FBG array battery failure early warning method, device, equipment, medium and product

Abstract

This invention discloses a battery failure early warning method, device, equipment, medium, and product based on an FBG array. The method includes acquiring multi-parameter data of a target battery using a multi-parameter FBG sensor array; wherein the multi-parameter data includes temperature data, strain data, refractive index data, and air pressure data of the target battery; performing multi-dimensional coupled feature extraction on the multi-parameter data to obtain a feature vector of the target battery; performing state identification on the feature vector to obtain the current state category and corresponding confidence level of the target battery; and performing graded early warning for the target battery based on the current state category and corresponding confidence level. Based on the multi-parameter FBG sensor array and multi-dimensional coupled feature extraction, simultaneous multi-parameter monitoring can be achieved, and the abnormal state precursors of the battery can be identified based on the coupling mechanism, thereby meeting the safety management requirements of high-performance lithium-ion batteries.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method, apparatus, device, medium, and product for early warning of battery failure based on FBG array. Background Technology

[0002] With the widespread adoption of electric vehicles and large-scale energy storage systems, the energy density and power density of lithium-ion batteries continue to improve, highlighting increasingly prominent issues related to battery thermal management and safety. Battery failure is essentially a multi-stage evolution process triggered by various factors, involving SEI (Solid Electrolyte Interphase) film decomposition, electrolyte redox, and positive electrode structure collapse. This process is accompanied by complex physicochemical changes such as drastic temperature fluctuations, electrode volume expansion and stress concentration, changes in electrolyte refractive index, and increased internal pressure.

[0003] FBG (Fiber Bragg Grating) sensors have become a research hotspot for battery internal state monitoring due to their advantages such as resistance to electromagnetic interference, high voltage resistance, small size, and single-point multiplexing. Existing FBG monitoring technologies acquire temperature and strain signals by attaching them to the battery surface or embedding them inside the battery. However, for stacked pouch lithium-ion batteries, due to the lack of inherent gaps inside, FBG sensors are difficult to embed in the active area of ​​the cell, and are only placed in the gap between the cell and the casing. This makes it impossible to directly and accurately acquire temperature and strain signals at the core location, resulting in severely insufficient monitoring accuracy. Furthermore, FBG monitoring technology is limited to macroscopic physical quantities such as temperature, strain, and pressure, lacking the key electrochemical state parameter of electrolyte refractive index. This makes it unable to capture early failure signals such as electrolyte decomposition, creating an electrochemical sensing blind spot. Simultaneously, FBG monitoring technology uses a single physical quantity threshold for early warning, ignoring the strong coupling relationships between various physical parameters inside the battery. Isolated analysis of each parameter makes it difficult to identify high-frequency weak failure characteristics, resulting in problems such as delayed warnings, high false alarm rates, and inability to identify early signs of failure. Therefore, existing FBG monitoring technology cannot achieve simultaneous monitoring of multiple parameters and accurate identification of early failure precursors based on coupling mechanisms, making it difficult to meet the safety management requirements of high-performance lithium-ion batteries. Summary of the Invention

[0004] This invention provides a battery failure early warning method, device, equipment, medium, and product based on FBG array. It can realize multi-parameter synchronous monitoring based on multi-parameter FBG sensing array and multi-dimensional coupling feature extraction, and identify the signs of abnormal battery state based on coupling mechanism, so as to meet the safety management requirements of high-performance lithium-ion batteries.

[0005] To achieve the above objectives, embodiments of the present invention provide a battery failure early warning method based on an FBG array, comprising: A multi-parameter FBG sensor array is used to acquire multi-parameter data of the target battery; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; The feature vector of the target battery is obtained by performing multi-dimensional coupled feature extraction on the multi-parameter data; The current state category and corresponding confidence level of the target battery are obtained by performing state identification on the feature vector; The target battery is given a graded warning based on the current state category and the corresponding confidence level.

[0006] As an improvement to the above scheme, the step of acquiring multi-parameter data of the target battery using a multi-parameter FBG sensor array includes: The reflectance spectrum of the target battery was acquired using a multi-parameter FBG sensor array; The characteristic wavelengths of each sensing unit in the multi-parameter FBG sensing array are identified based on the reflected spectrum. The temperature and strain data of the target battery are obtained by decoupling the characteristic wavelengths. Temperature and strain interference filtering is applied to the wavelength offset of the refractive index sensitive grating in the characteristic wavelength to obtain the refractive index data of the target battery; The gas pressure data inside the target battery is obtained by calculating the interference wavelength or phase drift of the FBG-FP gas pressure in the characteristic wavelength, so as to obtain the multi-parameter data of the target battery.

