Whole life cycle safety management method and system for aqueous battery

Abstract

The invention relates to the field of battery safety management, and particularly discloses an aqueous battery full-life-cycle safety management method and system, which abandons a traditional mode depending on a single macroscopic parameter or global fuzzy analysis, and deeply analyzes an electrochemical impedance spectrum and temperature dynamic state to realize the safety management of the whole life cycle of an aqueous battery. And accurately extracting a plurality of core precursor characteristics respectively representing electrode interface deterioration, dangerous side reaction germination and system thermodynamic instability trend. The multi-dimensional microscopic features are fused into a comprehensive risk score, so that accurate quantitative evaluation of the safety state is realized. Furthermore, an active decision is made based on the risk score, after the risk is identified, an electric field pulse can be actively applied to trigger a self-repairing mechanism in the electrolyte, and a protection layer is formed in situ at a risk source, so that post-event alarm of safety management is converted into beforehand intervention, and the safety of safety management is improved. And the running reliability of the whole life cycle of the water-based battery is fundamentally improved.

Description

Technical Field

[0001] This application relates to the field of battery safety management, and more specifically, to a method and system for full life-cycle safety management of aqueous batteries. Background Technology

[0002] Aqueous batteries, with their inherently safe electrolyte, abundant resources, low cost, and environmental friendliness, are considered a highly promising technology for large-scale energy storage systems (such as grid peak shaving and renewable energy grid integration). However, although aqueous batteries are far safer than traditional organic electrolyte lithium-ion batteries, they can still experience violent chemical side reactions under abuse conditions such as overcharging, over-discharging, rapid charging and discharging, or internal short circuits. These reactions can include hydrogen evolution, electrode material dissolution, and dendrite growth. These reactions release a large amount of heat, and when the heat generation rate exceeds the heat dissipation rate, thermal runaway can be triggered, leading to safety accidents such as battery swelling, electrolyte leakage, or even combustion and explosion.

[0003] Currently, for battery safety management, existing technologies mainly rely on the battery management system (BMS) to monitor macroscopic parameters such as voltage, current, and battery surface temperature. When these parameters exceed preset static thresholds, the system executes protective actions such as alarms or circuit disconnection. However, this approach is essentially a passive response mechanism, and the macroscopic parameters it monitors are often lagging indicators of the thermal runaway process. In other words, by the time the BMS detects a significant temperature surge or voltage drop, the chain reaction of exothermic reactions inside the battery has usually already begun, making intervention at this point often too late to effectively prevent thermal runaway. To achieve earlier warnings, some solutions introduce electrochemical impedance spectroscopy (EIS) as a probe of the internal state. However, these solutions typically employ a global, indiscriminate anomaly detection logic when analyzing impedance spectroscopy data. For example, they accumulate the fitting errors of all frequency points into a total residual and compare it with a fixed global threshold. The fundamental flaw of this method is that it ignores the strong correspondence between different frequency bands in the impedance spectrum and different physicochemical processes inside the battery. For example, early, weak signals associated with high-risk events such as dendrite growth are mainly present in the low-frequency region. Their contribution to the global residual is easily masked by measurement noise or relatively benign high-frequency errors, resulting in insufficient sensitivity to key early fault characteristics and thus missing the optimal opportunity for preventative intervention. In summary, existing technologies generally suffer from technical problems such as insensitivity to precursor signals of thermal runaway, a single dimension of risk assessment, and a lack of effective differentiation of fault nature.

[0004] Therefore, an optimized safety management solution for the entire life cycle of aqueous batteries is desired. Summary of the Invention

[0005] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a method and system for the full lifecycle safety management of aqueous batteries.

[0006] According to one aspect of this application, a method for full life-cycle safety management of aqueous batteries is provided, comprising: Obtain electrochemical impedance spectroscopy and battery surface temperature sequences; Thermal runaway precursor features were extracted from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, anomalous impedance characteristics, and normalized temperature rise rate. The thermal runaway risk score is determined based on the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. Control decisions are made based on thermal runaway risk scores to obtain control signals; In response to the non-empty control signal, an electric field pulse is applied to both ends of the battery, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

[0007] According to another aspect of this application, a full life-cycle safety management system for aqueous batteries is provided, comprising: The battery data acquisition module is used to acquire electrochemical impedance spectroscopy and battery surface temperature sequences; The thermal runaway precursor feature extraction module is used to extract thermal runaway precursor features from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, abnormal impedance characteristics and normalized temperature rise rate. The thermal runaway risk scoring module is used to determine the thermal runaway risk score based on the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. The control decision module is used to make control decisions based on thermal runaway risk scores to obtain control signals. The regulation response module is used to apply an electric field pulse to both ends of the battery in response to the regulation signal being non-empty, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

