A battery micro-overcharge damage detection method and system based on optical fiber sensing

Abstract

The present application belongs to the field of battery damage detection, and provides a battery micro-overcharge damage detection method and system based on optical fiber sensing, which acquires and decouples the temperature and strain parameters of the battery surface; based on the acquired data, the strain change rate of the characteristic voltage band is extracted as a detection feature, an adaptive dynamic detection threshold curve based on B-spline function is trained, the shape of the threshold curve is optimized by minimizing the classification loss function, and the optimal dynamic threshold curve is obtained; the optimal dynamic threshold curve is used as the detection basis to calculate the battery micro-overcharge index, and it is judged whether the battery micro-overcharge index exceeds the safety threshold, if it exceeds, the battery micro-overcharge damage is determined. The present application uses an adaptive dynamic threshold optimization algorithm to obtain the optimal classification threshold curve, and realizes the rapid and accurate classification of the micro-overcharge battery charging sequence.

Description

Technical Field

[0001] This invention belongs to the field of battery damage detection, specifically relating to a method and system for detecting micro-overcharge damage in batteries based on fiber optic sensing. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Lithium-ion batteries play a crucial supporting role in electric vehicles and electrochemical energy storage due to their high energy density, long lifespan, and environmental friendliness. In recent years, the demand for high-performance energy storage solutions from new energy vehicles, new energy power generation, and consumer electronics has exploded, collectively forming the core driving force for the continued expansion of the lithium-ion battery market. However, as a typical nonlinear time-varying complex system, lithium-ion batteries inevitably accumulate micro-overcharge damage during long-term service. Micro-overcharging is an extremely common and insidious form of electrical abuse, defined as a situation where a battery, after being fully charged, is continuously charged with a small amount of electricity, while the battery voltage does not significantly exceed its upper limit. This behavior leads to irreversible loss of active materials and active lithium ions. If a battery continuously experiences micro-overcharging, its lifespan will be significantly reduced, damage will accumulate, and the risk of thermal runaway will increase.

[0004] As a highly concealed form of damage, micro-overcharge damage exhibits external signals such as voltage and temperature almost identical to those of aged batteries. This makes it difficult to directly characterize micro-overcharge damage using traditional battery parameters, resulting in significant challenges for online diagnosis. Existing micro-overcharge damage detection methods place excessive demands on equipment computational performance, hindering accurate online diagnosis. Furthermore, most existing methods rely on traditional electrical parameters, which are insufficient for effectively characterizing the highly concealed micro-overcharge damage in batteries. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a method and system for detecting micro-overcharge damage in batteries based on fiber optic sensing. This invention acquires high-precision strain changes on the battery surface; extracts the strain change rate of characteristic voltage segments as detection features; and utilizes an adaptive dynamic threshold optimization algorithm to obtain the optimal classification threshold curve, thereby achieving rapid and accurate classification of charging sequences of micro-overcharged batteries.

[0006] According to some embodiments, the present invention adopts the following technical solution: A method for detecting micro-overcharge damage in batteries based on fiber optic sensing includes the following steps: Acquire and decouple battery surface temperature and strain parameters; Based on the acquired data, the strain change rate of the characteristic voltage band is extracted as the detection feature. An adaptive dynamic detection threshold curve based on the B-spline function is trained. The shape of the threshold curve is optimized by minimizing the classification loss function, and the optimal dynamic threshold curve is obtained by solving the problem. Based on the optimal dynamic threshold curve, the battery micro-overcharge index is calculated to determine whether the battery micro-overcharge index exceeds the safety threshold. If it does, the battery micro-overcharge damage is determined.

[0007] As an alternative implementation, the process of acquiring and decoupling the battery surface temperature and strain parameters includes: measuring the parameters at the geometric center of the front of the pouch battery using a dual fiber Bragg grating sensor. The two fiber Bragg grating sensors have the same parameters. The fiber Bragg grating sensors achieve selective reflection of a predetermined light wavelength through periodic modulation of the refractive index. The reflected wavelength is modulated by external physical quantities. The physical quantities are decoupled by detecting the wavelength change of the reflected light using a fiber demodulator.