[0007] As an improvement to the above scheme, the multi-parameter FBG sensing array is composed of three different types of sensing units fused together in series at intervals along the axial direction of a single-mode fiber, including: Temperature and strain dual-parameter sensing unit, refractive index sensing unit and micro pressure sensing unit, each sensing unit is attached to the surface of the target battery or implanted between the stacked layers of the target battery in a rectangular grid distribution; The temperature and strain dual-parameter sensing unit is composed of a dual-wavelength FBG at the same location, and is used to collect temperature data and strain data of the target battery. The refractive index sensing unit adopts a side-polished FBG and spin-coates a PDMS film on the surface of the polished area of ​​the side-polished FBG to collect the refractive index data of the target battery. The micro-pressure sensing unit uses a femtosecond laser to process an FBG-FP microcavity structure and is encapsulated with a PTFE gas-permeable but liquid-impermeable sheath, and is used to collect the pressure data of the target battery.

[0008] As an improvement to the above scheme, the step of extracting the feature vector of the target battery by performing multi-dimensional coupled feature extraction on the multi-parameter data includes: The multi-parameter data is preprocessed, and mutation features are extracted from the processed multi-parameter data within a preset sliding window to obtain the mutation features of the target battery. High-frequency fluctuation features are extracted from the processed multi-parameter data within the preset sliding window to obtain the high-frequency fluctuation features of the target battery. Multi-parameter coupling features are extracted from the processed multi-parameter data within the preset sliding window to obtain the multi-parameter coupling features of the target battery. The mutation feature, the high-frequency fluctuation feature, and the multi-parameter coupling feature are fused to obtain the feature vector of the target battery.

[0009] As an improvement to the above scheme, the step of extracting multi-parameter coupling features from the processed multi-parameter data within the preset sliding window to obtain the multi-parameter coupling features of the target battery includes: A four-dimensional parametric correlation matrix of temperature, strain, refractive index, and air pressure is constructed based on the processed multi-parameter data within the preset sliding window. Calculate the maximum eigenvalue, coupling coefficient, and decoupling residual of the four-dimensional parametric correlation matrix; wherein the coupling coefficient is the temperature-strain coupling coefficient; and the decoupling residual is the refractive index-temperature decoupling residual. The maximum eigenvalue, the coupling coefficient, and the decoupling residual are normalized. The multi-parameter coupling characteristics of the target battery are constructed based on the normalized maximum eigenvalue, coupling coefficient, and decoupling residual.

[0010] As an improvement to the above scheme, the step of providing graded early warnings for the target battery based on the current state category and the corresponding confidence level includes: If the current state category is a precursor to SEI membrane rupture or a precursor to electrolyte decomposition, and the corresponding confidence level is between the first preset confidence threshold and the second preset confidence threshold, then a warning at the attention level is issued to the target battery. If any parameter feature is higher than the preset feature threshold, and the corresponding confidence level is greater than the second preset confidence threshold, then an early warning level warning is issued to the target battery. If any parameter data reaches the preset emergency threshold, or the corresponding confidence level is greater than the third preset confidence threshold, then an alarm-level warning will be issued for the target battery.

[0011] To achieve the above objectives, embodiments of the present invention provide a battery failure early warning device based on an FBG array, comprising: The parameter data acquisition module is used to acquire multi-parameter data of the target battery using a multi-parameter FBG sensor array; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; The coupling feature extraction module is used to perform multi-dimensional coupling feature extraction on the multi-parameter data to obtain the feature vector of the target battery; A battery state recognition module is used to perform state recognition on the feature vector to obtain the current state category and corresponding confidence level of the target battery; The battery classification and early warning module is used to classify and warn the target battery according to the current state category and the corresponding confidence level.

[0012] To achieve the above objectives, this invention provides a battery failure early warning device based on an FBG array, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-mentioned battery failure early warning method based on an FBG array.

[0013] To achieve the above objectives, embodiments of the present invention also provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the above-described FBG array-based battery failure early warning method.

[0014] To achieve the above objectives, embodiments of the present invention also provide a computer program product, which is stored in a storage medium and executed by at least one processor to implement the steps of the above-described FBG array-based battery failure early warning method.

[0015] Compared with existing technologies, the present invention discloses a battery failure early warning method, device, equipment, medium, and product based on an FBG array. This method acquires multi-parameter data of a target battery using a multi-parameter FBG sensing array. The multi-parameter data includes temperature data, strain data, refractive index data, and air pressure data of the target battery. Multi-dimensional coupled feature extraction is performed on the multi-parameter data to obtain a feature vector of the target battery. State identification is performed on the feature vector to obtain the current state category and corresponding confidence level of the target battery. A graded early warning is then provided for the target battery based on the current state category and corresponding confidence level. By embedding a multi-parameter FBG sensor array between the layers of a stacked soft-pack battery, four-dimensional data of temperature, strain, refractive index, and air pressure in the core area of ​​the cell are accurately acquired. Electrolyte refractive index and internal micro-pressure parameters are introduced to capture early electrochemical failure signals such as electrolyte decomposition and SEI film rupture. Through multi-dimensional coupling feature extraction, a thermo-mechanical-electrochemical parameter correlation feature vector is established to solve the problems of inaccurate internal state monitoring, inability to identify early failure precursors, delayed warnings, and high false alarm rates in stacked soft-pack batteries, thus meeting the safety management requirements of high-performance lithium-ion batteries. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a battery failure early warning method based on an FBG array provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of a battery failure early warning system based on an FBG array provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a battery failure early warning device based on an FBG array provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of a battery failure early warning device based on an FBG array provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] It should be noted that the terms "comprising" and "specific" in this invention, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0019] In this invention, "failure" mainly refers to abnormal states and early precursors related to battery safety, including but not limited to SEI film rupture, electrolyte decomposition, dendrite growth, and early thermal runaway states; the "failure" is not limited to the final state in which the battery completely loses its working ability, but also includes abnormal evolution processes that can characterize the increase in battery safety risks.