[0008] Compared with existing technologies, this application provides a method and system for full life-cycle safety management of aqueous batteries. It constructs a closed-loop proactive safety management system encompassing perception, analysis, decision-making, and execution to address the problems of delayed response and insensitivity to early risk precursors in existing technologies. This solution abandons the traditional model relying on single macroscopic parameters or global fuzzy analysis, instead using deep analysis of electrochemical impedance spectroscopy and temperature dynamics to accurately extract multiple core precursor features characterizing electrode interface degradation, the budding of dangerous side reactions, and the trend of system thermodynamic instability. By integrating these multi-dimensional microscopic features into a comprehensive risk score, accurate quantitative assessment of the safety status is achieved. Furthermore, this solution overcomes the limitations of passive protection by making proactive decisions based on this risk score. Upon identifying a risk, it can proactively apply an electric field pulse to trigger the self-repair mechanism within the electrolyte, forming a protective layer in situ at the source of the risk. This transforms safety management from post-event warnings to pre-event intervention, fundamentally improving the operational reliability of aqueous batteries throughout their entire life-cycle. Attached Figure Description

[0009] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flowchart of a method for full lifecycle safety management of aqueous batteries according to an embodiment of this application; Figure 2 This is a data flow diagram illustrating the water-based battery lifecycle safety management method according to an embodiment of this application. Figure 3 This is a flowchart illustrating the process of extracting precursor features of thermal runaway from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate in the water-based battery life cycle safety management method according to embodiments of this application. Figure 4 This is a block diagram of a water-based battery lifecycle safety management system according to an embodiment of this application. Detailed Implementation

[0011] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0012] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0013] While this application makes various references to certain modules of the systems according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0014] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0015] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0016] Existing technologies for the safety management of aqueous batteries have fundamental flaws. They primarily rely on monitoring macroscopic parameters such as battery surface temperature, resulting in a severely delayed response. Even when electrochemical impedance spectroscopy (EIS) is employed, its global and indiscriminate analysis method is insensitive to weak early warning signs of thermal runaway, such as dendrite growth, thus missing the optimal intervention window. To address this issue, this application proposes a method for the full lifecycle safety management of aqueous batteries. This method first acquires EIS and surface temperature sequences in real time, reflecting the battery's internal microstate. Then, through in-depth analysis, it precisely extracts normalized charge transfer resistance (representing deterioration of interfacial dynamics), abnormal impedance characteristics indicating dangerous side reactions, and normalized temperature rise rate (reflecting system thermal instability). Next, these multi-dimensional warning features are weighted and fused into a quantified thermal runaway risk score. Based on this score, the system makes intelligent decisions. Once the risk score exceeds a preset control threshold, it no longer passively issues an alarm but actively generates a control signal, applying specific electric field pulses to both ends of the battery. This pulse can precisely stimulate pre-set dormant functional molecules in the electrolyte, causing them to form an activated protective layer in situ in the highest-risk areas on the negative electrode surface (such as dendrite tips), thereby physically suppressing and mitigating risks. In this way, safety management is fundamentally transformed from a delayed, passive response to a proactive, forward-looking control, effectively solving the technical problem of existing technologies being insensitive to early risk signals.

[0017] The technical solution of this application proposes a method for full life cycle safety management of aqueous batteries. Figure 1 This is a flowchart of a method for the full life-cycle safety management of aqueous batteries according to an embodiment of this application. Figure 2 This is a data flow diagram illustrating the water-based battery lifecycle safety management method according to an embodiment of this application. Figure 1 and Figure 2 As shown, the method for full life-cycle safety management of aqueous batteries according to an embodiment of this application includes the following steps: S100, acquiring an electrochemical impedance spectroscopy and a battery surface temperature sequence; S200, extracting thermal runaway precursor features from the electrochemical impedance spectroscopy and the battery surface temperature sequence to obtain normalized charge transfer resistance, abnormal impedance features, and normalized temperature rise rate; S300, determining a thermal runaway risk score based on the normalized charge transfer resistance, abnormal impedance features, and normalized temperature rise rate; S400, making a control decision based on the thermal runaway risk score to obtain a control signal; S500, in response to the control signal being non-empty, applying an electric field pulse to both ends of the battery, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

[0018] Specifically, in step S100, electrochemical impedance spectroscopy and battery surface temperature sequences are acquired. It should be understood that relying solely on macroscopic electrical parameters such as battery voltage and total current, or a single surface temperature point, is insufficient to effectively observe the microscopic changes in the electrochemical interface within the battery. These microscopic changes, such as an increase in charge transfer resistance or the appearance of nascent impedance characteristics, are precisely the early core precursors of thermal runaway. Therefore, in the technical solution of this application, by acquiring electrochemical impedance spectroscopy and battery surface temperature sequences, real-time data streams capable of characterizing the internal electrochemical kinetics, interface structural integrity, and system thermodynamic state of the battery are simultaneously captured. This provides a raw dataset containing multi-dimensional, highly sensitive information for subsequent extraction of thermal runaway precursor features and risk assessment, thus laying a data foundation for early and accurate warning of thermal runaway risks.