[0008] As an alternative implementation, the optical fibers in the two fiber Bragg grating sensors are a bare optical fiber and a polyimide-sheathed optical fiber, respectively. The reflection wavelength of the bare optical fiber is affected by both temperature and strain, while the reflection wavelength of the sheathed optical fiber is controlled only by temperature.

[0009] As an alternative implementation, the process of training an adaptive dynamic detection threshold curve based on a B-spline function includes: setting the threshold curve as... T (V), where V represents the voltage independent variable, and the adaptive dynamic threshold curve optimization algorithm parameterizes the threshold curve through B-spline interpolation: ; ; ; ; In the formula, Let B-spline be the control point parameter vector. k Let M be the node vector, and M be the number of control points. B m (V; k ) is the first m There are B-spline basis functions, where V(max) and V(min) represent the maximum and minimum values ​​of the voltage range, respectively.

[0010] As an alternative implementation, the process of optimizing the shape of the threshold curve by minimizing the classification loss function to obtain the optimal dynamic threshold curve includes: the objective function of the optimization problem includes a classification loss term, a maximum violation penalty term, and a smoothness regularization term.

[0011] As a further defined implementation method, the classification loss function is: ; in, J class ( θ ), J max ( θ )and J smooth ( θ These are the classification loss term, the maximum violation penalty term, and the smoothness regularization term, respectively. The classification loss term is: For each sequence i Its degree of violation is defined as: For positive samples with a label of 1, i.e., there is micro-overfilling impairment, ideally all points should be above the threshold curve: ; For negative samples with a label of 0, i.e., normal samples, ideally all points should be below the threshold curve: ; In the formula, and The first i The first sequence j The voltage value and the first j Each strain rate of change value, For voltage Threshold estimate of the rate of change of strain at a given point; sequence i The total loss is: ; In the formula N i Let i be the length of sequence i; The total category loss is: ; In the formula N The total number of training sequences; The maximum penalty for violation is: ; coefficient α Control the weighting percentage of this item; The smoothness regularization term is a second-order difference regularization term: ; In the formula, V k For the smoothness evaluation point, the second-order difference approximation is: ; in, λ represents the voltage step size between adjacent evaluation points, and λ is the weight of the smoothness regularization term.

[0012] As an alternative implementation method, in the process of solving for the optimal dynamic threshold curve... θ The following physical constraints must be met: ; in, For the first i The first sequence j For each strain rate of change, the interior point method is used in the optimization solution. The constraint is incorporated into the objective function through the obstacle function, and the optimal solution is iteratively approximated along the central path.

[0013] As an alternative implementation method, the process of calculating the battery micro-overcharge index based on the optimal dynamic threshold curve includes: The micro-overcharge index MOI is: ; n V is the length of the sequence. l and d S l These are the first and second digits of the sequence, respectively. l The voltage value and strain rate of change at each data point, 1( ) is an indicator function, when Returns 1 if the condition is met, otherwise returns 0.

[0014] A battery micro-overcharge damage detection system based on fiber optic sensing includes: The data acquisition module is used to acquire and decouple the battery surface temperature and strain parameters; The optimal dynamic threshold curve calculation module is used to extract the strain change rate of the characteristic voltage band as the detection feature based on the acquired data, train the adaptive dynamic detection threshold curve based on the B-spline function, optimize the shape of the threshold curve by minimizing the classification loss function, and solve for the optimal dynamic threshold curve. The micro-overcharge damage detection module is used to calculate the battery micro-overcharge index based on the optimal dynamic threshold curve, and to determine whether the battery micro-overcharge index exceeds the safety threshold. If it does, the battery micro-overcharge damage is determined.