[0020] Please see Figure 1 , Figure 1 This is a flowchart illustrating a battery failure early warning method based on an FBG array provided by an embodiment of the present invention. The battery failure early warning method based on an FBG array includes: S1, a multi-parameter FBG sensor array is used to acquire multi-parameter data of the target battery; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; S2, perform multi-dimensional coupled feature extraction on the multi-parameter data to obtain the feature vector of the target battery; S3, perform state recognition on the feature vector to obtain the current state category and corresponding confidence level of the target battery; S4, classify and issue warnings for the target battery according to the current state category and the corresponding confidence level.

[0021] For example, a multi-parameter FBG sensing array integrating a dual-parameter temperature and strain sensing unit, a side-polished refractive index sensing unit, and an FBG-FP micro-pressure sensing unit is embedded in the interlayer of the target battery to simultaneously acquire four-dimensional data (multi-parameter data) of the battery's internal temperature, strain, electrolyte refractive index, and internal pressure. The sampling frequency is no less than 100Hz, and the characteristic wavelength resolution is 1pm. The acquired four-dimensional data undergoes wavelet denoising and normalization preprocessing. Abrupt change features and high-frequency fluctuation features are extracted in parallel within a preset sliding window, and a four-dimensional correlation matrix of temperature-strain-refractive index-pressure is constructed. The coupling coefficient, refractive index-temperature decoupling residual, and maximum eigenvalue of the four-dimensional correlation matrix are calculated to form a feature vector. The feature vector is input into a lightweight classifier (such as a lightweight LightGBM (Light Gradient Boosting Machine) classification model, which uses a histogram-based decision tree algorithm. In one embodiment, the trained model file can be controlled to be less than 5MB, and the single inference time can be less than 10ms, depending on the processor performance and feature dimension). The output is the current state category of the battery (normal, precursor to SEI film rupture, precursor to electrolyte decomposition, dendrite growth, early failure) and the corresponding confidence score. Based on the current state category and the corresponding confidence score, a graded warning is issued for the target battery. This invention, through the simultaneous acquisition of multi-parameter data such as temperature, strain, refractive index, and gas pressure, incorporates the electrolyte refractive index and internal micro-pressure into the monitoring range. This can be used to characterize the precursory information of abnormal states related to electrolyte decomposition, SEI film changes, and gas generation processes, thereby supplementing the information dimensions when monitoring is based solely on single physical quantities such as temperature, strain, or pressure. By extracting multi-dimensional coupling features from the multi-parameter data, compared to a single threshold judgment method, it helps to comprehensively characterize the correlation changes between thermal, mechanical, and electrochemical parameters, and can be used to identify features such as changes in coupling relationships and abnormal fluctuations. Based on the current state category and the corresponding warning confidence level, a graded warning system is implemented, which can output corresponding warning information in the early stages of abnormal evolution and execute graded responses according to the risk level. The monitoring is carried out using an FBG sensor array, which has the characteristics of anti-electromagnetic interference and small size. It can be deployed on the surface or interlayer area of ​​stacked soft-pack batteries to obtain multi-parameter state information of the target area, thereby providing technical support for the safety monitoring of power batteries and energy storage batteries.

[0022] like Figure 2 As shown, Figure 2This is a schematic diagram of a battery failure early warning system based on an FBG array provided in an embodiment of the present invention. The system includes a multi-parameter FBG sensor array, a fiber optic demodulator, and a host computer (responsible for real-time data acquisition and early warning). The fiber optic demodulator incorporates a broadband light source. Spectral signals are transmitted from the demodulator via optical fiber to an FBG (Fiber Optic Grating) sensor array embedded within the battery under test. Changes in temperature, mechanical strain, electrolyte refractive index fluctuations, and internal air pressure within the battery alter the reflection wavelength of the FBG grating. The FBG sensor array, carrying reflection spectral signals with information on temperature, strain, refractive index, and air pressure, returns to the fiber optic demodulator for high-speed acquisition. The sampling frequency is set to at least 100Hz to ensure the ability to capture transient signals. The demodulator analyzes the wavelength changes in the reflection spectral signal, demodulating the optical signal into multi-parameter time-series data (multi-parameter data) of temperature, strain, refractive index, and air pressure. This data is transmitted to a host computer via a data cable. The host computer performs real-time demodulation, feature extraction, and status identification on the acquired spectral data, enabling early-stage, graded warnings of battery failure. The monitoring results are displayed in real-time as curves and a data interface. This system achieves a closed-loop processing chain from raw optical signals to battery safety status determination, solving the problems of single monitoring parameters and delayed warnings in existing technologies.