[0019] More specifically, in a specific example of this application, the process of obtaining the electrochemical impedance spectroscopy and battery surface temperature sequence includes the following steps. First, the battery management system continuously collects the battery surface temperature values ​​at a preset sampling frequency, such as 10 times per second, using multiple thermistors or thermocouples deployed on the battery casing surface. These discrete temperature data points are then organized chronologically to form a battery surface temperature sequence. Subsequently, at a preset diagnostic moment, such as the initial stage of each charging cycle or when abnormal fluctuations in current are detected, the battery management system controls its internal signal generation and acquisition module to apply a small-amplitude sinusoidal AC excitation signal to both ends of the battery. The frequency of this excitation signal is scanned within a specified range, for example, from 100 kHz to 0.01 Hz. Throughout the application of the excitation signal, the high-precision acquisition module synchronously records the battery's voltage and current responses. Finally, the system performs Fourier transforms on the excitation signal and response signal at each frequency point, calculating the amplitude ratio and phase difference between them to obtain a series of complex impedance values. These complex impedance values ​​together constitute the electrochemical impedance spectroscopy characterizing the current battery state.

[0020] Specifically, in step S200, pre-thermal runaway features are extracted from the electrochemical impedance spectroscopy and battery surface temperature sequence to obtain normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. It should be understood that since the original electrochemical impedance spectroscopy and battery surface temperature sequence are complex high-dimensional data, they cannot directly quantify the risk level of thermal runaway. The pre-thermal information contained within them needs to be made explicit through specific signal processing and model analysis. Therefore, in the technical solution of this application, pre-thermal runaway features are further extracted from the electrochemical impedance spectroscopy and battery surface temperature sequence to obtain normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. This transforms the original data stream into low-dimensional key feature indicators that can respectively characterize the dynamic state of the electrode interface, the risk of internal micro-short circuits, and the system's heat generation trend. In this way, a complex, multi-dimensional state monitoring problem can be reduced in dimensionality and focused on a few physical quantities strongly related to the core mechanism of thermal runaway, providing quantitative and standardized inputs for the subsequent construction of an accurate and robust comprehensive risk assessment model.

[0021] Figure 3 This document presents a flowchart illustrating the process of extracting precursor features of thermal runaway from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate, according to the full life-cycle safety management method for aqueous batteries as described in this application. Figure 3 As shown, step S200 includes: S210, performing nonlinear least squares fitting on the electrochemical impedance spectrum and the baseline impedance spectrum to obtain the current charge transfer resistance and the baseline charge transfer resistance; S220, determining the normalized charge transfer resistance based on the current charge transfer resistance and the baseline charge transfer resistance; S230, inputting the electrochemical impedance spectrum into the ECM model to obtain the equivalent circuit fitted impedance spectrum; S240, calculating the frequency-resolved residual vector between the electrochemical impedance spectrum and the equivalent circuit fitted impedance spectrum; S250, performing weighted anomaly scoring on the frequency-resolved residual vector to obtain the frequency-weighted residual; S260, performing adaptive threshold judgment on the frequency-weighted residual to obtain the anomalous impedance characteristics.

[0022] Specifically, the process for calculating the normalized charge transfer resistance is as follows: Accordingly, in step S210, the electrochemical impedance spectroscopy and baseline impedance spectroscopy are fitted using a nonlinear least squares method to obtain the current charge transfer resistance and baseline charge transfer resistance. It should be understood that since the original electrochemical impedance spectroscopy exists as a set of complex data points, it does not directly provide parameters with clear physical meaning, such as charge transfer resistance, and cannot be directly used to quantify the kinetic changes at the electrode interface. Therefore, in the technical solution of this application, the electrochemical impedance spectroscopy and baseline impedance spectroscopy are further fitted using a nonlinear least squares method to obtain the current charge transfer resistance and baseline charge transfer resistance, thereby resolving the complex, graphical impedance spectroscopy data into discrete, physically meaningful equivalent circuit element parameters. This allows for the precise decoupling and quantification of changes in electrode interface reactivity from the complex impedance spectrum, providing two key, directly comparable numerical foundations for subsequent calculations of relative change rates and risk assessment.

[0023] In a specific example of this application, the implementation process of nonlinear least squares fitting of electrochemical impedance spectroscopy and baseline impedance spectroscopy includes the following steps: First, a pre-defined equivalent circuit model capable of characterizing the electrochemical behavior of an aqueous battery is invoked, such as the Randles model which includes ohmic resistance, charge transfer resistance, and constant phase angle elements. Then, using the parameters of this equivalent circuit model as undetermined variables, and the sum of squared residuals between the measured electrochemical impedance spectroscopy data and the theoretical impedance spectroscopy data calculated by the model as the objective function, a nonlinear least squares optimization algorithm is initiated. This algorithm iteratively adjusts the parameter values ​​of each element in the model until the objective function converges to its minimum value. After the algorithm converges, the resistance value corresponding to the charge transfer process is extracted from the final parameter set and determined as the current charge transfer resistance. Finally, using the exact same equivalent circuit model and optimization algorithm, the same fitting operation is performed on the baseline impedance spectrum pre-stored in memory to obtain the baseline charge transfer resistance.

