Battery deformation real-time monitoring and early warning method based on multi-dimensional honeycomb structure

Abstract

The application discloses a battery deformation real-time monitoring and early warning method based on a multi-dimensional honeycomb structure and relates to the technical field of battery safety monitoring. The method comprises a monitoring target, a sensing network, a data acquisition terminal, a data processing terminal, a feature extraction terminal and a data analysis terminal. The monitoring target is a battery monomer or a battery module. The sensing network is composed of sensing units and temperature sensors interconnected in a honeycomb topology and is arranged on the surface of the battery or in the interval layer of the battery module. The wavelength offset and temperature data of the sensors are collected in real time through the data acquisition terminal. The real mechanical strain value is obtained through temperature compensation and calculation by the processing terminal, and the global deformation field is reconstructed. The feature extraction terminal extracts deformation characteristic parameters. The analysis terminal compares the deformation characteristic parameters with a machine learning early warning model according to a threshold value to make a judgment, and a graded early warning signal is sent when the conditions are met. The method realizes high-precision, global real-time monitoring and early safety warning of battery deformation.

Description

Technical Field

[0001] This invention belongs to the field of battery safety monitoring technology, and specifically discloses a method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure. Background Technology

[0002] Lithium-ion batteries are widely used in electric vehicles and energy storage power stations due to their high energy density and long cycle life. However, under abnormal conditions such as overcharging, over-discharging, and internal short circuits, batteries will produce gas and increase internal pressure, leading to bulging and deformation. Deformation is one of the most direct and critical physical warning signals before battery thermal runaway. Therefore, real-time, accurate, and comprehensive monitoring of battery deformation is of great significance for early warning of battery thermal runaway and ensuring system safety.

[0003] Existing battery deformation monitoring technologies are mostly single-point or local monitoring, such as placing one or a few sensors on the battery surface. This method has the following shortcomings: First, it cannot capture the deformation distribution over the entire battery area, and may miss local hot spots or abnormal deformation areas; second, it is difficult to distinguish between thermal expansion caused by temperature changes and mechanical deformation caused by internal faults, leading to false alarms or missed alarms; third, it lacks in-depth analysis of deformation patterns, such as distribution symmetry and gradient, resulting in limited early warning capabilities.

[0004] Therefore, it is necessary to invent a method for real-time monitoring and early warning of battery deformation based on a multidimensional honeycomb structure to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies in the prior art, this invention provides a real-time monitoring and early warning method for battery deformation based on a multi-dimensional cellular structure. The method involves a data acquisition terminal that collects sensor wavelength offset and temperature data in real time; a processing terminal that obtains the actual mechanical strain value through temperature compensation and calculation, and reconstructs the global deformation field; a feature extraction terminal that extracts deformation feature parameters; and an analysis terminal that combines threshold comparison of the deformation feature parameters with a machine learning early warning model to make judgments, and issuing graded early warning signals when conditions are met. This effectively solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure, specifically including the following steps:

[0007] S1. Arrange the sensor network on the surface of the battery cell or inside the battery module;

[0008] S2. The data acquisition terminal collects measurement data in real time;

[0009] S3. The data processing terminal processes and calculates the collected measurement data to obtain real-time deformation field distribution information of the entire battery domain.

[0010] S4. The feature extraction terminal extracts deformation feature parameters based on real-time deformation field distribution information;

[0011] S5. The data analysis terminal combines deformation characteristic parameters and early warning models to make judgments. When the judgment results meet the early warning conditions, an early warning signal is generated and issued.

[0012] The specific method of arranging the sensor network on the single battery cell is as follows: the sensor network is printed or embedded on a flexible insulating substrate to form an integrated sensing film, and then the sensing film is attached to the surface of the battery cell.

[0013] The specific way the sensor network is arranged in the battery module is as follows: the sensor network is woven or embedded in the spacer layer between the cells in the battery module, so that the sensor network is in contact with the surface of multiple battery cells.

[0014] The specific analysis method of the feature extraction terminal is as follows: extract deformation feature parameters from real-time deformation field distribution information; the deformation feature parameters include: maximum deformation, average deformation, deformation gradient, deformation distribution symmetry, and deformation rate.

[0015] Preferably, the sensing unit is a fiber Bragg grating sensor; the cellular topology is implemented by connecting multiple fiber Bragg grating sensors in series on a single optical fiber through wavelength division multiplexing or by implementing multiple fiber Bragg grating sensors on multiple optical fibers through space division multiplexing technology.