[0023] Specifically, step S1 includes: S11 uses a multi-parameter FBG sensor array to collect the reflection spectrum of the target battery; S12, Identify the characteristic wavelength of each sensing unit in the multi-parameter FBG sensing array based on the reflection spectrum; S13, Decouple the characteristic wavelength to obtain the temperature data and strain data of the target battery; S14, Temperature and strain interference filtering is performed on the wavelength offset of the refractive index sensitive grating in the characteristic wavelength to obtain the refractive index data of the target battery; S15, calculate the interference wavelength or phase drift of the FBG-FP gas pressure in the characteristic wavelength to obtain the gas pressure data inside the target battery, so as to obtain the multi-parameter data of the target battery.

[0024] For example, light emitted from a broadband light source is transmitted through an optical fiber circulator to a multi-parameter FBG sensing array inside the battery. The array includes a temperature and strain dual-parameter grating unit (temperature and strain dual-parameter sensing unit), a refractive index sensitive grating unit (refractive index sensing unit), and an FBG-FP (Fiber Bragg Grating Fabry-Perot) pressure sensing unit (micro-pressure sensing unit). The reflection spectral signal is returned to the optical fiber demodulator, where a Gaussian fitting peak-finding algorithm is used to analyze the reflection spectrum and identify the characteristic wavelength positions corresponding to each sensing unit. For example, the two center wavelengths of the temperature / strain dual grating are identified. and The center wavelength of the refractive index sensitive grating And the interference valley wavelength of the FBG-FP barometric pressure sensor To address the issue of cross-sensitivity between temperature and strain, a dual-grating finite difference method (which separates pure temperature changes from pure strain changes by constructing a system of two linear equations and performing matrix operations) is used to decouple the temperature and strain parameters. This involves pre-calibrating the temperature sensitivity coefficients of two gratings with different center wavelengths. , and strain sensitivity coefficient , In real-time monitoring, based on the measured wavelength shift... and A system of two linear equations in two variables is constructed, and the pure temperature change is solved through matrix operations. and pure strain change This enables independent measurement of the internal temperature and stress fields of the battery.

[0025] Because the refractive index sensitive grating employs a side-polished structure, its wavelength drift is simultaneously affected by three variables: temperature, strain, and refractive index. This is achieved through calculations... and Combined with the sensitivity coefficient of the grating, the total wavelength shift of the refractive index grating is... After subtracting the interference components caused by temperature and strain, the remaining wavelength shift is only related to the external refractive index change. Based on the pre-calibrated refractive index-wavelength response curve, the remaining wavelength shift is converted into the refractive index change value of the electrolyte inside the battery. This allows us to infer the aging or decomposition state of the electrolyte.

[0026] For the FBG-FP barometric pressure sensor, pressure changes are sensed by tracking the phase or wavelength shift of specific valleys in its interference spectrum. Due to the special encapsulation of the microcavity structure, it is insensitive to strain, and temperature effects are compensated for using reference grating data. Based on the pressure-wavelength sensitivity coefficient, the wavelength change of the FBG-FP sensor is converted into the internal pressure value of the battery. The gas pressure parameter is calculated to monitor battery gas production, and the temperature of the battery monitoring point is output in the same time dimension. ,strain Refractive index air pressure The four-dimensional data set (multi-parameter data) completes the collaborative acquisition of multiple parameters in a single sampling cycle. This invention introduces refractive index and micro-pressure parameters to construct a complete thermo-mechanical-electrochemical sensing dimension, solving the problem of failing to identify early electrochemical failures such as electrolyte decomposition.

[0027] Specifically, the multi-parameter FBG sensing array is composed of three different types of sensing units fused together in series at intervals along the axial direction of a single-mode fiber, including: Temperature and strain dual-parameter sensing unit, refractive index sensing unit and micro pressure sensing unit, each sensing unit is attached to the surface of the target battery or implanted between the stacked layers of the target battery in a rectangular grid distribution; The temperature and strain dual-parameter sensing unit is composed of a dual-wavelength FBG at the same location, and is used to collect temperature data and strain data of the target battery. The refractive index sensing unit adopts a side-polished FBG and spin-coates a PDMS film on the surface of the polished area of ​​the side-polished FBG to collect the refractive index data of the target battery. The micro-pressure sensing unit uses a femtosecond laser to process an FBG-FP microcavity structure and is encapsulated with a PTFE gas-permeable but liquid-impermeable sheath, and is used to collect the pressure data of the target battery.

[0028] For example, the multi-parameter FBG sensing array is composed of three different types of sensing units fused together in series along the axial direction of a single-mode fiber. The spacing between the sensing units in the multi-parameter FBG sensing array is 10mm to 50mm (depending on the cell size). The temperature and strain dual-parameter sensing unit can be a common single-mode FBG with a center wavelength in the 1550nm band, and the grating length is set to 10mm to ensure sufficient reflectivity and bandwidth. To solve the problem of cross-sensitivity between temperature and strain, two FBGs with a center wavelength difference greater than 10nm are arranged near the same location (wavelength planning is performed based on the total number of array channels, demodulation band, and reflection spectrum bandwidth of each sensing unit). The difference in the sensitivity coefficients of the two FBGs to temperature and strain is used to construct a set of equations, and matrix operations are used to decouple and separate the temperature and strain signals.

