An in-situ spatial and temporal thermal monitoring system for lithium metal batteries based on distributed optical fiber

Abstract

A system and method for in-situ spatiotemporal thermal monitoring of lithium metal batteries based on distributed optical fiber (OFDR) utilizes distributed optical fiber sensing technology combined with a helical arrangement of single-mode fiber (SMF) to convert one-dimensional temperature data into a two-dimensional planar distribution, enabling precise monitoring of the internal temperature field of the battery. Furthermore, by introducing a super-resolution algorithm based on the SRGAN model, low-resolution temperature images are converted into high-resolution images, reconstructing details and improving both spatial and temporal resolution. This integrated approach of physical sensing and super-resolution algorithms reduces the risk of thermal runaway, extends battery life, achieves more uniform lithium deposition, and improves battery cycle life, thus providing strong technical support for the safe operation and performance optimization of lithium metal batteries. In addition, it allows for the investigation of hotspot generation mechanisms and verification of the effectiveness of battery protection strategies. This invention provides real-time, non-destructive, and high-precision super-resolution monitoring of large measurement areas for lithium metal batteries.

Description

Technical Field

[0001] This invention relates to battery thermal monitoring technology, and in particular to an in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers. Background Technology

[0002] As the global transition to clean energy accelerates, the demand for innovative energy storage technologies continues to grow. Lithium metal electrodes are emerging as a transformative solution, with theoretical capacities an order of magnitude higher than traditional lithium-ion batteries. This significant advancement makes lithium metal batteries (LMBs) central to achieving the high energy densities required for future technologies. However, the practical application of these batteries is hampered by critical safety concerns, as their performance degradation and safety exhibit strong temperature dependence, particularly the risk of thermal runaway, which could lead to severe fires and explosions. Therefore, rigorous thermal monitoring throughout the entire lifecycle of LMBs is crucial. By mitigating safety risks while leveraging the high capacity of lithium metal, these efforts will drive the development of next-generation high-energy-density, resource-efficient battery systems, thereby contributing to the clean energy transition.

[0003] With advancements in battery mechanism research and thermal management technology, researchers have gradually recognized that the internal temperature of a battery is the most critical parameter determining thermal runaway. Current technologies for monitoring the internal temperature of lithium metal batteries include electrochemical impedance spectroscopy (EIS), ultrasonic imaging, fiber Bragg grating, and optical frequency domain reflectometry.

[0004] Electrochemical impedance spectroscopy (EIS) is a quasi-steady-state frequency domain measurement method. By measuring electrical parameters such as current, potential, and resistance, it obtains the EIS spectrum of the battery. The macroscopic battery temperature can be monitored using the EIS curve. This method is real-time, and the measurement results are easy to process. Its disadvantages include being an external measurement method, only able to measure the overall battery temperature, unable to detect local hot spots, and being a single-point detection method with a small detection area.

[0005] Ultrasonic imaging refers to the use of ultrasonic technology to monitor batteries. By exciting and receiving ultrasonic signals on the battery surface, and utilizing the changes in signal characteristics caused by variations in battery performance and structural differences during the propagation of ultrasonic guided waves inside the battery, the relationship between ultrasonic transmission parameters and battery state can be established, thereby obtaining information about the battery's internal temperature. This method is highly sensitive, low-cost, easy to use, and fast. However, it relies on fixed-point monitoring of the battery's exterior and cannot detect the overall heterogeneity of commercially available pouch batteries.

[0006] The fiber Bragg grating method is fabricated by laterally exposing the fiber core to strong ultraviolet light with a periodic pattern. The changes in the central wavelength of the Bragg reflected light are used to determine the changes in corresponding physical quantities within the cell. This method offers high resolution and can achieve in-situ cell temperature monitoring. While this method allows for location-based monitoring, temperature monitoring can only be performed in specially treated areas of the fiber, and distributed monitoring is not possible.

[0007] Optical frequency domain reflectometry (OFDR) utilizes backscattering of Rayleigh scattering caused by microscopic inhomogeneities in the refractive index of optical fibers. The system consists of a linearly tunable laser, an interferometer structure (reference arm and measuring arm), a photodetector, and a data acquisition unit. By analyzing the frequencies in the spectrum, the location of scattering points in the sensing fiber can be determined, thus achieving fiber localization. The wavelength shift generated by the Rayleigh scattering mode can be used to obtain the change in the corresponding physical quantity at that location. This method has high sensitivity and is distributed, but the number of detectable points remains limited.