[0015] The data acquisition module includes two fiber Bragg grating sensors with identical parameters, arranged in parallel. One of them is a bare optical fiber, and the other is a polyimide-sheathed protective optical fiber. Both fiber Bragg grating sensors are connected to an optical fiber demodulator.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention eliminates the need to disassemble and damage the battery; it enables online detection of micro-overcharge damage solely by utilizing data from the battery's normal constant-current charging voltage plateau period.

[0017] This invention has high sensitivity to micro-overcharge damage and can accurately detect micro-overcharge damage. Even after only one cycle of micro-overcharge exceeding the upper limit by 0.1V, this invention can still achieve accurate detection in the next cycle.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0020] Figure 1 This is a flowchart of a lithium-ion battery micro-overcharge damage detection method according to one embodiment. Figure 2 This is a schematic diagram of a dual-fiber sensing principle according to one embodiment; Figure 3 Here is an example of a battery #1 IC curve; Figure 4 The image shows the temperature and strain curves at the measuring point #1 of the battery in one embodiment, where (a) is the temperature curve and (b) is the strain curve. Figure 5 The following is a micro-overcharge diagnostic result of one embodiment, wherein (a) is the strain change rate of battery #1, (b) is the diagnostic result of battery #1, (c) is the strain change rate of battery #2, (d) is the diagnostic result of battery #2, (e) is the strain change rate of battery #3, and (f) is the diagnostic result of battery #3. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0022] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0023] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0024] Where there is no conflict, the embodiments and features described in this application may be combined with each other.

[0025] Example 1 A method for detecting micro-overcharge damage in batteries based on fiber optic sensing, such as Figure 1 As shown, firstly, a dual-fiber sensing scheme is used to acquire and decouple the battery surface temperature and strain parameters. Then, the strain change rate of the characteristic voltage band is extracted as a detection feature, and an adaptive dynamic detection threshold curve based on B-spline functions is trained. Finally, using the optimal dynamic threshold curve as the detection basis, the Mild Overcharge Index (MOI) is defined, and accurate detection of battery micro-overcharge damage is achieved by judging whether the MOI exceeds the safety threshold.

[0026] The main steps are as follows: First, parameters at the geometric center of the front of the pouch battery are measured using a dual Bragg grating (FBG) sensing scheme. The principle is as follows: Figure 2 As shown. The two fiber optic sensors have identical parameters, differing only in the presence or absence of a protective sheath. The FBG sensor achieves selective reflection of specific light wavelengths through periodic modulation of the refractive index. Its reflected wavelength is modulated by external physical quantities such as temperature and strain. The wavelength change of the reflected light is detected by a fiber optic demodulator, thereby decoupling physical quantities such as temperature and strain. The two optical fibers used in this invention are a bare fiber and a polyimide-sheathed fiber. The reflected wavelength of the bare fiber is simultaneously affected by both temperature and strain, while the wavelength change of the sheathed fiber is controlled solely by temperature because it is isolated from strain. The specific measurement principles of both are shown in the following equations: (1) (2) In the formula, Δ λ TS and λ TS0 These represent the change in the reflected wavelength of the bare fiber and the initial reflected wavelength, respectively; Δ λ T and λ T0These are the change in reflected wavelength of the sheathed optical fiber and the initial reflected wavelength, respectively. p e It is the strain optical sensitivity coefficient, Δ ε It is the strain change experienced by the grating. α f and ξ These are the coefficient of thermal expansion and the temperature optical sensitivity coefficient, respectively. Since the two optical fibers have identical manufacturing parameters, for the same measurement point, there are... λ TS0 = λ T0 = λ 0. Solving the simultaneous equations, the strain and temperature changes are as follows: (3) (4) An Adaptive Dynamic Threshold Curve Optimization (ADTCO) algorithm is proposed to achieve binary classification of strain rate change sequences during battery charging by finding an optimal dynamic threshold curve. The core idea of ​​the algorithm is to parameterize the threshold curve as a B-spline function and optimize its shape by minimizing the classification loss function, enabling it to effectively distinguish between battery sequence samples that have undergone micro-overcharging.