[0016] Preferably, the measurement data includes the offset of the center wavelength of the fiber Bragg grating sensor and temperature sensor data.

[0017] Preferably, the data processing and calculation includes:

[0018] Based on the mapping relationship between wavelength offset and strain and temperature, the initial strain value of strain-temperature coupling is calculated;

[0019] By using temperature sensor data, the initial strain value is decoupled by temperature compensation to obtain the true mechanical strain value;

[0020] Each calculated real mechanical strain value is reconstructed into a continuous deformation field covering the entire monitoring target area through an interpolation algorithm.

[0021] Preferably, the specific analysis method of the data analysis terminal is as follows:

[0022] The maximum deformation is compared with a preset absolute safety threshold; if it is exceeded, an alarm is triggered immediately.

[0023] Based on a pre-built early warning model, the time series data of deformation characteristic parameters are dynamically analyzed to identify abnormal deformation and issue an early warning when the deformation trend deviates from the normal baseline.

[0024] By combining the deformation state classification, deformation characteristic parameters, and the location information corresponding to the deformation characteristic parameters, a diagnostic early warning signal containing the fault location and type is generated.

[0025] The technical effects and advantages of this invention are as follows:

[0026] 1. Real-time high-precision deformation monitoring of the entire battery domain: By adopting a sensor network based on a multi-dimensional cellular topology, it is possible to densely deploy points with high spatial resolution to achieve real-time, continuous, and dynamic monitoring of the deformation field of the entire battery cell surface or the battery module, overcoming the limitation of traditional point measurement that cannot fully reflect the deformation state of the battery.

[0027] 2. Excellent anti-interference and temperature compensation capabilities: The sensor network integrates a dedicated temperature sensor and adopts fiber optic grating sensing technology. Through the data processing terminal, the coupling relationship between strain and temperature is calculated, which effectively eliminates the interference of temperature changes on deformation measurement results and significantly improves the accuracy and reliability of mechanical strain measurement.

[0028] 3. Flexible deployment and strong applicability: The sensor network can be made into a flexible film and attached to the surface of the cell, or woven and embedded in the spacer layer of the battery module, and also has electrical insulation. This integrated design enables it to adapt to various complex battery structures and small spaces, achieve integrated deployment without affecting the performance of the battery pack, and can simultaneously monitor the deformation and interaction forces of multiple cells in the battery module.

[0029] 4. Provides multi-level intelligent early warning and fault diagnosis: The system not only sets absolute safety thresholds for immediate alarms, but more importantly, it uses feature extraction and machine learning-based early warning models to perform dynamic time-series analysis of deformation characteristic parameters. This enables early identification of abnormal deformation trends and proactive safety warnings. Simultaneously, combined with the precise positioning capabilities of cellular units, it can accurately determine the location and type of abnormal deformation, generating diagnostic early warning signals with guiding significance. This provides accurate and efficient decision support for the safety management and maintenance of the battery system. Attached Figure Description

[0030] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0031] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0032] Figure 2 This is a schematic diagram of the steps of the present invention. Detailed Implementation

[0033] 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.

[0034] This invention provides, for example Figure 1 The method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure, as shown, includes: a monitoring target, a sensor network, a data acquisition terminal, a data processing terminal, a feature extraction terminal, and a data analysis terminal. The monitoring target is a single battery cell or a battery module, such as... Figure 2 As shown, the specific steps include the following:

[0035] S1. Arrange the sensor network on the surface of the battery cell or inside the battery module;

[0036] Furthermore, in the above technical solution, the sensor network consists of multiple sensing units and temperature sensors interconnected in a cellular topology.

[0037] The sensing unit is a fiber optic grating (FBG) sensor, which serves as the core sensing unit for measuring strain.

[0038] The temperature sensor can be an integrated dedicated FBG temperature sensor, a thermocouple, or a resistance temperature detector (RTD) for synchronous temperature monitoring and temperature compensation.

[0039] The cellular topology is implemented by connecting multiple fiber Bragg grating sensors in series on a single optical fiber through wavelength division multiplexing, or by implementing multiple fiber Bragg grating sensors on multiple optical fibers through space division multiplexing technology.

[0040] It should be further explained that the cellular topology adopts a hexagonal cellular grid layout, with each fiber grating sensor located at the apex of the cell. Each fiber grating sensor has a different center wavelength and is connected by optical fibers to form a mesh structure. Temperature sensors are arranged at key nodes or the center of the cellular grid to ensure the representativeness of the temperature field monitoring.