[0008] Fiber Bragg grating and optical frequency domain reflectometry offer advantages over electrochemical impedance spectroscopy and ultrasonic imaging, including in-situ, non-destructive, multi-point, and distributed monitoring. However, current monitoring technologies still present a limited number of detectable points relative to the electrode surface area, resulting in one-dimensional temperature data. Therefore, achieving precise operation and distributed temperature monitoring of the entire lithium metal electrode throughout the battery's lifespan remains a challenge.

[0009] Missing data can be filled in using regression analysis, interpolation methods, and Bayesian inference. However, these traditional interpolation methods are prone to loss of detail or blurring. In recent years, machine learning-based super-resolution techniques have been applied to convert low-resolution images into high-resolution images while reconstructing details.

[0010] It should be noted that the information disclosed in the background section above is only for understanding the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0011] The main objective of this invention is to overcome the deficiencies in the aforementioned background technology and provide an in-situ spatiotemporal thermal monitoring system and method for lithium metal batteries based on distributed optical fibers.

[0012] To achieve the above objectives, the present invention adopts the following technical solution:

[0013] A distributed optical fiber-based in-situ spatiotemporal thermal monitoring system for lithium metal batteries includes:

[0014] A single-mode fiber (SMF) is connected to the inside of the battery. The SMF has a helical arrangement structure, which makes the fiber distributed along a two-dimensional plane, and is used to realize temperature monitoring in physical space.

[0015] An OFDR system for capturing temperature data of a one-dimensional continuous point set along the single-mode fiber (SMF), the OFDR system comprising:

[0016] A linearly tunable laser is used to emit optical signals;

[0017] The interferometer structure includes a reference arm and a measuring arm, used to measure reflected light signals propagating along an optical fiber;

[0018] Photodetectors and data acquisition instruments are used to collect optical signal data and convert it into electrical signal data.

[0019] The data processing unit is used to convert the one-dimensional temperature data captured by the OFDR system into two-dimensional planar distribution data corresponding to the helical arrangement structure of the single-mode fiber (SMF) to achieve accurate monitoring of the internal temperature distribution of the battery.

[0020] Furthermore, the data processing module includes a super-resolution algorithm module for performing super-resolution processing of temperature data to improve the spatial resolution of temperature monitoring. The super-resolution algorithm module includes:

[0021] The SRGAN model is used for machine learning to achieve super-resolution of temperature mapping. The SRGAN model includes:

[0022] The generator network, preferably using ResNet blocks, is used to generate super-resolution images;

[0023] Discriminator network, a CNN structure, is used to evaluate high-resolution and super-resolution images.

[0024] Furthermore, the generator network comprises a deep network consisting of five ResNet blocks, each ResNet block enhancing the information flow through skip connections, wherein each ResNet block contains 64 convolutional kernels of size 9×9 with a stride of 1 and padding of 4; the discriminator network is a convolutional neural network (CNN) using the Leaky ReLU activation function; the output layer contains a dense block followed by a sigmoid function for binary scoring of high-resolution and super-resolution images to evaluate image quality.

[0025] Furthermore, the system includes an external temperature experimental setup for constructing a training dataset, the setup comprising:

[0026] Multiple ceramic heating elements of different shapes are connected to a single-mode fiber (SMF) arranged in a spiral to simulate the temperature distribution inside a battery.

[0027] An infrared camera is used to collect temperature data of the ceramic heating element to form a dataset for training a super-resolution algorithm model; wherein the temperature of the ceramic heating element is controlled by a power supply, and the dataset is used for machine learning to achieve super-resolution temperature mapping.

[0028] Furthermore, the temperature change video captured by the infrared camera is used to record the temperature distribution on the battery surface; frames are extracted from the video as images and synchronized with the data acquisition time of the OFDR system to form a raw dataset; wherein, the raw dataset is processed by data augmentation techniques, including random rotation, reflection, translation, scaling, brightness and contrast adjustment, Gaussian blur, noise addition, and cropping.

[0029] Furthermore, it also includes:

[0030] A battery testing system is used to perform capacity and voltage readouts of batteries to monitor their electrochemical performance.

[0031] The computer, including the data processing unit, is used to collect and process data from the battery testing system and the infrared camera, as well as to manage the data acquisition and analysis of the OFDR system;

[0032] Scanning electron microscopy (SEM) is used to observe the morphology of the anode after battery cycling in order to assess the physical state and structural changes of the battery.

[0033] The computer synchronizes data from the OFDR system, infrared camera, and battery testing system, and performs data analysis and image processing.

[0034] Furthermore, the data processing unit identifies areas without temperature readings and fills the data in these areas with a preset room temperature value, such as 22°C.

[0035] Furthermore, a lithium anode protection strategy is applied to the monitored lithium metal battery, the lithium anode protection strategy including:

[0036] Pyramid-patterned lithium anodes increase surface area and improve ion transport pathways by forming pyramid structures on the surface of lithium anodes.