[0029] The residual cladding thickness in the side-polished area of ​​the refractive index sensing unit is 1μm to 3μm, preferably 2μm. The PDMS (Polydimethylsiloxane) film thickness is preferably 100nm to 2μm, more preferably about 500nm. The refractive index sensing unit adopts a side-polished FBG, in which the cladding on one side of the optical fiber is removed through a precision polishing process until the residual cladding thickness is controlled between 1μm and 3μm, allowing the external electrolyte to directly modulate the evanescent field of the grating. To prevent the electrolyte from corroding the optical fiber core and to enhance the signal-to-noise ratio, a PDMS film with a thickness of about 500nm is spin-coated on the surface of the polished area. This film has good permeability to changes in electrolyte composition and can isolate corrosive substances. Its characteristic of changing refractive index with electrolyte concentration amplifies the sensing sensitivity, enabling the unit to accurately capture the slight refractive index drift in the early stage of electrolyte decomposition.

[0030] The micro-pressure sensing unit features an FBG-FP microcavity with an air gap length of 50 μm to 100 μm. This unit monitors gas generation reactions within the battery and employs an intrinsic FBG-FP cavity structure based on femtosecond laser micromachining technology. This structure forms an FP interferometer by etching an air gap microcavity with a length of 50 to 100 micrometers between two FBG segments. When a side reaction occurs inside the battery, generating gas and causing an increase in pressure, the gas density within the microcavity changes, altering the refractive index and resulting in a phase or wavelength shift in the interference spectrum. To ensure accurate internal pressure sensing without being affected by electrolyte immersion, the sensing unit is encapsulated within a micro-sheath made of a PTFE (Polytetrafluoroethylene) microporous membrane that is permeable to air but impermeable to liquid. Each sensing unit is attached to the battery surface or directly embedded between battery stack layers according to the rectangular grid distribution rule based on the battery size, forming a collaborative sensing network covering the entire battery area. The sensor spacing is adjustable from 10mm to 50mm to balance measurement accuracy and cost.

[0031] Specifically, step S2 includes: S21, preprocess the multi-parameter data, extract mutation features from the processed multi-parameter data within a preset sliding window, and obtain the mutation features of the target battery. S22, High-frequency fluctuation features are extracted from the processed multi-parameter data within the preset sliding window to obtain the high-frequency fluctuation features of the target battery. S23, perform multi-parameter coupling feature extraction on the processed multi-parameter data within the preset sliding window to obtain the multi-parameter coupling features of the target battery; S24, the mutation feature, the high-frequency fluctuation feature and the multi-parameter coupling feature are fused to obtain the feature vector of the target battery.

[0032] For example, the preprocessing includes Gaussian fitting for peak finding, wavelet thresholding for denoising, and normalization; the abrupt change features are extracted using the CUSUM (Cumulative Sum) algorithm; and the high-frequency fluctuation features are extracted using FFT (Fast Fourier Transform) and sample entropy calculation. The fiber optic demodulator acquires reflectance spectral data through a high-speed photodetector array, uses a Gaussian fitting algorithm to accurately locate the peak wavelengths of each FBG and the interference valley positions of the FP cavity, with a resolution of 1 pm. Using a pre-calibrated sensitivity coefficient matrix, the wavelength data is decoupled and converted into temperature data. ,strain Refractive index air pressure The four-dimensional physical parameter time series is obtained. To eliminate environmental noise interference, a wavelet threshold denoising algorithm is introduced. The db4 wavelet basis is selected to perform three-level decomposition and reconstruction of the original signal. The processed multi-parameter data is calculated based on the sliding window technique. The window length is preferably 2 to 10 seconds, more preferably 5 to 10 seconds, and the sliding step size is set to 1 second to ensure the real-time performance of feature extraction.

[0033] Within each sliding window, three types of feature vectors are computed in parallel. The first type is abrupt change features, used to identify step events such as SEI membrane rupture. For example, the CUSUM algorithm is used to monitor the processed multi-parameter data and calculate statistics. ,in, This is the historical average. To allow for deviation, when When the preset alarm threshold is exceeded, it is identified as a sudden change point, and its amplitude and duration are recorded; the second type is high-frequency fluctuation characteristics, which are used to capture random processes such as dendrite growth. A fast Fourier transform is performed on the signal within the window to calculate the energy spectral density integral value in the high-frequency band from 10Hz to 45Hz, and the sample entropy of the signal is calculated to quantify the complexity and irregularity of the signal; the third type is multi-parameter coupling characteristics, which construct a four-dimensional correlation matrix of temperature-strain-refractive index-pressure. Calculate its largest eigenvalue The condition number is used to determine the significant jump in the distribution of eigenvalues ​​of the matrix when the internal coupling mechanism of the battery changes. The ratio of the temperature change rate to the strain change rate is also calculated as the coupling coefficient to distinguish between normal thermal expansion and abnormal mechanical deformation.