[0024] Accordingly, in step S220, a normalized charge transfer resistance is determined based on the current charge transfer resistance and the baseline charge transfer resistance. It should be understood that directly using the absolute values ​​of the current charge transfer resistance and the baseline charge transfer resistance, or their simple difference, cannot form a risk characterization with a uniform scale and clear physical meaning. The absolute magnitude and range of variation of these values ​​will differ significantly depending on the battery's health status, temperature, and other operating conditions, making them difficult to directly use for subsequent cross-dimensional feature fusion. Therefore, in the technical solution of this application, a normalized charge transfer resistance is further determined based on the current charge transfer resistance and the baseline charge transfer resistance. This transforms a resistance change with physical units and an indefinite range into a dimensionless, standardized risk characteristic value constrained to the interval between 0 and 1. This ensures that the charge transfer resistance feature has a standardized contribution comparable to other features when subsequently weighted summation is performed to calculate the comprehensive risk score, thereby guaranteeing the robustness and accuracy of the final risk assessment model.

[0025] Specifically, in this embodiment of the application, determining the normalized charge transfer resistance based on the current charge transfer resistance and the baseline charge transfer resistance includes: calculating the normalized charge transfer resistance using the following formula: ; in, For the current charge transfer resistance, Baseline charge transfer resistance, The relative rate of change of charge transfer resistance. and These are hyperparameters calibrated based on experimental data, controlling the slope and center point respectively. The risk center point of the rate of change of resistance is defined, and This controls the slope or sensitivity of the response of the normalized value near the center point. This is the normalized charge transfer resistance.

[0026] Specifically, the process for calculating the abnormal impedance characteristics is as follows: It is understandable that the existing anomaly detection mechanisms based on equivalent circuit models suffer from a technical bottleneck rooted in their global, indiscriminate anomaly judgment logic. This logic, when calculating the fitting residuals, indiscriminately accumulates the fitting errors at all frequency points of the electrochemical impedance spectroscopy into a total residual square sum, and compares it with a fixed global threshold. This approach essentially ignores the crucial and unique information that there is a strong correspondence between different frequency bands in the electrochemical impedance spectroscopy and different physicochemical processes within the battery. Specifically, residual fluctuations in the high-frequency region are usually associated with benign aging of the ohmic internal resistance or electrolyte interface film, while residual anomalies in the low-frequency region often indicate the formation of dangerous new interfaces such as impaired diffusion or dendrite growth. The global residual sum calculation method cannot effectively distinguish between these two distinct types of deviations. This leads to a residual value caused by a relatively benign, large deviation in the high-frequency region potentially being confused with a residual value generated by a low-frequency deviation representing a fatal risk but with a weak initial signal, thus causing misjudgment or missed detection. More critically, early signals of thermal runaway precursors, such as dendrite initiation, appear as extremely small anomalies in the low-frequency band of the impedance spectrum. Their contribution to the global residual sum is easily drowned out by measurement noise and fitting errors across the entire frequency band, preventing them from effectively triggering the fixed global threshold. This makes the mechanism insufficiently sensitive to key early fault characteristics, easily overlooking them and missing the optimal opportunity for preventative intervention. At its root, this deficiency lies in its reduction of a vector problem rich in structured information (i.e., the distribution of errors in the frequency domain) to a scalar problem with limited information (i.e., the sum of errors), sacrificing diagnostic accuracy and predictability. To address these technical shortcomings, an anomaly feature identification method based on frequency weighting and dynamic thresholds is proposed.

[0027] Accordingly, in steps S230 and S240, the electrochemical impedance spectroscopy (EIS) is input into the ECM model to obtain the equivalent circuit fitting impedance spectrum, and the frequency-resolved residual vector between the EIS and the equivalent circuit fitting impedance spectrum is calculated. It should be understood that compressing all fitting error information into a single scalar results in the loss of the complete distribution information of the error in the frequency domain. This makes it impossible to distinguish between a residual caused by a relatively benign large deviation in the high-frequency region and a low-frequency deviation representing fatal risks such as dendrite growth but with a weak initial signal, thus leading to misjudgment or missed judgment. Therefore, in the technical solution of this application, the EIS is further input into the ECM model to obtain the equivalent circuit fitting impedance spectrum, and the frequency-resolved residual vector between the EIS and the equivalent circuit fitting impedance spectrum is calculated. This generates an error map that clearly indicates in which specific frequency band the model failed to predict, providing a data basis for accurately locating and quantifying the model fitting deviation in the frequency domain. This provides a vectorized input that retains complete structured information for subsequent frequency-weighted analysis, making it possible to amplify specific frequency band anomalous signals related to precursors of thermal runaway.

[0028] In a specific example of this application, the implementation process includes the following steps: First, using the equivalent circuit model parameters obtained by fitting with the nonlinear least squares method in the previous steps, a theoretical complex impedance value is calculated for each angular frequency point used in the electrochemical impedance spectroscopy measurement. All these theoretical complex impedance values ​​together constitute the equivalent circuit fitted impedance spectrum. Then, iterating through all angular frequency points, at each angular frequency point, the measured complex impedance value and the theoretical complex impedance value calculated by the model are extracted. The system subtracts these two complex numbers to obtain a complex number difference, and then calculates the square of the Euclidean distance of this difference in the complex plane, i.e., the square of the modulus of the difference, which is used as the residual value at that frequency point. This process is described by the following formula: ; in, Represents an angular frequency of The residual value calculated at the point; This is the complex impedance value actually measured at this frequency point; This is the theoretical complex impedance value at that frequency point obtained by fitting the equivalent circuit model. Finally, the residual values ​​calculated for all frequency points are arranged in frequency order to form a frequency-resolved residual vector that corresponds one-to-one with each frequency point.