[0037] Copper mesh patterned lithium anodes utilize micropatterns created on the surface of lithium anodes to guide the uniform distribution and deposition of lithium ions.

[0038] PLA-coated lithium anodes enhance thermal stability by applying a protective coating to the surface of the lithium anode, taking advantage of its chemical and mechanical barrier properties.

[0039] Monitoring was conducted using the OST-SRTM system to evaluate the effectiveness of different lithium anode protection strategies.

[0040] Furthermore, the pyramid-patterned lithium anode uses a pyramid-shaped mold made with high-precision 3D printing technology to imprint a microstructure pattern on the lithium anode. The mold size is 50mm×50mm×0.5mm, the height of a single pyramid is 50μm, the width is 50μm, and the spacing is 40μm. A pressure of 2MPa is applied during the imprinting process.

[0041] The copper mesh patterned lithium anode uses a 250-mesh copper mesh to apply a pressure of 1 MPa to the lithium surface and performs two imprinting processes to improve the uniformity of surface modification. The first imprinting is done in the direct direction, and the second imprinting is done after rotating the mesh by 45 degrees.

[0042] The PLA-coated lithium anode is prepared by dissolving PLA in DMSO solution, stirring at 130°C for 5 hours and continuously stirring at 80°C for 24 hours to obtain a uniform milky white solution, preparing a PLA solution with a mass ratio of 1.2%, and then scraping the solution onto the lithium metal surface to form a PLA coating.

[0043] A method for in-situ spatiotemporal thermal monitoring of lithium metal batteries, using the aforementioned distributed optical fiber-based in-situ spatiotemporal thermal monitoring system for lithium metal batteries.

[0044] The present invention has the following beneficial effects:

[0045] This invention provides an in-situ spatiotemporal thermal monitoring system and method for lithium metal batteries based on distributed optical fiber. The innovative monitoring system provides real-time, non-destructive, and high-precision temperature monitoring for lithium metal batteries, covering a large measurement area and possessing super-resolution capabilities. The system utilizes the high sensitivity and distributed characteristics of distributed optical fiber sensing (OFDR) technology, combined with the design of a helical arrangement of single-mode fiber (SMF), to convert one-dimensional temperature data into a two-dimensional planar distribution, thereby achieving precise monitoring of the internal temperature field of the battery. Furthermore, by introducing a super-resolution algorithm based on the SRGAN model, the system can convert low-resolution temperature images into high-resolution images, reconstructing details and improving spatial and temporal resolution. This integrated physical sensing and super-resolution algorithm method not only explores the mechanism of hotspot generation and verifies the effectiveness of battery protection strategies, but also reduces the risk of thermal runaway, extends battery life, achieves more uniform lithium deposition, and improves battery cycle life, thus providing strong technical support for the safe operation and performance optimization of lithium metal batteries. These technical features work together to greatly improve the safety and energy density of lithium metal batteries, laying a solid foundation for the development of sustainable energy applications.

[0046] Other beneficial effects of the embodiments of the present invention will be further described below. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the OST-SRTM system according to an embodiment of the present invention. (a) Structure of the LMB pouch cell and the position of the SMF in the cell. (b) Backscattered Rayleigh signal (RBS) along the optical fiber and the principle of 1D temperature location. (c) Temperature conversion from one dimension to two dimensions. (d) Application of super-resolution algorithm in temperature distribution. (e) Dendrite growth process of lithium anode (top) and temperature map of lithium electrode after 50 cycles (from left to right: bare lithium, pyramid-shaped lithium, copper mesh patterned lithium, polylactic acid PLA-lithium).

[0048] Figure 2 This is a schematic diagram of the experimental setup for monitoring internal thermal events in a battery cell according to an embodiment of the present invention. (a) Experimental setup: ① Battery testing system for capacity and capacity readout; ② Infrared camera for external temperature data acquisition; ③ OFDR system for optical signal monitoring; ④ Laptop computer for all data acquisition and processing; ⑤ Scanning electron microscope (SEM) images for observing the anode morphology after cycling. (b) Relationship and function of optical, electrical, and thermal signal processing. (c) Anode morphology simulation. (d) Lithium anode protection strategies: pyramid pattern, copper mesh pattern, PLA coating.

[0049] Figure 3 This is a schematic diagram illustrating the principle of super-resolution OFDR temperature data processing based on SRGAN in an embodiment of the present invention. (a) Dataset constructed using OFDR and infrared thermal imaging. (b) Structure of the SRGAN model. (c) Changes in validation loss, PSNR, and SSIM during training.