[0027] Assuming the threshold curve is T (V), where V represents the voltage independent variable, and ADTCO parameterizes the threshold curve using B-spline interpolation: (5) (6) (7) (8) In the formula, Let B-spline be the control point parameter vector. k Let M be the node vector, and M be the number of control points. B m (V; k ) is the first m There are B-spline basis functions, where V(max) and V(min) represent the maximum and minimum values ​​of the voltage range, respectively.

[0028] In this embodiment, the objective function of the optimization problem consists of the following three parts: (9) In the formula, J class ( θ), J max ( θ )and J smooth ( θ These are the classification loss term, the maximum violation penalty term, and the smoothness regularization term, respectively.

[0029] The specific calculations for each item are shown below: Classification loss item J class ( θ ) For each sequence i Its degree of violation is defined as: For a positive sample with a label of 1 (exhibiting micro-overfilling impairment), ideally all points would be above the threshold curve: (10) For negative samples (normal samples) with a label of 0, ideally all points should be below the threshold curve: (11) In the formula, and The first i The first sequence j The voltage value and the first j Each strain rate of change value, For voltage Threshold estimate of the rate of change of strain.

[0030] sequence i The total loss is: (12) In the formula N i Let i be the length of sequence i.

[0031] The total category loss is: (13) In the formula N This represents the total number of training sequences.

[0032] Maximum penalty for violation J max ( θ ) To prevent the influence of extreme outliers in the sequence, the following maximum violation penalty is added to the optimization objective function: (14) Through coefficients αThe weighting of this item is controlled. Since the physical quantities measured by the FBG sensor are relatively stable and the proportion of abnormal extreme values ​​is small, in this embodiment... α Only 0.1 is taken.

[0033] Smoothness regularization term J smooth ( θ ) To ensure the smoothness of the optimal threshold curve, the following second-order difference regularization term is introduced: (15) In the formula, V k For the smoothness evaluation point, the second-order difference approximation is: (16) In this embodiment, the smoothness regularization term weight λ = 0.01.

[0034] The ADTCO algorithm optimizes variables. θ To minimize the objective function J ( θ ), can be represented as: (17) In the formula, θ The following physical constraints must be met: (18) In optimization problems, the interior-point method is used. The interior-point method is an efficient algorithm for solving linear programming and convex optimization problems. It incorporates constraints into the objective function through obstacle functions, iteratively approximating the optimal solution along the central path, and is widely used in engineering and scientific computing.

[0035] For the optimal threshold curve problem, the interior point method calculates sequentially in each iteration. J ( θ And update parameters θ Until convergence. Ultimately, the optimal parameter vector is obtained. θ * Output optimal threshold curve T (V).

[0036] For a certain battery charging segment, the strain rate sequence (V, d) S The micro-overcharge index MOI is defined as: (19) (20) In the formula, n V is the length of the sequence. l and d S l These are the first and second digits of the sequence, respectively. lThe voltage value and strain rate of change at each data point, 1( ) is an indicator function, when Returns 1 if the condition is met, otherwise returns 0. For a normal charging sequence, most of the d in the sequence... S Less than the optimal threshold T (V), MOI approaches 0; conversely, for battery sequences with micro-overcharge damage, MOI approaches 1.

[0037] After all batteries under test were capacitively charged, they were first discharged to 0% SOC at a 0.5C rate (1.95 A), then allowed to rest for 60 minutes before being charged at 0.5C to the upper limit voltage. The upper limit voltage included a normal upper limit voltage of 4.4 V and a slightly over-limit voltage of 4.5 V / 4.6 V. Afterward, they were charged at a constant voltage until the current was less than 0.05C. After resting for 60 minutes, they were discharged at 0.5C to the standard lower limit cutoff voltage of 2.8 V. This charge-discharge cycle was repeated. The test corresponding to each upper limit voltage was conducted independently. During the cycle, data such as voltage, current, temperature, and strain were simultaneously collected using a nebula charge-discharge device and a fiber optic demodulator. Specifically, battery #1 underwent 10 normal cycles followed by 136 cycles of slightly overcharge at a cutoff voltage of 4.5 V; battery #2 underwent 51 cycles of slightly overcharge at a cutoff voltage of 4.5 V; and battery #3 underwent 3 normal cycles followed by 1 cycle of slightly overcharge at a cutoff voltage of 4.6 V, and finally 3 more normal cycles.