[0041] Furthermore, in the above technical solution, the sensor network is arranged on the single battery cell in the following specific way: the sensor network is printed or embedded on a flexible insulating substrate to form an integrated sensing film, and then the sensing film is attached to the surface of the battery cell.

[0042] In a preferred embodiment of the present invention, the specific implementation of the sensor network being arranged in a single battery cell can be as follows: a flexible insulating material, specifically polyimide or PET film, is selected as the carrier of the sensor network; the sensor is patterned and printed on a flexible substrate using photolithography, screen printing or laser etching technology, or the sensor network is embedded between two layers of flexible insulating film, and then encapsulated into an integrated sensor film by hot pressing or adhesive bonding, and the sensor film is attached to the surface of the battery cell to ensure close contact with the battery surface in order to accurately sense deformation and temperature changes;

[0043] Furthermore, in the above technical solution, the sensor network is arranged in the battery module in the following specific way: the sensor network is woven or embedded in the spacer layer between the cells in the battery module, so that the sensor network is in contact with the surface of multiple battery cells, and the deformation and interaction force of multiple battery cells in the battery module are monitored synchronously.

[0044] In a preferred embodiment of the present invention, the sensor network is arranged in the battery module as follows: a flexible insulating material, specifically a polyimide or silicone-based composite material, is selected as the substrate of the sensor network, and the shape and size of the substrate are completely consistent with the shape and size of the spacer layer in the battery module; the sensor network is fabricated on the selected flexible substrate using precision printing, photolithography, or flexible circuit board embedding technology, and covered with a thin flexible protective layer, such as another layer of polyimide or silicone-based composite material, to form a new spacer layer in the battery module, and the new spacer layer replaces the original spacer layer in the battery module.

[0045] S2. The data acquisition terminal collects measurement data in real time;

[0046] Furthermore, in the above technical solution, the measurement data includes the offset of the center wavelength of the fiber optic grating sensor and temperature sensor data.

[0047] In a preferred embodiment of the present invention, the specific implementation of the data acquisition terminal for real-time acquisition of measurement data is as follows: a multi-channel fiber optic demodulator is used as the core device of the data acquisition terminal, which supports high-speed and high-precision wavelength demodulation. Each channel corresponds to one or more fiber optic grating sensors connected in series or in parallel through wavelength division multiplexing or space division multiplexing technology. The temperature sensor data is acquired synchronously through an integrated temperature acquisition module, such as a thermocouple or an RTD acquisition card.

[0048] S3. The data processing terminal processes and calculates the collected measurement data to obtain real-time deformation field distribution information of the entire battery domain.

[0049] Furthermore, in the above technical solution, the data processing and calculation include:

[0050] Based on the mapping relationship between wavelength offset and strain and temperature, the initial strain value of strain-temperature coupling is calculated;

[0051] It should be further explained that, based on the physical characteristics of the fiber Bragg grating sensor, a mapping model is established between the wavelength offset Δλ and the strain ε and temperature T. In the formula, K ε K is the strain sensitivity coefficient. T Here, ΔT is the temperature sensitivity coefficient, and ΔT is the deviation between the temperature measured by the temperature sensor and the reference temperature.

[0052] Example of temperature compensation calculation:

[0053] Strain sensitivity coefficient: K ε =1.2pm / με, temperature sensitivity coefficient: K T =10.0 pm / ℃, at a certain moment, the data acquisition terminal obtained a wavelength offset Δλ=120.0 pm; the temperature sensor measured a temperature deviation ΔT=5.0℃ from the reference temperature;

[0054] Calculation process:

[0055] K ε ·ε+K T ΔT = 120.0;

[0056] ε=(120-K T ·ΔT) / K ε =(120-10×5) / 1.2=70 / 1.2≈58.33με

[0057] Furthermore, the temperature sensitivity coefficient K T The method for determining it is as follows:

[0058] Place the FBG sensor to be calibrated in a high-precision constant temperature chamber, and change the temperature of the constant temperature chamber in a certain temperature step, such as 5℃, to cover the entire temperature range of the battery operation. At each temperature stabilization point, use a demodulator to accurately record the center wavelength value λ of the FBG sensor.

[0059] Calculate the wavelength offset Δλ and temperature change ΔT for each temperature point relative to a reference temperature, such as 25℃.

[0060] Plot a scatter plot of Δλ / λ0 versus ΔT, perform linear fitting on the data points, and the slope of the resulting straight line is the temperature sensitivity coefficient K of the FBG sensor. T

[0061] Furthermore, the strain sensitivity coefficient K ε The method for determining it is as follows:

[0062] The FBG sensor to be calibrated is tightly bonded to a beam or tensile specimen of known material properties, such as aluminum alloy, which must be placed in a constant temperature environment. The bonding process is consistent with the process used on the battery.