[0050] Figure 4 The results verify the effectiveness of the lithium anode protection strategy in this invention. (a) Rate performance, (b) bare lithium, pyramidal lithium, copper lithium, and PLA-Li coin cells at 1–5 mA cm⁻¹. -2 (c) Resistance of bare lithium and PLA-Li at different cycle numbers. (d) Cycle performance of bare lithium, pyramidal lithium, copper mesh patterned lithium, and PLA-Li pouch cells. (e) and (f) Super-resolution images of different cells at 10, 30, and retention cycles (48, 56, 60, and 70, respectively). Box plot. (g) Bare lithium, pyramidal lithium, copper mesh patterned lithium, and PLA-Li after 50 cycles. SEM images (from top to bottom). (h) Radar images of different monitoring methods (① real-time, ② non-destructive, ③ precision, ④ equipment volume, ⑤ measurement area). Detailed Implementation

[0051] The embodiments of the present invention will be described in detail below. It should be emphasized that the following description is merely exemplary and is not intended to limit the scope and application of the present invention.

[0052] It should be noted that when a component is referred to as "fixed to" or "set on" another component, it can be directly on or indirectly on that other component. When a component is referred to as "connected to" another component, it can be directly connected to or indirectly connected to that other component. Furthermore, a connection can be used for fixing, coupling, or communication.

[0053] It should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0054] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0055] See Figures 1 to 3 This invention provides an in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers, comprising a single-mode fiber (SMF), an OFDR system, and a data processing unit. The SMF is used to construct a sensor and is connected to the inside of the battery. The SMF has a helical arrangement, distributing the fiber along a two-dimensional plane for temperature monitoring in physical space. The OFDR system captures temperature data of a one-dimensional continuous set of points along the SMF. The OFDR system includes a linearly tunable laser, an interferometer structure, a photodetector, and a data acquisition unit. The linearly tunable laser emits optical signals; the interferometer structure includes a reference arm and a measuring arm for measuring reflected optical signals propagating along the fiber; the photodetector and data acquisition unit acquire the optical signal data and convert it into electrical signal data. The data processing unit converts the one-dimensional temperature data captured by the OFDR system into two-dimensional planar distribution data corresponding to the helical arrangement of the SMF, thereby achieving accurate monitoring of the internal temperature distribution of the battery.

[0056] In a preferred embodiment, the data processing module includes a super-resolution algorithm module for performing super-resolution processing of temperature data, thereby improving the spatial resolution of temperature monitoring. See also... Figure 3 b) The super-resolution algorithm module includes: an SRGAN model for machine learning to achieve super-resolution of temperature mapping; the SRGAN model includes: a generator network, composed of ResNet blocks, for generating super-resolution images; and a discriminator network, a CNN structure, for evaluating high-resolution (HR) and super-resolution (SR) images. Figure 3 As shown in Figure b, in a preferred embodiment, the generator network comprises a deep network consisting of five ResNet blocks. Each ResNet block enhances the information flow by skipping connections to prevent gradient vanishing. The convolutional layer parameters of each ResNet block can be "n64k9s1p4", that is, 64 convolutional kernels of size 9×9 with a stride of 1 and padding of 4. The discriminator network is a convolutional neural network (CNN) that uses the Leaky ReLU activation function (ρ = 0.2) to avoid neuron death caused by negative output and ensure the effectiveness of network training. The output layer contains a dense block followed by a sigmoid function to perform binary scoring on high-resolution (HR) and super-resolution (SR) images to evaluate image quality.

[0057] In some embodiments, the system further includes an external temperature experimental apparatus for constructing a training dataset. The apparatus includes: ceramic heating elements of different shapes connected to a helically arranged single-mode fiber (SMF) to simulate the temperature distribution inside the battery; and an infrared camera for acquiring temperature data from the ceramic heating elements to form a dataset for training a super-resolution algorithm model. The temperature of the ceramic heating elements is controlled by a power supply, and the dataset is used for machine learning to achieve super-resolution temperature mapping. In a preferred embodiment, the temperature change video acquired by the infrared camera is used to record the temperature distribution on the battery surface. Frames are extracted from the video as images and synchronized with the data acquisition time of the OFDR system to form an original dataset. The original dataset is processed using data augmentation techniques to enhance the model's generalization ability. These data augmentation techniques include random rotation, reflection, translation, scaling, brightness and contrast adjustment, Gaussian blur, noise addition, and cropping to avoid overfitting and improve the accuracy and robustness of the super-resolution algorithm.