[0034] More specifically, step S23 includes: S231, construct a four-dimensional parameter correlation matrix of temperature, strain, refractive index and air pressure based on the processed multi-parameter data within the preset sliding window; S232, calculate the maximum eigenvalue, coupling coefficient, and decoupling residual of the four-dimensional parametric correlation matrix; wherein, the coupling coefficient is the temperature-strain coupling coefficient; and the decoupling residual is the refractive index-temperature decoupling residual; S233, normalize the maximum eigenvalue, coupling coefficient and decoupling residual; S234, construct the multi-parameter coupling characteristics of the target battery based on the normalized maximum eigenvalue, temperature-strain coupling coefficient, and refractive index-temperature decoupling residual.

[0035] For example, select the temperature sequence within the current sliding window. Strain sequence Refractive index sequence Pressure sequence The mean and standard deviation of each sequence were calculated and normalized. Then, the correlation coefficient between any two variables was calculated using the Pearson correlation coefficient formula, such as the correlation coefficient between temperature and strain. Correlation coefficient between temperature and refractive index Etc., construct a 4×4 symmetric matrix. (A four-dimensional parametric correlation matrix), where the elements on the main diagonal are all 1s, and the off-diagonal elements reflect the degree of linear correlation between the physical quantities. Regarding the correlation matrix... Perform eigenvalue decomposition to obtain a set of eigenvalues. Extract the largest eigenvalue As a primary coupling characteristic, during normal battery operation, parameters such as temperature and strain are highly coupled, and the main energy of the signal is concentrated in one principal direction. The values ​​are relatively large and stable; however, when early failure occurs inside the battery (such as a micro-short circuit causing localized temperature rise without overall strain), the synergistic relationship between the parameters is broken, signal independence increases, and the correlation matrix tends to diagonalize, leading to... Significantly reduced, through monitoring The rate of decrease is used to determine the degree of instability of the coupling mechanism. Differential operations are performed on the temperature and strain signals respectively to obtain the temperature change rate sequence. and strain rate of change sequence Calculate the ratio of the two. As a temperature-strain coupling coefficient, during normal charge and discharge, this coefficient is governed by the thermal expansion coefficient of the battery material and remains within a specific range. However, under abnormal conditions such as lithium plating or mechanical extrusion, strain growth no longer solely depends on thermal expansion, leading to... Abnormal jumps or deviations from the normal threshold range are detected to differentiate between normal thermal expansion and abnormal mechanical deformation. Considering that the refractive index is affected by both temperature and electrolyte concentration, a temperature sequence is used. For refractive index sequence Linear regression fitting was performed to establish a refractive index-temperature baseline model under normal conditions. Calculate the measured refractive index Compared with model predictions Refractive index-temperature decoupling residuals between In some embodiments, the RMS (Root Mean Square) of the decoupled residual sequence can also be used as an auxiliary statistical indicator. When the electrolyte is not decomposed, the residual RMS approaches zero; when the electrolyte decomposes and the component concentration changes, the refractive index will produce an additional drift independent of temperature, resulting in a significant increase in the decoupled residual, thereby achieving accurate capture of electrolyte failure precursors. This invention exploits the coupling characteristics between temperature, strain, and refractive index (such as the temperature-strain coupling coefficient and the refractive-temperature decoupled residual). Based on physical mechanisms, these characteristics are more sensitive to failure precursors and have strong anti-interference capabilities. By capturing weak abrupt changes in coupling relationships, early warnings can be issued minutes to hours before failure occurs, and the multi-parameter interlocking logic significantly reduces the false alarm rate.

[0036] Specifically, step S4 includes: S41, if the current state category is a precursor to SEI membrane rupture or a precursor to electrolyte decomposition, and the corresponding confidence level is between the first preset confidence threshold and the second preset confidence threshold, then a warning at the attention level is issued to the target battery. S42, if any parameter feature is higher than the preset feature threshold and the corresponding confidence level is greater than the second preset confidence threshold, then a warning level warning is issued to the target battery. S43, if any parameter data reaches the preset emergency threshold, or the corresponding confidence level is greater than the third preset confidence threshold, then an alarm-level warning is issued for the target battery.

[0037] For example, if the current state category is a precursor to SEI film rupture or electrolyte decomposition, and the corresponding confidence level is between a first preset confidence threshold (which can be set as needed, such as 0.6) and a second preset confidence threshold (which can be set as needed, such as 0.8), then a attention-level warning is issued to the target battery, such as increasing the sampling frequency and marking the abnormal area of ​​the target battery; if any parameter feature (such as refractive index drift rate) is higher than a preset feature threshold (which can be set as needed), and the corresponding confidence level is greater than the second preset confidence threshold, then a warning-level warning is issued to the target battery, such as triggering an audible and visual alarm and prompting manual intervention; if any parameter data (such as temperature or air pressure) reaches a preset emergency threshold (which can be set as needed), or the corresponding confidence level is greater than a third preset confidence threshold (which can be set as needed, such as 0.9), then an alarm-level warning is issued to the target battery, such as alerting the BMS (Battery Management System). The battery management system sends a cut-off command and highlights the location of the fault on the heat map of the visualization interface to achieve precise and proactive safety protection.