[0063] An axial strain ε of known magnitude is applied to the specimen using a precision tensile testing machine, while a demodulator is used to record the change Δλ of the center wavelength of the FBG sensor.

[0064] Plot a scatter plot of Δλ / λ0 versus the applied strain ε, perform linear fitting on the data points, and the slope of the resulting straight line is the strain sensitivity coefficient K of the FBG sensor. ε .

[0065] By using temperature sensor data, temperature compensation decoupling is performed on the initial strain value to eliminate the influence of temperature changes on the deformation measurement results and obtain the true mechanical strain value.

[0066] Each calculated real mechanical strain value is reconstructed into a continuous deformation field covering the entire monitoring target area through an interpolation algorithm.

[0067] It should be further explained that after calculating the true mechanical strain value of each sensing unit, the true mechanical strain value is reconstructed into a continuous deformation field covering the entire domain of the battery cell or battery module through the interpolation algorithm, specifically such as triangulation interpolation, inverse distance weighting, Kriging interpolation or spline interpolation, forming a visualized strain distribution map.

[0068] It should be further explained that, taking the triangulation difference as an example, the specific method for reconstructing the real mechanical strain value into a continuous deformation field is as follows:

[0069] Triangular mesh generation: Using fiber optic grating sensors as data points, Delaunay triangulation is performed to generate a triangular mesh covering the entire battery surface. The coordinates of the vertices of each triangle in the triangular mesh and the actual mechanical strain value after temperature compensation are obtained.

[0070] Define the output grid: Define a sufficiently dense regular grid that covers all sensor points, for example, one point per 1 mm;

[0071] Interpolating the points on the output mesh yields the strain field: For any point P, interpolation can be performed based on the coordinates of the three vertices of the triangle containing P and the actual mechanical strain value after temperature compensation. The specific interpolation process is as follows:

[0072] Suppose point P is inside triangle T, and its coordinates are A(x). P y P , ε P The vertices of triangle T are A(x) and A(x). a ya , ε a ), B(x) b y b , ε b ), C(x) c y c , ε c x and y represent the coordinates of the corresponding point in the triangular mesh, and ε is the actual mechanical strain value of the corresponding point;

[0073] Calculate the centroid coordinates (λ) of point P within triangle T. a , λ b , λ c The calculation formula is: In the formula, Area(U, V, W) represents the area of ​​the triangle formed by points U, V, and W;

[0074] Calculate the strain value ε at point P P The calculation formula is: ;

[0075] Convert the strain value into deformation: For each point on the output grid, its deformation δ can be calculated by the formula δ=ε⋅L0, where ε is the actual mechanical strain value at that point and L0 is the effective base length of the sensing unit;

[0076] The value of L0 depends on the type and deployment method of the sensor, and there are mainly two cases:

[0077] For a single FBG sensor based on wavelength division multiplexing, L0 is the physical length of the grating itself;

[0078] When multiple FBGs are connected in series on a single optical fiber to form a sensor network, the L0 of each sensor can be understood as the distance between two adjacent sensor nodes, i.e., the side length of the cellular grid.

[0079] It should be further explained that, in the case where the sensor network is arranged in the battery module, the calculated actual mechanical strain value should be the sum of the strain values ​​of multiple surfaces in contact with the sensor network, and the calculation result reflects the relative deformation at a certain point.

[0080] Example of deformation calculation:

[0081] Vertices of triangle T:

[0082] A: Coordinates (x) a y a )=(0,0), the true mechanical strain value ε a =0.01;

[0083] B: Coordinates (x) b y b)=(4,0), the actual mechanical strain value ε b =0.02;

[0084] C: coordinates (x) c y c )=(0,3), the true mechanical strain value ε c =0.03;

[0085] Point P: coordinates (x) p y p )=(1,1), effective base length L0=100mm;

[0086] Calculation steps:

[0087] Calculate the area of ​​triangle ABC:

[0088] Area(A, B, C)=(1 / 2)×|(4−0)×(3−0)−(0−0)×(0−0)|=(1 / 2)×|12−0|=6;

[0089] Area(P, B, C)=(1 / 2)×|(4−1)×(3−1)−(0−1)×(0−1)|=(1 / 2)×|6−(-1)|=2.5;