[0058] In some embodiments, the system further includes: a battery testing system for performing capacity and voltage readouts of the battery to monitor its electrochemical performance; a computer (such as a laptop computer) including the data processing unit for collecting and processing data from the battery testing system and the infrared camera, and managing data acquisition and analysis of the OFDR system; and a scanning electron microscope (SEM) for observing the anode morphology of the battery after cycling to assess the physical state and structural changes of the battery; wherein the computer synchronizes data from the OFDR system, the infrared camera, and the battery testing system, and performs data analysis and image processing to ensure the accuracy and integrity of the experimental data.

[0059] In a preferred embodiment, the data processing unit identifies areas without temperature readings and fills the data in these areas with a preset room temperature value, such as 22°C, to ensure the integrity and continuity of the temperature distribution data.

[0060] In some embodiments, a lithium anode protection strategy is applied to the monitored lithium metal battery, the lithium anode protection strategy including:

[0061] (1) Pyramid-patterned lithium anode: By forming a pyramid structure on the surface of the lithium anode to increase the surface area and improve the ion transport path, the battery performance is improved. Specifically, in a preferred embodiment, the pyramid-patterned lithium anode is constructed by imprinting a microstructure pattern onto the lithium anode using a pyramid-shaped mold made with high-precision 3D printing technology. The mold size is 50mm × 50mm × 0.5mm, the height of each pyramid is 50μm, the width is 50μm, and the spacing is 40μm. A pressure of 2MPa is applied during the imprinting process.

[0062] (2) Copper mesh patterned lithium anode: By creating micropatterns on the surface of the lithium anode to guide the uniform distribution and deposition of lithium ions, electrochemical performance can be enhanced. Specifically, in a preferred embodiment, the copper mesh patterned lithium anode uses a 250-mesh copper mesh to apply a pressure of 1 MPa to the lithium surface and performs two imprinting processes to improve the uniformity of surface modification. The first imprinting is performed in the direct direction, and the second imprinting is performed after rotating the mesh by 45 degrees.

[0063] (3) PLA-coated lithium anode, which improves thermal stability by applying a protective coating to the surface of the lithium anode to utilize its chemical and mechanical barrier properties. Specifically, in a preferred embodiment, the PLA-coated lithium anode is prepared by dissolving PLA in a DMSO solution, stirring at 130°C for 5 hours and continuously stirring at 80°C for 24 hours to obtain a uniform milky white solution, preparing a PLA solution with a mass ratio of 1.2%, and then scraping the solution onto the lithium metal surface to form a PLA coating.

[0064] The effectiveness of the above lithium anode protection strategies can be evaluated by monitoring using the OST-SRTM system of this invention.

[0065] This invention also provides an in-situ spatiotemporal thermal monitoring method for lithium metal batteries, using the aforementioned in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers.

[0066] This invention presents an innovative in-situ spatiotemporal thermal monitoring method for lithium metal batteries, providing real-time, non-destructive, and high-precision temperature monitoring covering a large measurement area, and possessing super-resolution capabilities. The system utilizes the high sensitivity and distributed characteristics of optical fiber distributed sensing (OFDR) technology, combined with the design of a helical arrangement of single-mode fiber (SMF), to convert one-dimensional temperature data into a two-dimensional planar distribution, thereby achieving precise monitoring of the internal temperature field of the battery. Furthermore, by introducing a super-resolution algorithm based on the SRGAN model, the system can convert low-resolution temperature images into high-resolution images, reconstructing details and improving spatial and temporal resolution. This integrated physical sensing and super-resolution algorithm method not only explores the mechanism of hotspot generation and verifies the effectiveness of battery protection strategies, but also reduces the risk of thermal runaway, extends battery life, achieves more uniform lithium deposition, and improves battery cycle life, thus providing strong technical support for the safe operation and performance optimization of lithium metal batteries. These technical features work together to greatly improve… lithium metal batteries Safety and energy density lay a solid foundation for the development of sustainable energy applications. Base.

[0067] The following describes specific embodiments of the present invention.

[0068] A system integrating physical sensing and super-resolution algorithms is proposed to achieve operational spatiotemporal super-resolution thermal monitoring (OST-SRTM) of lithium metal batteries (LMBs). OFDR is a distributed fiber optic sensing technology based on Rayleigh backscattering (RBS) with a spatial resolution of several sub-millimeters. A helical structure is constructed using single-mode fiber (SMF). Figure 1 The algorithm transforms a one-dimensional (1D) continuous set of points into discrete points distributed on a two-dimensional (2D) plane. Then, we employ a super-resolution algorithm (...). Figure 1 (d). The lithium anode was patterned using pyramidal imprints and a copper mesh, and an artificial solid electrolyte interface layer was formed by coating with polylactic acid (PLA). Figure 1(e). The developed system allows for high-precision evaluation of the effectiveness of these lithium anode protection strategies, rather than requiring post-cycle battery disassembly for destructive techniques such as scanning electron microscopy. By applying the protection strategies and our developed system, we observed extended battery life, lower temperature rise, more uniform temperature distribution, and fewer hot spots during battery capacity degradation cycles. These all contribute to more uniform lithium deposition, longer battery cycle life, and a reduced risk of thermal runaway. This novel monitoring approach, by combining state-of-the-art optical technology with energy storage diagnostics and creating a robust framework for evaluating thermal management technologies, significantly improves the safety and energy density of lithium-ion batteries (LMBs) for sustainable energy applications.