[0029] Accordingly, in step S250, a weighted anomaly score is performed on the frequency-resolved residual vector to obtain a frequency-weighted residual score. It should be understood that since the residual values ​​in the frequency-resolved residual vector are equally weighted, the absolute values ​​of the weak low-frequency residual signals, which are directly related to dangerous fault modes such as dendrite growth, are easily drowned out by measurement noise or benign fluctuations in other frequency bands, leading to insufficient sensitivity of the system to key risk signals and thus causing missed detections. Therefore, in the technical solution of this application, a weighted anomaly score is further performed on the frequency-resolved residual vector to obtain a frequency-weighted residual score, thereby introducing a mechanism similar to a mathematical magnifying glass specifically used for observing the low-frequency band of electrochemical impedance spectroscopy. This specifically amplifies those weak low-frequency residual signals that are directly related to dangerous fault modes, overcoming the deficiency of insufficient sensitivity in the original mechanism. In this way, even if the absolute value of the low-frequency residual signal representing early dendrite growth is very small, after being amplified by a huge weight, its contribution to the final score will become extremely significant, thereby achieving early and high-sensitivity capture of key risk signals and effectively solving the problem of missed detection caused by weak fault signals.

[0030] In a specific example of this application, the implementation process of weighted anomaly scoring of the frequency-resolved residual vector includes the following operations. First, iterate through each element in the frequency-resolved residual vector generated in the previous steps, i.e., each angular frequency point. residual values ​​on For each angular frequency point The system calculates the corresponding weight function value. The weighting function is defined as the reciprocal of the angular frequency, i.e. = 1 / Subsequently, the residual value at that frequency point was... Its corresponding weight Multiplying these values ​​yields the weighted residual for that frequency point. Finally, the weighted residuals calculated at all frequency points are summed to obtain a final scalar value that highlights low-frequency anomalies; this value is the frequency-weighted residual score. Expressed using the following formula: ; in, It is the final calculated frequency-weighted residual; At angular frequency The residual value at the point; At angular frequency The weighting function at a point is defined as the reciprocal of the angular frequency. It is worth mentioning that the reciprocal frequency weighting function used is like a mathematical magnifying glass specifically designed for observing the low-frequency range of electrochemical impedance spectroscopy; the lower the frequency of the residual, the greater the weight it is assigned.

[0031] Accordingly, in step S260, an adaptive threshold judgment is performed on the frequency-weighted residual to obtain abnormal impedance characteristics. It should be understood that using a fixed, unchanging judgment threshold cannot adapt to the constantly changing health status of the battery throughout its lifespan due to aging. Aging batteries are inherently more fragile, and their safety monitoring standards should be more stringent. A fixed high threshold may be insensitive to early faults in aging batteries, while a fixed low threshold may generate frequent false alarms for new batteries in good condition. Therefore, in the technical solution of this application, an adaptive threshold judgment is further performed on the frequency-weighted residual to obtain abnormal impedance characteristics, thereby introducing the ability to adapt to individual battery conditions into the safety monitoring system, allowing the judgment standard to be adaptively adjusted according to the actual condition of the battery. This ensures that batteries entering the end of their lifespan are monitored with greater vigilance, maintaining sensitivity to early faults while avoiding excessively frequent false alarms for new batteries in good condition, thus achieving an optimized safety monitoring strategy throughout the battery's entire lifespan.

[0032] Specifically, in this embodiment of the application, an adaptive threshold judgment is performed on the frequency-weighted residual to obtain abnormal impedance characteristics, including: determining an adaptive threshold based on battery health status indicators; determining abnormal impedance characteristics based on a comparison between the frequency-weighted residual and the adaptive threshold, wherein when the frequency-weighted residual is greater than the adaptive threshold, the abnormal impedance characteristics are determined to be abnormal.

[0033] In a specific example of this application, the implementation process of adaptive threshold judgment for frequency-weighted residuals includes the following steps: First, a dynamic threshold function associated with the battery health status index is constructed, which ensures that the older the battery and the lower its SoH value, the lower the corresponding judgment threshold. Based on the battery health status index SoH evaluated and provided by the battery management system, the adaptive threshold is determined using the following formula: ; in, It is an adaptive threshold; It is the baseline threshold for a battery in brand new condition; It is a battery health status indicator that is assessed by the battery management system and represents the percentage of the battery's current capacity relative to its factory capacity. This is a normal coefficient used to adjust the sensitivity of the threshold to changes in SoH. Subsequently, the frequency-weighted residuals calculated in the previous steps are... With the dynamically generated adaptive threshold A comparison is then made. Ultimately, based on this comparison result, the abnormal impedance characteristics are determined when the frequency-weighted residual is... Greater than the adaptive threshold If the impedance characteristics are normal, then an abnormality is determined to exist; otherwise, the abnormal impedance characteristics are determined to be normal.

