimization process. The calculation of the cooling medium flow rate directly depends on the estimation of the real-time heat generation rate in the interface degradation region. The adjustment of the flow channel arrangement density is related to the spatial layout constraints and thermal diffusivity of the battery pack. The power control of the semiconductor cooling chip forms a closed-loop feedback with the goal of stabilizing the temperature of the target area below the set point. The thermal management parameter configuration table is a set of instructions that connects the strategy generation module and the underlying battery thermal management system execution mechanism.

[0100] See Figure 5This is a bar chart comparing the current allocation ratio before and after the adjustment of lithium battery safety response strategies, reflecting the optimization effect of resource allocation under different safety intervention measures. The chart clearly shows the optimization logic of "resources shifting from current path adjustment to thermal management intervention," reflecting the system's effectiveness evaluation results for different risk mitigation methods. This dynamic adjustment ensures that current resources are concentrated in the links that can more directly reduce short-circuit risks, improving the overall efficiency of safety response. It can provide a quantitative basis for the current scheduling algorithm of the battery management system (BMS), ensuring that in actual operating conditions, thermal management, current path adjustment, and electrolyte replenishment can allocate current resources in the optimal ratio. By comparing the changes in current ratio before and after the adjustment, the optimization effect of the safety response strategy can be verified, providing data support for subsequent algorithm iterations.

[0101] In one embodiment of the present invention, the model optimization module starts working. The model optimization module collects actual impedance change data and thermal runaway characteristic parameters of the target lithium battery within a preset verification period. The preset verification period is set to 100 charge-discharge cycles. The actual impedance change data is obtained by repeatedly measuring impedance spectra in the high-frequency range. The thermal runaway characteristic parameters include voltage drop initiation temperature, peak gas generation rate, and surface temperature rise rate. In some embodiments, the model optimization module performs deviation analysis processing on the actual impedance change data and the predicted impedance distribution previously generated by the short-circuit diagnosis module. The predicted impedance distribution is the result of the short-circuit identification model's estimation of the battery's internal impedance spatial state during the diagnosis process. The deviation analysis processing calculates the relative error between the actual measured value and the predicted value at each spatial coordinate point, and performs statistical averaging and variance calculation on the errors at all coordinate points to generate a first calibration coefficient that comprehensively reflects the model's prediction accuracy. A data comparison example shows that the predicted ohmic impedance at coordinates (10, 20, 5) mm is 5.2 milliohms, while the average measured value of the actual impedance change data at that coordinate is 5.8 milliohms. The deviation analysis processing calculates the first calibration coefficient based on such data point pairs.

[0102] It is understandable that the model optimization module performs a time-domain comparison of the thermal runaway characteristic parameters with the predicted risk level previously output by the short-circuit diagnosis module. The time-domain comparison process first aligns the time series of the predicted risk level with the time points when the thermal runaway characteristic parameters appear. Then, it analyzes whether significant changes in the thermal runaway characteristic parameters actually occurred within the time window predicted as high risk. By calculating the temporal correlation and strength between the predicted risk level and the evolution of the thermal runaway characteristic parameters, a second calibration coefficient is generated.

[0103] In practical implementation, the model optimization module adjusts the decision threshold of the short-circuit identification model based on the first and second calibration coefficients. The decision threshold is the numerical boundary used in the short-circuit identification model to distinguish different short-circuit risk levels. The adjustment process uses a weighted update algorithm. The weighted update algorithm scales or shifts each threshold in the decision threshold set according to the magnitude and sign of the first and second calibration coefficients to generate an optimized short-circuit identification model. The decision threshold adjustment formula is expressed as:

[0104]

[0105] in: The first one in the optimized short-circuit identification model A new value for the decision threshold. The first short-circuit identification model before optimization The original values ​​of each decision threshold. It is the learning rate constant. It is the first calibration coefficient Weighting factors It is the second calibration coefficient Weighting factors.

[0106] In some embodiments, the optimized short-circuit identification model is applied to the internal short-circuit diagnosis task of subsequent batches of lithium batteries. The subsequent batches of lithium batteries are of the same model as the target lithium battery but were produced at different times. After the model optimization module completes the optimization, the system automatically replaces the model file loaded in the short-circuit diagnosis module with the optimized short-circuit identification model. When the subsequent batches of lithium batteries are diagnosed, their internal impedance characteristics will be input into the optimized short-circuit identification model for analysis, generating a calibrated short-circuit risk level and potential short-circuit area identification.

[0107] Optionally, the preset verification cycle is set based on the typical accelerated life test cycle of the battery. The acquisition frequency of actual impedance change data is higher than that during routine diagnosis. Thermal runaway characteristic parameters are triggered and recorded in a dedicated battery safety test cabinet by applying overcharge or heating abuse conditions. The calculation of the first and second calibration coefficients is performed periodically. After each preset verification cycle of data acquisition is completed, a model optimization process is executed. It can be understood that the deviation analysis process not only generates an overall first calibration coefficient, but also generates a spatial error distribution map. The spatial error distribution map is used to identify which spatial regions of the battery have larger prediction deviations in the short-circuit identification model. The time-domain comparison process focuses on examining the time difference between the jump from the predicted risk level to the appearance of actual thermal runaway characteristic parameters. This time difference is the key to evaluating the model's early warning capability. The optimized short-circuit identification model has a decision threshold that is closer to the safety boundary of the actual battery.