[0069] To investigate the internal temperature distribution and uneven lithium deposition of pouch cells, the experimental setup for signal acquisition and processing of the cells is as follows: Figure 2 As shown in a and b. Assemble Li / LFP pouch cells in a drying chamber. Place bare lithium, Li, or PLA-Li (65×60mm) cells... 2 (100μm thickness), separator and LFP cathode (65×60mm) 2 The cells were laminated together with a mass loading of 400 mg. Then, the molded SMF was bonded to the back of the copper current collector. Finally, after electrolyte injection, the entire cell was encapsulated with an aluminum-plastic film. The electrolyte was 1.0 M LiPF6, EC / DMC / DEC (1:1:1, v:v). One end of an optical fiber extended from the cell and connected to the OFDR system, while the other end was coiled into the mold. During cycling, the cell was connected to a cell testing system. One end of the optical fiber was connected to the OFDR testing system to monitor the internal temperature, while the external temperature was monitored by an infrared camera. To promote Li… + The uniformity of deposition provides three lithium anode protection strategies ( Figure 2 d). The first two are based on surface patterns: (1) Pressing a pyramid-shaped mold into lithium: A pyramid-shaped mold was made using high-precision 3D printing technology. The mold was designed to be 50mm × 50mm × 0.5mm in size, and the individual pyramids on the mold were characterized by a height of 50μm, a width of 50μm, and a spacing of 40μm between the pyramids. In order to press the pyramid structure onto the lithium anode, a pressure of 2MPa was applied, thereby embedding a microstructure into the lithium surface. The pattern is called a pyramid diagram. (2) Rolling copper mesh to create micropatterns: using 250 meshA copper mesh was applied to the lithium surface with a pressure of 1 MPa. To improve the uniformity of the surface modification, two imprinting processes were performed: initially in the direct direction, and then a second imprinting was performed after rotating the mesh by 45 degrees. (3) Coating with polylactic acid: An appropriate amount of PLA was dissolved in DMSO solution, stirred at 130°C for 5 h to dissolve, and then stirred continuously at 80°C for 24 h to obtain a uniform milky white solution. A PLA solution with a concentration of 1.2% (mass ratio) was prepared according to the PLA content. The 1.2% PLA solution was scraped onto the lithium metal surface with a knife to obtain a lithium metal sheet with a PLA coating. The OST-SRTM system was used to perform more accurate spatiotemporal monitoring of different lithium anodes.

[0070] External temperature experiments were conducted by connecting ceramic heating elements of different shapes to helically arranged optical fibers. Simultaneously, data was collected using an infrared camera to construct a training dataset. Using this dataset, we applied machine learning to achieve super-resolution temperature mapping. Figure 3 As shown in Figure a, T-shaped, circular, and two-point ceramic heating elements were placed below the SMF. The temperature of the elements was controlled by a power supply. Temperature data was recorded simultaneously using an OFDR system and an infrared camera. The OFDR system captured the data in a one-dimensional format and then processed it into a spiral structure. Areas without temperature readings were filled with room temperature values ​​(22°C). Simultaneously, the infrared camera recorded video of temperature changes in the same area. The video was framed into images and synchronized with the data acquisition time of the OFDR system to form the original dataset. To avoid overfitting due to the limited diversity of the dataset, data augmentation techniques were employed. Images in the original dataset were transformed using random rotation, reflection, translation, scaling, brightness and contrast adjustments, Gaussian blur, noise addition, and cropping. These modifications generated a more diverse and robust training dataset. A network was built based on the SRGAN model, as follows... Figure 3 As shown in b, the generator network is a deep network consisting of 5 ResNet blocks, designed to enhance information flow between layers by skipping connections and preventing gradient vanishing as network depth increases. For example, "n64k9s1p4" refers to 64 convolutional kernels of size 9×9 with a stride of 1 and padding of 4. The discriminator network is a CNN that uses Leaky ReLU as the activation function to avoid negative outputs that produce dead neurons. At the end of the network, a dense block is followed by a sigmoid function for binary classification, scoring high-resolution (HR) and super-resolution (SR) images. In the optimal model of this invention, we achieved a PSNR of 25 dB and an SSIM of 0.7 on the test set, successfully converting one-dimensional temperature data from the OFDR system into a super-resolution two-dimensional temperature map.