[0038] like Figure 3 As shown in the IC curves for all cycles of battery #1, it can be seen that during the micro-overcharge cycle, the two IC peaks near 3.9 V and 4.0 V rapidly decrease to the lower right as the number of micro-overcharge cycles increases, exhibiting rapid aging characteristics. However, since normal aging can also cause similar changes, IC curve analysis alone is insufficient to effectively distinguish between batteries damaged by micro-overcharge and those aging normally.

[0039] Figure 4 The temperature and strain changes at various measurement points for all CC charging cycles of battery #1 are presented. Only the segments before 4.4 V are included for unified analysis. Simultaneously, the initial measurements of temperature and strain parameters at the start of CC charging are subtracted to ensure that the starting point of all sequences is located at the origin. It can be seen that the temperature changes at the measurement points during CC charging do not exhibit a clear pattern in all cycles before and after the battery undergoes micro-overcharging. Compared to temperature, the strain values ​​show a significant change before and after micro-overcharging. Specifically, in the first 11 normal cycles, the maximum strain at measurement point 2 is below 600 μST; however, after micro-overcharging, the maximum strain spikes to over 800 μST. During the voltage plateau period in the middle of battery charging, the strain change rates before and after micro-overcharging show a significant separation.

[0040] Figure 5The optimal threshold curve is shown to detect micro-overcharge during the 4.0 V-4.2 V voltage plateau period of batteries #1-3. (a) shows the strain rate of change curve for battery #1 in the characteristic voltage range, where the black dashed line represents the ADTCO dynamic threshold. It can be seen that in the first 11 normal charging cycles, all strain rates of change within the characteristic voltage range are below the threshold curve. That is, the MOI value remains constant at 0 in the first 11 cycles, indicating a safe cycle, as shown in (b). In the subsequent 135 damaged cycles, almost all strain rate of change values ​​are above the threshold curve, and the MOI value approaches 1, triggering an overcharge alarm. Notably, even in the first damaged cycle (the 12th cycle) with only one 4.5 V micro-overcharge, the MOI reaches as high as 0.934, demonstrating the sensitivity of the proposed method to micro-overcharge damage.

[0041] The strain rate distribution of battery #2 is as follows Figure 5 As shown in (c). Compared to battery #1, battery #2 exhibits stronger strain rate fluctuations under the same micro-overcharge conditions, which may be due to subtle differences in battery manufacturing or sensor installation. Fortunately, as Figure 5 As shown in (d), this strong strain variation does not affect the detection of micro-overcharge damage. In the first normal cycle, although the strain rate of change fluctuated abnormally high values ​​around 4.0 V and 4.05 V, the MOI for the entire characteristic voltage range was only 0.187, far below the warning threshold of 0.5. Meanwhile, in the second damage cycle, the MOI rapidly increased to 0.911, and remained at 1 for all subsequent damage cycles, indicating that the proposed method can achieve accurate detection of micro-overcharge.

[0042] Figure 5 (e) and Figure 5 (f) shows the strain rate curve and diagnostic results of battery #3 in the characteristic voltage range. Similar to the curves of batteries #1-2, all strain rates were below the threshold curve in the first 4 safety cycles, but after a 4.6 V micro-overcharge, all values ​​were above the threshold line.

[0043] The above embodiments demonstrate that the present invention can accurately identify battery micro-overcharge damage, maintain high diagnostic sensitivity, verify its applicability in battery micro-overcharge damage detection tasks, and has high engineering application value.