[0034] By implementing the aforementioned optimization mechanism, high sensitivity and specificity are achieved in detecting early signs of thermal runaway, especially early, weak electrochemical signals related to dendrite growth. This overcomes the technical shortcomings of existing methods, which, due to their use of global, undifferentiated residual analysis, are insensitive to key risk signals and prone to misinterpretation. By introducing a frequency-resolved and weighted amplification mechanism, it can accurately focus on and amplify low-frequency anomalies that truly indicate danger from complex impedance spectral data. Combined with an adaptive threshold that dynamically adjusts according to the battery's health status, it ensures optimal monitoring sensitivity regardless of the battery's stage in its life cycle. In this way, the battery management system's early warning capability for core causes of thermal runaway, such as internal micro-short circuits, is enhanced, providing a valuable time window for subsequent proactive safety interventions. This fundamentally improves the proactive safety management level and the reliability of aqueous batteries throughout their entire life cycle.

[0035] Specifically, the process for calculating the normalized temperature rise rate is as follows: It is understandable that directly using the raw absolute value of the temperature rise rate, such as a value in degrees Celsius per second, makes its safety implications difficult to define due to variations in external factors such as ambient temperature, heat dissipation conditions, and operating conditions. Furthermore, its physical unit value cannot be directly and mathematically integrated with other dimensionless electrochemical characteristics. Therefore, in the technical solution of this application, a normalized temperature rise rate is further calculated to transform a raw thermodynamic index with physical dimensions, influenced by external operating conditions, into a standardized dimensionless characteristic value that characterizes the inherent risk of heat generation imbalance. This ensures that the thermodynamic dimension risk assessment has comparable and consistent weighting in the final multi-dimensional feature fusion, greatly improving the accuracy and adaptability of the final thermal runaway risk score.

[0036] Specifically, in a specific example of this application, the process of calculating the normalized temperature rise rate includes the following operations: First, a sliding window with a preset time length is applied to the acquired battery surface temperature sequence, for example, capturing temperature data points within the most recent 30 seconds. Then, using the temperature values ​​within the window as the dependent variable and the corresponding timestamps as the independent variables, linear regression analysis is performed to calculate the slope of the regression line, which is then determined as the current temperature rise rate. Next, the system substitutes this temperature rise rate into a preset normalization function, such as an sigmoid function, to map it to the interval between 0 and 1. Through the calculation of this function, a dimensionless numerical value is finally obtained, which is the normalized temperature rise rate.

[0037] Specifically, in step S300, a thermal runaway risk score is determined based on normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. It should be understood that since these three precursor features—normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate—characterize risks from three different dimensions—electrochemical interface kinetic degradation, internal micro-short circuit initiation, and system macroscopic thermodynamic instability—using any single feature for judgment is one-sided and cannot comprehensively and accurately reflect the overall safety status of the battery, easily leading to missed or incorrect judgments. Therefore, in the technical solution of this application, a thermal runaway risk score is further determined based on normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. This combines multiple risk indicator factors from different sources and with different physical meanings into a unified, quantitative, comprehensive risk index through weighted fusion. This enables cross-validation and complementarity of risk information from different dimensions, resulting in a more robust and reliable safety status assessment result than any single feature, providing a clear and unambiguous basis for subsequent control decisions.

[0038] More specifically, in this embodiment, the thermal runaway risk score is determined based on normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. This includes calculating a weighted sum of the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate as the thermal runaway risk score. Specifically, the process of determining the thermal runaway risk score includes the following operations: First, pre-set weighting coefficients for the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate are read from memory. These weighting coefficients are pre-calibrated based on the contribution of different characteristics to thermal runaway and their indication of urgency. Then, the three precursor feature values ​​obtained in the preceding steps are multiplied by their corresponding weighting coefficients, and the product of these three values ​​is summed to obtain the final thermal runaway risk score.

[0039] Specifically, in step S400, a control decision is made based on the thermal runaway risk score to obtain a control signal. It should be understood that since a single thermal runaway risk score is merely a quantitative value and does not directly drive any physical operation, without a clear decision-making logic to associate it with specific system responses, the score will lose its practical significance in guiding safety management. Therefore, in the technical solution of this application, a control decision is further made based on the thermal runaway risk score to obtain a control signal, thereby constructing a decision engine that maps continuous risk assessment values ​​to discrete, hierarchical system behaviors (i.e., no action, alarm, and proactive intervention). This ensures that the system can take appropriate response measures matching the severity of the risk, avoiding overreaction to low-risk states while ensuring timely and proactive intervention to high-risk states, achieving optimal allocation of safety management resources.