[0108] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis, characterized in that, The system includes: The data acquisition module is used to acquire the dynamic impedance spectrum data set of the target lithium battery during the charge and discharge cycle process. The dynamic impedance spectrum data set includes multi-frequency excitation signal sequence, voltage response waveform sequence and temperature distribution monitoring data. The impedance field reconstruction module is used to perform three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set to generate the internal impedance characteristics of the target lithium battery. The internal impedance characteristics include the ohmic impedance gradient, the charge transfer impedance accumulation, and the diffusion impedance fluctuation coefficient. The short-circuit diagnosis module is used to call a pre-trained short-circuit identification model to perform nonlinear anomaly analysis on the internal impedance characteristics, and generate the short-circuit risk level and potential short-circuit area identification of the target lithium battery. include: The ohmic impedance gradient is input into the impedance anomaly detection layer of the short-circuit identification model, and the spatial coordinates and impedance change trajectory of the ohmic impedance anomaly region are determined by the gradient mutation identification algorithm. The accumulated charge transfer impedance is input into the interface state analysis layer of the short-circuit identification model to perform charge transfer barrier accumulation calculation and generate the interface reactivity degradation probability and reaction rate decay prediction value. The diffusion impedance fluctuation coefficient is input into the mass transfer analysis layer of the short-circuit identification model, and the electrolyte ion diffusion coefficient and concentration polarization evolution data are calculated based on the ion migration retardation model. By integrating the impedance change trajectory, the interface reaction activity degradation probability, and the ion diffusion coefficient, a comprehensive anomaly index for the target lithium battery is generated, and the short-circuit risk level is determined based on the comparison result between the comprehensive anomaly index and the preset safety threshold. Based on the spatial superposition of the spatial coordinates, the predicted reaction rate decay value, and the concentration polarization evolution data, the geometric boundaries of the impedance anomaly region, the interface degradation region, and the mass transfer impeded region are identified. An environmental compensation module is used to perform battery condition compensation and correction processing on the short-circuit risk level to generate a corrected short-circuit risk level. The battery condition compensation and correction processing is based on the correlation between the temperature distribution monitoring data and the electrochemical impedance temperature. The strategy generation module is used to generate a set of safety response strategies based on the potential short-circuit area identifier. The set of safety response strategies includes current path adjustment schemes and thermal management intervention schemes.

2. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 1, characterized in that, The impedance field reconstruction module performs three-dimensional impedance distribution reconstruction processing on the dynamic impedance spectrum data set to generate the internal impedance characteristics of the target lithium battery, including: The multi-frequency excitation signal sequence is divided into multiple excitation sub-bands according to the frequency band range, and each excitation sub-band corresponds to an impedance sampling interval; For each of the aforementioned excitation sub-bands, the following processing is performed: The three-dimensional potential distribution topology of the target lithium battery is constructed based on the voltage response waveform sequence. The three-dimensional potential distribution topology includes spatial variation data of the anode potential field, cathode potential field, and electrolyte potential field. The three-dimensional potential distribution topology is coupled with the excitation sub-band for analysis to generate the impedance distribution reconstruction result of the current sampling interval. The impedance distribution reconstruction result includes the spatial distribution matrix of the ohmic impedance component, the charge transfer impedance component and the diffusion impedance component. The impedance distribution reconstruction results from multiple consecutive sampling intervals are subjected to frequency domain fusion processing to calculate the ohmic impedance gradient, charge transfer impedance accumulation, and diffusion impedance fluctuation coefficient; wherein, The ohmic impedance gradient is the rate of change of the ohmic impedance component along the battery thickness direction. The cumulative charge transfer impedance is the integral of the charge transfer impedance component along the normal direction of the electrode interface. The diffusion impedance fluctuation coefficient is the ratio of the standard deviation to the mean of the diffusion impedance component within a specified frequency range.

3. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 2, characterized in that, The impedance field reconstruction module couples the three-dimensional potential distribution topology with the excitation sub-band to generate the impedance distribution reconstruction result for the current sampling interval, including: Based on the correspondence between the anode potential field and the high-frequency excitation signal in the excitation sub-band, an anode potential-excitation mapping equation is established, and the distribution function of the ohmic impedance component is obtained by solving the anode potential-excitation mapping equation. Based on the correlation characteristics between the cathode potential field and the low-frequency excitation signal, a charge transfer impedance calculation model is constructed. The charge transfer impedance calculation model includes dynamic calibration parameters of the active area of ​​the electrode material and the interface capacitance. By combining the spatial gradient of the electrolyte potential field and the ionic conductivity, an iterative calculation process for diffusion impedance is established, which includes a feedback correction mechanism for potential increment and diffusion impedance increment. The output results of the distribution function, the charge transfer impedance calculation model, and the diffusion impedance iterative calculation process are subjected to spatial grid fusion processing to generate three-dimensional impedance distribution data containing ohmic impedance, charge transfer impedance, and diffusion impedance.

4. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 1, characterized in that, The environmental compensation module performs battery condition compensation correction processing on the short-circuit risk level to generate a corrected short-circuit risk level, including: Extract the extreme temperatures and temperature change gradients from the temperature distribution monitoring data, and calculate the dynamic compensation amount of the electrochemical impedance temperature coefficient as a function of temperature. The temperature drift compensation calculation is performed on the ohmic impedance gradient based on the dynamic compensation amount to generate the corrected ohmic impedance gradient. Based on the correlation between the temperature change gradient and electrode reaction kinetics, the cumulative charge transfer impedance is modified by the reaction rate to generate a modified cumulative charge transfer impedance. Based on the characteristics of electrolyte conductivity change at extreme temperatures, the diffusion impedance fluctuation coefficient is adjusted by ion migration adaptation to generate a corrected diffusion impedance fluctuation coefficient. The corrected ohmic impedance gradient, charge transfer impedance accumulation, and diffusion impedance fluctuation coefficient are input into the short-circuit identification model for recalculation, generating the short-circuit risk level after compensating for operating conditions.

5. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 4, characterized in that, The environmental compensation module performs temperature drift compensation calculations on the ohmic impedance gradient based on the dynamic compensation amount, generating a corrected ohmic impedance gradient, including: Obtain the reference impedance temperature coefficient of the target lithium battery at the reference temperature and the dynamic compensation amount, and establish an impedance temperature coefficient-temperature correlation function; The impedance temperature drift is calculated based on the impedance temperature coefficient-temperature correlation function, whereby the impedance temperature drift is the product of the temperature change and the impedance temperature coefficient change. The impedance temperature drift is superimposed on the calculation of the ohmic impedance gradient to generate an ohmic impedance gradient correction value that includes temperature compensation effect. The ohmic impedance gradient correction value is subjected to relaxation effect compensation processing, which is based on the product factor of the electrochemical relaxation time constant and the temperature holding time.

6. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 1, characterized in that, The strategy generation module generates a set of security response strategies based on the potential short-circuit region identifiers, including: For the identification of the impedance abnormality region, the optimal current path adjustment scheme is calculated. The optimal current path adjustment scheme is achieved by adjusting the current distribution ratio of the parallel branches inside the battery. Based on the identification of the interface degradation area, a thermal management intervention scheme is constructed, which includes the selection of local cooling areas and the optimized configuration of cooling intensity parameters; Based on the identification of the mass transfer obstruction region, an electrolyte replenishment strategy is generated. The electrolyte replenishment strategy dynamically adjusts the replenishment dose and replenishment location according to the predicted ion migration rate. The optimal current path adjustment scheme, the thermal management intervention scheme, and the electrolyte replenishment strategy are prioritized to generate a set of safety response strategies that include execution order and control parameters.

7. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 6, characterized in that, The strategy generation module constructs a thermal management intervention plan, including: Extract the thermal characteristic parameters of the interface degradation region, and calculate the heat generation rate and thermal diffusivity. The flow rate of the cooling medium is selected based on the heat generation rate, and the flow rate is directly proportional to the heat generation rate. The arrangement density of the cooling channels is adjusted based on the thermal diffusivity to achieve a dynamic balance between cooling efficiency and heat accumulation rate. The power of the semiconductor cooling chip is dynamically adjusted based on real-time temperature monitoring data to ensure that the interface temperature is maintained below the material phase transition point. Generate a configuration table of thermal management parameters, including cooling medium flow rate, channel layout density, and cooling power.

8. The lithium battery internal short-circuit diagnostic system based on impedance spectroscopy analysis according to claim 1, characterized in that, The system also includes: The model optimization module is used to collect actual impedance change data and thermal runaway characteristic parameters of the target lithium battery within a preset verification period. The actual impedance change data and the predicted impedance distribution are subjected to deviation analysis to generate the first calibration coefficient. The thermal runaway characteristic parameters are compared with the predicted risk level in the time domain to generate a second calibration coefficient. The decision threshold of the short-circuit identification model is adjusted according to the first calibration coefficient and the second calibration coefficient to generate an optimized short-circuit identification model. The optimized short-circuit identification model was then applied to subsequent batches of lithium battery internal short-circuit diagnosis tasks.

9. A method for internal short circuit of lithium batteries based on impedance spectroscopy analysis, characterized in that, It includes all modules and method flows of the lithium battery internal short-circuit system based on impedance spectroscopy analysis as described in any one of claims 1 to 8.

Images