[0044] Example 2 A battery micro-overcharge damage detection system based on fiber optic sensing includes: The data acquisition module is used to acquire and decouple the battery surface temperature and strain parameters; The optimal dynamic threshold curve calculation module is used to extract the strain change rate of the characteristic voltage band as the detection feature based on the acquired data, train the adaptive dynamic detection threshold curve based on the B-spline function, optimize the shape of the threshold curve by minimizing the classification loss function, and solve for the optimal dynamic threshold curve. The micro-overcharge damage detection module is used to calculate the battery micro-overcharge index based on the optimal dynamic threshold curve, and to determine whether the battery micro-overcharge index exceeds the safety threshold. If it does, the battery micro-overcharge damage is determined.

[0045] The data acquisition module includes two fiber Bragg grating sensors with identical parameters, arranged in parallel. One of them is a bare optical fiber, and the other is a polyimide-sheathed protective optical fiber. Both fiber Bragg grating sensors are connected to an optical fiber demodulator.

[0046] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of one or more computer-usable storage media (including, but not limited to, disk storage, etc.) containing computer-usable program code. CD - ROM It takes the form of a computer program product implemented on (such as optical memory, etc.).

[0047] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0048] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0049] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art without creative effort within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting micro-overcharge damage in batteries based on fiber optic sensing, characterized in that, Includes the following steps: Acquire and decouple battery surface temperature and strain parameters; Based on the acquired data, the strain change rate of the characteristic voltage band is extracted as the detection feature. An adaptive dynamic detection threshold curve based on the B-spline function is trained. The shape of the threshold curve is optimized by minimizing the classification loss function, and the optimal dynamic threshold curve is obtained by solving the problem. Based on the optimal dynamic threshold curve, the battery micro-overcharge index is calculated to determine whether the battery micro-overcharge index exceeds the safety threshold. If it does, the battery micro-overcharge damage is determined. The process of optimizing the shape of the threshold curve by minimizing the classification loss function and obtaining the optimal dynamic threshold curve includes: the objective function of the optimization problem includes the classification loss term, the maximum violation penalty term, and the smoothness regularization term; The classification loss function is: ； in, J class ( θ ), J max ( θ )and J smooth ( θ These are the classification loss term, the maximum violation penalty term, and the smoothness regularization term, respectively. The classification loss term is: For each sequence i Its degree of violation is defined as: For positive samples with a label of 1, i.e., there is micro-overfilling impairment, ideally all points should be above the threshold curve: ； For negative samples with a label of 0, i.e., normal samples, ideally all points should be below the threshold curve: ； In the formula, and The first i The first sequence j The voltage value and the first j Each strain rate of change value, For voltage Threshold estimate of the rate of change of strain at a given point; sequence i The total loss is: ； In the formula N i Let i be the length of sequence i; The total classification loss item is: ； In the formula N The total number of training sequences; The maximum penalty for violation is: ； coefficient α Control the weighting percentage of this item; The smoothness regularization term is a second-order difference regularization term: ； In the formula, V k For the smoothness evaluation point, the second-order difference approximation is: ； in, λ represents the voltage step size between adjacent evaluation points, and λ is the weight of the smoothness regularization term.

2. The method for detecting micro-overcharge damage in batteries based on fiber optic sensing as described in claim 1, characterized in that, The process of acquiring and decoupling the surface temperature and strain parameters of the battery includes: measuring the parameters at the geometric center of the front of the pouch battery using dual fiber Bragg grating sensors. The two fiber Bragg grating sensors have the same parameters. The fiber Bragg grating sensors achieve selective reflection of a predetermined light wavelength through periodic modulation of the refractive index. The reflected wavelength is modulated by external physical quantities. The wavelength change of the reflected light is detected by a fiber demodulator, thereby decoupling the physical quantities.

3. The method for detecting micro-overcharge damage in batteries based on fiber optic sensing as described in claim 2, characterized in that, The optical fibers in the two fiber Bragg grating sensors are a bare optical fiber and a polyimide-sheathed optical fiber, respectively. The reflection wavelength of the bare optical fiber is affected by both temperature and strain, while the reflection wavelength of the sheathed optical fiber is controlled only by temperature.