[0038] This invention discloses a battery failure early warning method based on an FBG array. It acquires multi-parameter data of a target battery using a multi-parameter FBG sensor array. This multi-parameter data includes temperature, strain, refractive index, and pressure data of the target battery. Multi-dimensional coupling feature extraction is performed on the multi-parameter data to obtain a feature vector of the target battery. State identification is performed on the feature vector to obtain the current state category and corresponding confidence level of the target battery. A graded early warning is then issued for the target battery based on the current state category and corresponding confidence level. By embedding a multi-parameter FBG sensor array between the layers of a stacked soft-pack battery, it accurately acquires four-dimensional data (temperature, strain, refractive index, and pressure) of the core area of ​​the cell. Electrolyte refractive index and internal micro-pressure parameters are introduced to capture early electrochemical failure signals such as electrolyte decomposition and SEI film rupture. Through multi-dimensional coupling feature extraction, a thermo-mechanical-electrochemical parameter correlation feature vector is established. This method solves the problems of inaccurate internal state monitoring, inability to identify early failure precursors, delayed early warning, and high false alarm rate in stacked soft-pack batteries, thus meeting the safety management requirements of high-performance lithium-ion batteries.

[0039] See Figure 3 , Figure 3 This is a schematic diagram of a battery failure early warning device 10 based on an FBG array provided in an embodiment of the present invention. The battery failure early warning device 10 based on an FBG array includes: The parameter data acquisition module 11 is used to acquire multi-parameter data of the target battery using a multi-parameter FBG sensor array; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; The coupling feature extraction module 12 is used to perform multi-dimensional coupling feature extraction on the multi-parameter data to obtain the feature vector of the target battery; The battery state recognition module 13 is used to perform state recognition on the feature vector to obtain the current state category and corresponding confidence level of the target battery; The battery classification and early warning module 14 is used to classify and warn the target battery according to the current state category and the corresponding confidence level.

[0040] The FBG array-based battery failure early warning device 10 provided in this embodiment of the invention can realize all the processes of the FBG array-based battery failure early warning method of the above embodiments. The functions and technical effects of each module in the device are the same as the functions and technical effects of the FBG array-based battery failure early warning method of the above embodiments, and will not be repeated here.

[0041] See Figure 4 , Figure 4 This is a schematic diagram of the structure of an FBG array-based battery failure early warning device 20 provided in an embodiment of the present invention. The FBG array-based battery failure early warning device 20 of this embodiment includes: a processor 21, a memory 22, and a computer program stored in the memory 22 and executable on the processor 21. When the processor 21 executes the computer program, it implements the steps in the above-described FBG array-based battery failure early warning method embodiment. Alternatively, when the processor 21 executes the computer program, it implements the functions of each module in the above-described FBG array-based battery failure early warning device embodiment.

[0042] For example, the computer program may be divided into one or more modules, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the FBG array-based battery failure early warning device 20.

[0043] The FBG array-based battery failure early warning device 20 can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. The FBG array-based battery failure early warning device 20 may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will understand that the schematic diagram is merely an example of the FBG array-based battery failure early warning device 20 and does not constitute a limitation on the FBG array-based battery failure early warning device 20. It may include more or fewer components than shown, or combine certain components, or use different components. For example, the FBG array-based battery failure early warning device 20 may also include input / output devices, network access devices, buses, etc.

[0044] The processor 21 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 21 is the control center of the FBG array-based battery failure early warning device 20, connecting all parts of the device through various interfaces and lines.

[0045] The memory 22 can be used to store the computer programs and / or modules. The processor 21 implements various functions of the FBG array-based battery failure warning device 20 by running or executing the computer programs and / or modules stored in the memory 22 and calling the data stored in the memory 22. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0046] The module integrated into the FBG array-based battery failure early warning device 20, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by the processor 21, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable medium may be appropriately added to or subtracted from the content as required by the legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium may not include electrical carrier signals and telecommunication signals.

[0047] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0048] This invention also provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the FBG array-based battery failure early warning method as described in the above embodiments.

[0049] Furthermore, embodiments of the present invention also provide a computer program product, which is stored in a storage medium and executed by at least one processor to implement the steps of the FBG array-based battery failure early warning method described above.

[0050] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A battery failure early warning method based on FBG array, characterized in that, include: A multi-parameter FBG sensor array is used to acquire multi-parameter data of the target battery; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; The feature vector of the target battery is obtained by performing multi-dimensional coupled feature extraction on the multi-parameter data; The current state category and corresponding confidence level of the target battery are obtained by performing state identification on the feature vector; The target battery is given a graded warning based on the current state category and the corresponding confidence level.