[0040] More specifically, in this embodiment, the control decision based on the thermal runaway risk score to obtain the control signal includes: if the thermal runaway risk score is less than the alarm threshold, the control signal is empty; if the thermal runaway risk score is between the alarm threshold and the control threshold, the control signal is a warning signal; if the thermal runaway risk score is greater than the control threshold, the control signal is a state control action. That is, in a specific example of this application, the implementation process of the control decision based on the thermal runaway risk score includes the following operations: First, the thermal runaway risk score calculated in the preceding steps is sequentially compared with the alarm threshold and the control threshold pre-stored in the memory. The decision logic is as follows: if the thermal runaway risk score is less than the alarm threshold, the system determines that the battery is in a safe operating state, the control signal is empty, and no operation is performed. If the thermal runaway risk score is greater than or equal to the alarm threshold but less than the control threshold, the system determines that the battery has shown early risk precursors, generates a control signal, which is a warning signal, used to activate the alarm indication on the user interface or send a warning message to the higher-level monitoring center. If the thermal runaway risk score is greater than or equal to the control threshold, the system determines that the battery is in a high-risk state and requires immediate active intervention. The control signal generated at this time is a state control action, which will trigger the subsequent electric field pulse application module.

[0041] Specifically, in step S500, in response to the non-empty control signal, an electric field pulse is applied to both ends of the battery. Under the excitation of the electric field pulse, dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface. It should be understood that issuing only a warning signal is a passive response; it does not fundamentally eliminate or suppress the physical causes leading to increased risk scores, such as budding dendrites or continuously deteriorating electrode interfaces. If external intervention is not timely, the risk of thermal runaway will continue to escalate. Therefore, in the technical solution of this application, in response to the non-empty control signal, an electric field pulse is applied to both ends of the battery. This transforms the abstract decision command into a physical action that can directly intervene in the microscopic electrochemical processes inside the battery, actively triggering a preset self-healing mechanism. In this way, safety management can be transformed from passive post-event warnings to proactive pre-event intervention and active in-event control, fundamentally resolving risks before they evolve into irreversible failures, thereby greatly improving the intrinsic reliability of the battery throughout its entire life cycle.

[0042] More specifically, in a concrete example of this application, when the control signal is a state control action, the process of applying electric field pulses to both ends of the battery includes the following operations. First, after receiving the state control action command, the control signal execution module in the battery management system immediately starts a high-voltage pulse generation unit. This unit generates one or a series of specific electric field pulses according to preset parameters, such as a pulse amplitude of 5 volts, a pulse width of 100 microseconds, and a pulse frequency of 1 kHz. Subsequently, the electric field pulse is precisely applied to both ends of the battery through a drive circuit connected to the positive and negative electrodes. This electric field pulse establishes an instantaneous high field strength in the electrolyte, especially at the tips of dendrites or interface defects, where the electric field lines are highly concentrated. Under the excitation of the aforementioned electric field pulse, functional molecules that were originally dormant and stable in the electrolyte gain sufficient energy to be activated and preferentially migrate and accumulate in the region with the highest field strength on the negative electrode surface. Finally, these activated functional molecules undergo in-situ electrochemical reactions on the negative electrode surface, such as electropolymerization or deposition, forming a physically robust and ionically conductive activated protective layer. This protective layer can effectively passivate the activity of dendrite tips, smooth the electrode surface, and suppress subsequent dangerous side reactions, thereby eliminating the risk of thermal runaway at the nascent stage.

[0043] In summary, the method for full life-cycle safety management of aqueous batteries according to the embodiments of this application is explained. It constructs a closed-loop proactive safety management system of perception-analysis-decision-execution to solve the problems of delayed response and insensitivity to early risk precursors in existing technologies. This solution abandons the traditional mode of relying on a single macroscopic parameter or global fuzzy analysis, and instead uses deep analysis of electrochemical impedance spectroscopy and temperature dynamics to accurately extract multiple core precursor features that characterize electrode interface degradation, the initiation of dangerous side reactions, and the trend of system thermodynamic instability. By integrating these multi-dimensional microscopic features into a comprehensive risk score, an accurate quantitative assessment of the safety status is achieved. Furthermore, this solution breaks through the limitations of passive protection. Based on the risk score, proactive decision-making is carried out. After identifying a risk, an electric field pulse can be actively applied to trigger the self-repair mechanism inside the electrolyte, forming a protective layer in situ at the source of the risk. This transforms safety management from post-event alarm to pre-event intervention, fundamentally improving the operational reliability of aqueous batteries throughout their entire life cycle.

[0044] Furthermore, a full life-cycle safety management system for aqueous batteries is also provided.