4. The method for detecting micro-overcharge damage in batteries based on fiber optic sensing as described in claim 1, characterized in that, The process of training an adaptive dynamic detection threshold curve based on B-spline functions Includes: setting the threshold curve as T (V), where V represents the voltage independent variable, and the adaptive dynamic threshold curve optimization algorithm parameterizes the threshold curve through B-spline interpolation: ； ； ； ； In the formula, Let B-spline be the control point parameter vector. k Let M be the node vector, and M be the number of control points. B m (V; k ) is the first m There are B-spline basis functions, where V(max) and V(min) represent the maximum and minimum values ​​of the voltage range, respectively.

5. The method for detecting micro-overcharge damage in batteries based on fiber optic sensing as described in claim 1, characterized in that, In the process of obtaining the optimal dynamic threshold curve, θ The following physical constraints must be met: ； in, For the first i The first sequence j For each strain rate of change, the interior point method is used in the optimization solution. The constraint is incorporated into the objective function through the obstacle function, and the optimal solution is iteratively approximated along the central path.

6. The method for detecting micro-overcharge damage in batteries based on fiber optic sensing as described in claim 1, characterized in that, The process of calculating the battery micro-overcharge index based on the optimal dynamic threshold curve includes: The micro-overcharge index MOI is: ； n V is the length of the sequence. l and d S l These are the first and second digits of the sequence, respectively. l The voltage value and strain rate of change at each data point, 1( ) is an indicator function, when Returns 1 if the condition is met, otherwise returns 0.

7. A battery micro-overcharge damage detection system based on fiber optic sensing, characterized in that, include: The data acquisition module is used to acquire and decouple the battery surface temperature and strain parameters; The optimal dynamic threshold curve calculation module is used to extract the strain change rate of the characteristic voltage band as the detection feature based on the acquired data, train the adaptive dynamic detection threshold curve based on the B-spline function, optimize the shape of the threshold curve by minimizing the classification loss function, and solve for the optimal dynamic threshold curve. The micro-overcharge damage detection module is used to calculate the battery micro-overcharge index based on the optimal dynamic threshold curve, and to determine whether the battery micro-overcharge index exceeds the safety threshold. If it does, the battery micro-overcharge damage is determined. The process of optimizing the shape of the threshold curve by minimizing the classification loss function and obtaining the optimal dynamic threshold curve includes: the objective function of the optimization problem includes the classification loss term, the maximum violation penalty term, and the smoothness regularization term; The classification loss function is: ； in, J class ( θ ), J max ( θ )and J smooth ( θ These are the classification loss term, the maximum violation penalty term, and the smoothness regularization term, respectively. The classification loss term is: For each sequence i Its degree of violation is defined as: For positive samples with a label of 1, i.e., there is micro-overfilling impairment, ideally all points should be above the threshold curve: ； For negative samples with a label of 0, i.e., normal samples, ideally all points should be below the threshold curve: ； In the formula, and The first i The first sequence j The voltage value and the first j Each strain rate of change value, For voltage Threshold estimate of the rate of change of strain at a given point; sequence i The total loss is: ； In the formula N i Let i be the length of sequence i; The total classification loss item is: ； In the formula N The total number of training sequences; The maximum penalty for violation is: ； coefficient α Control the weighting percentage of this item; The smoothness regularization term is a second-order difference regularization term: ； In the formula, V k For the smoothness evaluation point, the second-order difference approximation is: ； in, λ represents the voltage step size between adjacent evaluation points, and λ is the weight of the smoothness regularization term.

8. The battery micro-overcharge damage detection system based on fiber optic sensing as described in claim 7, characterized in that, The data acquisition module includes two fiber Bragg grating sensors with identical parameters, arranged in parallel. One of them is a bare optical fiber, and the other is a polyimide-sheathed protective optical fiber. Both fiber Bragg grating sensors are connected to an optical fiber demodulator.

Images