[0090] Area(A, P, C)=(1 / 2)×|(1−0)×(3−0)−(0−0)×(0−1)|= (1 / 2)×|3−0|=1.5;

[0091] Area(A,B,P)=(1 / 2)×|(4−0)×(1−0)−(1−0)×(0−0)|=(1 / 2)×|4−0|=2;

[0092] Calculate the centroid coordinates of point P:

[0093] λ a =2.5 / 6=5 / 12≈0.4167;

[0094] λ b =1.5 / 6=1 / 4≈0.25;

[0095] λ c =2 / 6=1 / 3≈0.3333;

[0096] Interpolation calculation of strain value ε at point P p :

[0097] ε p=(5 / 12)×0.01+(1 / 4)×0.02+(1 / 3)×0.03=(0.05 / 12)+(0.06 / 12)+(0.12 / 12)=(0.23 / 12)≈0.019167;

[0098] Convert the strain value into deformation δ:

[0099] δ=ε p ⋅L0=0.019167×100=1.9167mm;

[0100] Conclusion: The deformation of point P is calculated to be 1.9167 mm.

[0101] S4. The feature extraction terminal extracts deformation feature parameters based on real-time deformation field distribution information;

[0102] Furthermore, in the above technical solution, the specific analysis method of the feature extraction terminal is as follows: extracting deformation feature parameters from real-time deformation field distribution information; the deformation feature parameters include: maximum deformation, average deformation, deformation gradient, deformation distribution symmetry, and deformation rate.

[0103] It should be further explained that the process for extracting deformation feature parameters is as follows:

[0104] Calculate the maximum deformation δ max : Iterate through the deformation δ of all points in the entire deformation field; find the maximum value among them, and denote it as δ. max This is used for subsequent comparisons of absolute safety thresholds;

[0105] Calculate the average deformation δ avg The arithmetic mean of the deformation δ in all deformation fields is used to reflect the overall deformation level. The formula is as follows: In the formula, N is the number of points in the output mesh divided in the real-time deformation field;

[0106] Calculating the deformation gradient ∇δ: Numerical differentiation methods, such as the central difference method, are used to calculate the gradient at each point in the deformation field; formulas can be used... and Calculate the gradient components in the x and y directions respectively, and then apply the formula... The overall gradient is synthesized, and δ in the formula is... x+Δx，y δ x-Δx，y δ x，y+Δy and δ x，y-Δy The value represents the deformation of the point at the corresponding coordinates. Δx and Δy represent the distance between two adjacent points in the output mesh of the real-time deformation field in the corresponding direction. The gradient value reflects the severity of the deformation change and can be used to identify local stress concentration areas.

[0107] It should be further explained that, for points on the boundary, the gradient components in the x and y directions are calculated using the forward or backward difference method.

[0108] Example of deformation gradient calculation:

[0109] A 3x3 grid with a grid spacing of Δx = Δy = 1 mm has the following deformation data:

[0110] δ 0，0 =10mm, δ 0，1 =12mm, δ 0，2 =15mm, δ 1，0 =11mm, δ 1，1 =20mm, δ 1，2 =18mm, δ 2，0 =13mm, δ 2，1 =17mm, δ 2，2 =16mm;

[0111] Calculate the gradient at the center point (1, 1):

[0112] ∂δ / ∂x≈(δ 1+1，1 -δ 1-1，1 ) / (2×1)=(δ 2，1 -δ 0，1 ) / 2 = 2.5 μm / mm;

[0113] ∂δ / ∂y≈(δ 1，1+1 -δ 1，1-1 ) / (2×1)=(δ 1，2 -δ 1，0 ) / 2 = 3.5 μm / mm;

[0114] Deformation gradient ∇δ=(2.5) 2 +3.5 2 ) 1 / 2 ≈4.3μm / mm;

[0115] Conclusion: The deformation gradient ∇δ at coordinates (1, 1) is 4.3 μm / mm.

[0116] Calculate the deformation distribution symmetry S: Mirror the deformation field along the battery's central axis or a preset symmetry axis; calculate the difference in deformation within the symmetrical region; use the correlation coefficient or root mean square error as a symmetry index; for example, using the formula... Calculate the symmetry S of the deformation distribution; the closer S is to 1, the better the symmetry; a lower S indicates asymmetrical deformation distribution, which may indicate local anomalies. In the formula, RMSE() is the root mean square error formula, and δ left and δ right The deformation of points in the output mesh of a real-time deformation field that is symmetric about the axis of symmetry;

[0117] Calculate the deformation rate dδ / dt: Combine historical deformation field data with the formula Calculates the rate of change of the current deformation relative to the previous time step; rate calculations can be performed on the entire domain or key regions, such as the point of maximum deformation; used to determine whether the deformation development trend is abnormal, δ in the formula t Δt represents the deformation at time t of a point in the output grid divided in the real-time deformation field, where Δt is a short time interval, such as 1 second.