2. The battery failure early warning method based on FBG array as described in claim 1, characterized in that, The method of acquiring multi-parameter data of the target battery using a multi-parameter FBG sensor array includes: The reflectance spectrum of the target battery was acquired using a multi-parameter FBG sensor array; The characteristic wavelengths of each sensing unit in the multi-parameter FBG sensing array are identified based on the reflected spectrum. The temperature and strain data of the target battery are obtained by decoupling the characteristic wavelengths. Temperature and strain interference filtering is applied to the wavelength offset of the refractive index sensitive grating in the characteristic wavelength to obtain the refractive index data of the target battery; The gas pressure data inside the target battery is obtained by calculating the interference wavelength or phase drift of the FBG-FP gas pressure in the characteristic wavelength, so as to obtain the multi-parameter data of the target battery.

3. The battery failure early warning method based on FBG array as described in claim 1, characterized in that, The multi-parameter FBG sensor array is composed of three different types of sensing units fused together in series at intervals along the axial direction of a single-mode fiber, including: Temperature and strain dual-parameter sensing unit, refractive index sensing unit and micro pressure sensing unit, each sensing unit is attached to the surface of the target battery or implanted between the stacked layers of the target battery in a rectangular grid distribution; The temperature and strain dual-parameter sensing unit is composed of a dual-wavelength FBG at the same location, and is used to collect temperature data and strain data of the target battery. The refractive index sensing unit adopts a side-polished FBG and spin-coates a PDMS film on the surface of the polished area of ​​the side-polished FBG to collect the refractive index data of the target battery. The micro-pressure sensing unit uses a femtosecond laser to process an FBG-FP microcavity structure and is encapsulated with a PTFE gas-permeable but liquid-impermeable sheath, and is used to collect the pressure data of the target battery.

4. The battery failure early warning method based on FBG array as described in claim 1, characterized in that, The step of extracting the feature vector of the target battery by performing multi-dimensional coupled feature extraction on the multi-parameter data includes: The multi-parameter data is preprocessed, and mutation features are extracted from the processed multi-parameter data within a preset sliding window to obtain the mutation features of the target battery. High-frequency fluctuation features are extracted from the processed multi-parameter data within the preset sliding window to obtain the high-frequency fluctuation features of the target battery. Multi-parameter coupling features are extracted from the processed multi-parameter data within the preset sliding window to obtain the multi-parameter coupling features of the target battery. The mutation feature, the high-frequency fluctuation feature, and the multi-parameter coupling feature are fused to obtain the feature vector of the target battery.

5. The battery failure early warning method based on FBG array as described in claim 4, characterized in that, The step of extracting multi-parameter coupling features from the processed multi-parameter data within the preset sliding window to obtain the multi-parameter coupling features of the target battery includes: A four-dimensional parametric correlation matrix of temperature, strain, refractive index, and air pressure is constructed based on the processed multi-parameter data within the preset sliding window. Calculate the maximum eigenvalue, coupling coefficient, and decoupling residual of the four-dimensional parametric correlation matrix; wherein the coupling coefficient is the temperature-strain coupling coefficient; and the decoupling residual is the refractive index-temperature decoupling residual. The maximum eigenvalue, the coupling coefficient, and the decoupling residual are normalized. The multi-parameter coupling characteristics of the target battery are constructed based on the normalized maximum eigenvalue, coupling coefficient, and decoupling residual.

6. The battery failure early warning method based on FBG array as described in claim 1, characterized in that, The step of classifying and issuing early warnings for the target battery based on the current state category and the corresponding confidence level includes: If the current state category is a precursor to SEI membrane rupture or a precursor to electrolyte decomposition, and the corresponding confidence level is between the first preset confidence threshold and the second preset confidence threshold, then a warning at the attention level is issued to the target battery. If any parameter feature is higher than the preset feature threshold, and the corresponding confidence level is greater than the second preset confidence threshold, then an early warning level warning is issued to the target battery. If any parameter data reaches the preset emergency threshold, or the corresponding confidence level is greater than the third preset confidence threshold, then an alarm-level warning will be issued for the target battery.

7. A battery failure early warning device based on FBG array, characterized in that, include: The parameter data acquisition module is used to acquire multi-parameter data of the target battery using a multi-parameter FBG sensor array; wherein, the multi-parameter data includes temperature data, strain data, refractive index data and air pressure data of the target battery; The coupling feature extraction module is used to perform multi-dimensional coupling feature extraction on the multi-parameter data to obtain the feature vector of the target battery; A battery state recognition module is used to perform state recognition on the feature vector to obtain the current state category and corresponding confidence level of the target battery; The battery classification and early warning module is used to classify and warn the target battery according to the current state category and the corresponding confidence level.

8. A battery failure early warning device based on FBG array, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the FBG array-based battery failure early warning method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the FBG array-based battery failure early warning method as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product is stored in a storage medium, and the program product is executed by at least one processor to implement the steps of the FBG array-based battery failure early warning method as described in any one of claims 1-6.

Images