[0071] The OST-SRTM system was used to validate the protection method. High-load commercial LiFePO4 was used as the cathode, paired with different lithium metal anodes, to evaluate the protective effect throughout the cycling process and monitor temperature changes. By comparing the cycle stability and capacity retention of different pouch cells, PLA-Li was found to have the best electrochemical performance. The SMF's temperature response spectrum was tested to observe the internal structure of the battery. The temperature variation trend was studied, and the electrochemical and temperature characteristics were evaluated. A pouch cell using bare lithium batteries was used. No. The capacity decreased at 48 cycles, while cells treated with pyramid patterns, Cu patterns, and PLA showed capacity decreases at 56, 60, and 70 cycles. Figure 4 d). Our designed system measured the space-time temperature maps of four batteries ( ). Figure 4 (e, f). During the first 10 cycles, the battery capacity and performance remained stable, with no obvious hot spots observed. Box plots combined with scatter density plots illustrate the concentration and dispersion of temperature distribution for each electrode type after cycling. Bare lithium exhibited the widest temperature range (from -5°C to 8°C) and a relatively high median, indicating a significant temperature rise during cycling. The protected anode showed a more concentrated temperature distribution with a narrower range, suggesting that the protection strategy could partially suppress temperature changes. As cycling progressed, heat accumulation became apparent. Pyramid-type and Cu-type Li showed smaller temperature rises and greater temperature stability, while PLA-Li exhibited the smallest temperature rise and the most stable distribution. In capacity retention cycles, compared to bare lithium (5.3°C), the protected anode showed lower average temperature rises (3.1°C, 2.6°C, and 0.7°C) and fewer hot spot areas in later capacity degradation cycles. The scatter density plot of bare lithium showed larger temperature fluctuations, indicating uneven distribution and the presence of multiple hot spots. PLA-Li demonstrated the best thermal management capabilities, making it suitable for applications requiring high thermal stability. After 100 cycles, cells retaining larger capacities exhibited more heat accumulation; eventually, the temperatures of all cells converged. This indicates that under normal cycling conditions, as battery capacity decreases, remaining capacity continues to influence internal temperature. Non-destructive methods have been developed for monitoring battery status, which can be categorized into electrical, acoustic, and optical methods. Most electrical monitoring methods focus on single-point and external parameters such as impedance, voltage, current, and thermal conductivity. Acoustic methods convert a single point into a two-dimensional representation through scanning. Due to its advantages such as resistance to electromagnetic interference, high sensitivity, corrosion resistance, and multi-point monitoring capabilities, optical methods have become the most promising technology for monitoring the internal temperature distribution and strain changes of batteries. Compared with other optical sensors, our developed OST-SRTM system features in-situ observation, two-dimensional spatiotemporal (1 frame / 3 seconds), and high resolution (16 points / cm). 2 Super-resolution up to 1551 dots / cm 2 Advantages such as )

[0072] In some alternatives, optical fibers of different shapes can be used for two-dimensional data conversion; Cu and pyramidal structures with different parameters can be used for negative electrode protection.

[0073] In summary, this invention proposes an in-situ spatiotemporal thermal monitoring system and method for lithium metal batteries, effectively addressing the challenges of distributed thermal sensing within the battery. The system design comprises two core components: First, based on OFDR technology, a modeled helical optical fiber structure is designed, which not only reduces stress but also expands the sensing range; second, an SRGAN super-resolution machine learning algorithm based on a self-built dataset is developed, significantly improving the sensing resolution.

[0074] The main innovative contributions of this invention include the following aspects: 1. A method based on helical junctions is proposed. Data structure two 2D processing technology enables the system to convert one-dimensional temperature data into a two-dimensional plane. 1. The distribution of temperature data was improved, enabling more accurate monitoring of internal temperature changes within the battery. 2. A temperature sensing dataset based on helical optical fibers and heating ceramics was constructed, providing a foundation for the training and validation of super-resolution algorithms. 3. A super-resolution temperature monitoring machine learning algorithm based on SRGAN was implemented, which improves the spatial resolution of temperature monitoring and provides a new technical means for battery thermal management. 4. Lithium anode protection technologies based on Cu patterning, pyramid patterning, and PLA coating were developed. These technologies can effectively reduce the occurrence of hot spots during battery capacity decay, improving battery safety and performance. 5. Distributed optical fiber technology was applied to temperature acquisition in lithium metal pouch batteries, achieving accurate monitoring of spatiotemporal temperature distribution.