[0118] S5. The data analysis terminal combines deformation characteristic parameters and early warning models to make judgments. When the judgment results meet the early warning conditions, an early warning signal is generated and issued.

[0119] Furthermore, in the above technical solution, the specific analysis method of the data analysis terminal is as follows:

[0120] The maximum deformation is compared with a preset absolute safety threshold; if it is exceeded, an alarm is triggered immediately.

[0121] Based on a pre-built early warning model, the time series data of deformation characteristic parameters are dynamically analyzed to identify abnormal deformation and issue an early warning when the deformation trend deviates from the normal baseline.

[0122] It should be further explained that the early warning model uses historical deformation data, including time series of deformation feature parameters under normal and abnormal states, to train a machine learning model, such as LSTM, GRU, CNN, or ensemble learning algorithms. The model input is the time series of deformation feature parameters, and the output is the classification of deformation state, such as normal, slightly abnormal, or severely abnormal.

[0123] By combining the deformation state classification, deformation characteristic parameters, and the location information corresponding to the deformation characteristic parameters, a diagnostic early warning signal containing the fault location and type is generated.

[0124] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure, characterized in that, include: The system comprises a monitoring target, a sensor network, a data acquisition terminal, a data processing terminal, a feature extraction terminal, and a data analysis terminal. The monitoring target is a single battery cell or a battery module. The sensor network consists of multiple interconnected sensing units and temperature sensors in a cellular topology. The specific steps include: S1. Arrange the sensor network on the surface of the battery cell or inside the battery module; S2. The data acquisition terminal collects measurement data in real time; S3. The data processing terminal processes and calculates the collected measurement data to obtain real-time deformation field distribution information of the entire battery domain. S4. The feature extraction terminal extracts deformation feature parameters based on real-time deformation field distribution information; S5. The data analysis terminal combines deformation characteristic parameters and early warning models to make judgments. When the judgment results meet the early warning conditions, an early warning signal is generated and issued. The specific method of arranging the sensor network on the single battery cell is as follows: the sensor network is printed or embedded on a flexible insulating substrate to form an integrated sensing film, and then the sensing film is attached to the surface of the battery cell. The specific way the sensor network is arranged in the battery module is as follows: the sensor network is woven or embedded in the spacer layer between the cells in the battery module, so that the sensor network is in contact with the surface of multiple battery cells. The specific analysis method of the feature extraction terminal is as follows: extract deformation feature parameters from real-time deformation field distribution information; the deformation feature parameters include: maximum deformation, average deformation, deformation gradient, deformation distribution symmetry, and deformation rate.

2. The method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure as described in claim 1, characterized in that: The sensing unit is a fiber Bragg grating sensor; the cellular topology is implemented by connecting multiple fiber Bragg grating sensors in series on a single optical fiber through wavelength division multiplexing or by implementing multiple fiber Bragg grating sensors on multiple optical fibers through space division multiplexing technology.

3. The method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure as described in claim 1, characterized in that: The measurement data includes the offset of the center wavelength of the fiber optic grating sensor and temperature sensor data.

4. The method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure as described in claim 1, characterized in that: The data processing and calculation include: Based on the mapping relationship between wavelength offset and strain and temperature, the initial strain value of strain-temperature coupling is calculated; By using temperature sensor data, the initial strain value is decoupled by temperature compensation to obtain the true mechanical strain value; Each calculated real mechanical strain value is reconstructed into a continuous deformation field covering the entire monitoring target area through an interpolation algorithm.

5. The method for real-time monitoring and early warning of battery deformation based on a multi-dimensional honeycomb structure as described in claim 1, characterized in that: The specific analysis method of the data analysis terminal is as follows: The maximum deformation is compared with a preset absolute safety threshold; if it is exceeded, an alarm is triggered immediately. Based on a pre-built early warning model, the time series data of deformation characteristic parameters are dynamically analyzed to identify abnormal deformation and issue an early warning when the deformation trend deviates from the normal baseline. By combining the deformation state classification, deformation characteristic parameters, and the location information corresponding to the deformation characteristic parameters, a diagnostic early warning signal containing the fault location and type is generated.

Images