[0045] Figure 4 This is a block diagram of a water-based battery lifecycle safety management system according to an embodiment of this application. Figure 4As shown, the aqueous battery lifecycle safety management system 100 according to an embodiment of this application includes: a battery data acquisition module 110 for acquiring electrochemical impedance spectroscopy and battery surface temperature sequence; a thermal runaway precursor feature extraction module 120 for extracting thermal runaway precursor features from the electrochemical impedance spectroscopy and battery surface temperature sequence to obtain normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate; a thermal runaway risk scoring module 130 for determining a thermal runaway risk score based on the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate; a regulation decision module 140 for making regulation decisions based on the thermal runaway risk score to obtain a regulation signal; and a regulation response module 150 for applying an electric field pulse to both ends of the battery in response to the regulation signal being non-empty, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

[0046] As described above, the aqueous battery lifecycle safety management system 100 according to the embodiments of this application can be implemented in various wireless terminals, such as servers with aqueous battery lifecycle safety management algorithms. In one possible implementation, the aqueous battery lifecycle safety management system 100 according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the aqueous battery lifecycle safety management system 100 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the aqueous battery lifecycle safety management system 100 can also be one of many hardware modules of the wireless terminal.

[0047] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for full life-cycle safety management of aqueous batteries, characterized in that, include: Obtain electrochemical impedance spectroscopy and battery surface temperature sequences; Thermal runaway precursor features were extracted from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, anomalous impedance characteristics, and normalized temperature rise rate. The thermal runaway risk score is determined based on the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. Control decisions are made based on thermal runaway risk scores to obtain control signals; In response to the non-empty control signal, an electric field pulse is applied to both ends of the battery, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

2. The method for full life-cycle safety management of aqueous batteries according to claim 1, characterized in that, Precursor features of thermal runaway were extracted from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, anomalous impedance characteristics, and normalized temperature rise rate, including: Nonlinear least squares fitting was performed on the electrochemical impedance spectroscopy and baseline impedance spectroscopy to obtain the current charge transfer resistance and baseline charge transfer resistance; Determine the normalized charge transfer resistance based on the current charge transfer resistance and the baseline charge transfer resistance.

3. The method for full life-cycle safety management of aqueous batteries according to claim 2, characterized in that, Based on the current charge transfer resistance and the baseline charge transfer resistance, the normalized charge transfer resistance is determined, including calculating the normalized charge transfer resistance using the following formula: ; in, For the current charge transfer resistance, Baseline charge transfer resistance, and These are hyperparameters calibrated based on experimental data, which control the slope and center point respectively. This is the normalized charge transfer resistance.

4. The method for full life-cycle safety management of aqueous batteries according to claim 2, characterized in that, Precursor features of thermal runaway were extracted from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, anomalous impedance characteristics, and normalized temperature rise rate, including: Electrochemical impedance spectroscopy is input into the ECM model to obtain the equivalent circuit fitting impedance spectrum. Calculate the frequency-resolved residual vector between the electrochemical impedance spectrum and the equivalent circuit fitted impedance spectrum; A weighted anomaly score is applied to the frequency-resolved residual vector to obtain the frequency-weighted residual score. An adaptive threshold judgment is performed on the frequency-weighted residual to obtain the abnormal impedance characteristics.

5. The method for full life-cycle safety management of aqueous batteries according to claim 4, characterized in that, Adaptive thresholding of the frequency-weighted residuals is used to obtain abnormal impedance characteristics, including: Determine adaptive thresholds based on battery health status indicators; Abnormal impedance characteristics are determined by comparing the frequency-weighted residual with an adaptive threshold. When the frequency-weighted residual is greater than the adaptive threshold, the abnormal impedance characteristics are determined to be abnormal.

6. The method for full life-cycle safety management of aqueous batteries according to claim 5, characterized in that, Based on battery health status indicators, an adaptive threshold is determined, including: determining the adaptive threshold using the following formula, wherein the formula is: ; in, It is an adaptive threshold; It is the baseline threshold for a battery in brand new condition; It is an indicator of battery health status; This is a normal coefficient used to adjust the sensitivity of the threshold to changes in SoH.

7. The method for full life-cycle safety management of aqueous batteries according to claim 1, characterized in that, A thermal runaway risk score is determined based on normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate, including: calculating the weighted sum of normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate as the thermal runaway risk score.

8. The method for full life-cycle safety management of aqueous batteries according to claim 1, characterized in that, Control decisions are made based on thermal runaway risk scores to obtain control signals, including: If the thermal runaway risk score is less than the alarm threshold, the control signal is empty; If the thermal runaway risk score is between the alarm threshold and the control threshold, the control signal is a warning signal; If the thermal runaway risk score is greater than the control threshold, the control signal is a state control action.

9. A full life-cycle safety management system for aqueous batteries, characterized in that, include: The battery data acquisition module is used to acquire electrochemical impedance spectroscopy and battery surface temperature sequences; The thermal runaway precursor feature extraction module is used to extract thermal runaway precursor features from electrochemical impedance spectroscopy and battery surface temperature sequences to obtain normalized charge transfer resistance, abnormal impedance characteristics and normalized temperature rise rate. The thermal runaway risk scoring module is used to determine the thermal runaway risk score based on the normalized charge transfer resistance, abnormal impedance characteristics, and normalized temperature rise rate. The control decision module is used to make control decisions based on thermal runaway risk scores to obtain control signals. The regulation response module is used to apply an electric field pulse to both ends of the battery in response to the regulation signal being non-empty, wherein dormant functional molecules in the electrolyte form an activated protective layer in situ on the negative electrode surface under the excitation of the electric field pulse.

Images