[0075] Thanks to the distributed monitoring and high sensitivity of OFDR technology, the two-dimensional data representation achieved through a spiral structure, and the super-resolution capability of a machine learning algorithm based on the SRGAN model, this invention enables real-time, non-destructive, and high-precision monitoring of the temperature distribution of lithium anodes, covering a large measurement area and possessing super-resolution capabilities. The inventors discovered that the presence of localized hotspots is a key factor leading to thermal runaway and affecting lithium deposition. The pyramid, copper mesh pattern, and PLA strategies employed in this invention effectively reduce the occurrence of these hotspots. Since even small temperature changes can have a significant impact on battery performance, this invention is of great significance for precise thermal management of lithium metal batteries. This invention not only promotes the development of battery diagnostic technology but also provides a solid foundation for developing safer and more efficient energy storage solutions.

[0076] The above description provides a further detailed explanation of the present invention in conjunction with specific / preferred embodiments, and it should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various substitutions or modifications can be made to these described embodiments without departing from the concept of the present invention, and all such substitutions or modifications should be considered within the scope of protection of the present invention. In the description of this specification, the reference to terms such as "an embodiment," "some embodiments," "preferred embodiment," "example," "specific example," or "some examples," etc., indicates that the specific features, structures, materials, or characteristics described in connection with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples. Although the embodiments of the present invention and their advantages have been described in detail, it should be understood that various changes, substitutions, and modifications can be made herein without departing from the scope of protection of the patent application.

Claims

1. A distributed optical fiber-based in-situ spatiotemporal thermal monitoring system for lithium metal batteries, characterized in that, include: A single-mode fiber (SMF) is connected to the inside of the battery. The SMF has a helical arrangement structure, which makes the fiber distributed along a two-dimensional plane, and is used to realize temperature monitoring in physical space. An OFDR system for capturing temperature data of a one-dimensional continuous point set along the single-mode fiber (SMF), the OFDR system comprising: A linearly tunable laser is used to emit optical signals; The interferometer structure includes a reference arm and a measuring arm, used to measure reflected light signals propagating along an optical fiber; Photodetectors and data acquisition instruments are used to collect optical signal data and convert it into electrical signal data. A data processing unit is used to convert the one-dimensional temperature data captured by the OFDR system into two-dimensional planar distribution data corresponding to the helical arrangement structure of the single-mode fiber (SMF), so as to achieve accurate monitoring of the internal temperature distribution of the battery. The data processing unit includes a super-resolution algorithm module for performing super-resolution processing of temperature data to improve the spatial resolution of temperature monitoring. The super-resolution algorithm module includes an SRGAN model for machine learning to achieve super-resolution temperature mapping. The SRGAN model includes: a generator network, which uses ResNet blocks to form a deep network. Each ResNet block enhances the information flow through skip connections to generate super-resolution images; a discriminator network, which is a convolutional neural network (CNN) structure, used to evaluate high-resolution and super-resolution images. The convolutional neural network uses the Leaky ReLU activation function; the output layer contains a dense block followed by a sigmoid function to perform binary scoring on high-resolution and super-resolution images to evaluate image quality.

2. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fiber as described in claim 1, characterized in that, The generator network comprises a deep network consisting of five ResNet blocks, each containing 64 9×9 convolutional kernels with a stride of 1 and padding of 4.

3. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers as described in claim 1, characterized in that, The system further includes an external temperature experimental setup for constructing a training dataset, the setup comprising: Multiple ceramic heating elements are connected to a single-mode fiber (SMF) arranged in a spiral pattern to simulate the temperature distribution inside the battery; An infrared camera is used to collect temperature data of the ceramic heating element to form a dataset for training a super-resolution algorithm model; wherein the temperature of the ceramic heating element is controlled by a power supply, and the dataset is used for machine learning to achieve super-resolution temperature mapping.

4. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fiber as described in claim 3, characterized in that, The infrared camera captures video of temperature changes to record the temperature distribution on the battery surface. Frames are extracted from the video as images and synchronized with the data acquisition time of the OFDR system to form a raw dataset. The raw dataset is then processed by data augmentation techniques, including random rotation, reflection, translation, scaling, brightness and contrast adjustment, Gaussian blur, noise addition, and cropping.

5. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fiber as described in any one of claims 1 to 4, characterized in that, Also includes: A battery testing system is used to perform capacity and voltage readouts of batteries to monitor their electrochemical performance. The computer, including the data processing unit, is used to collect and process data from the battery testing system and the infrared camera, as well as to manage the data acquisition and analysis of the OFDR system; Scanning electron microscopy (SEM) is used to observe the morphology of the anode after battery cycling in order to assess the physical state and structural changes of the battery. The computer synchronizes data from the OFDR system, infrared camera, and battery testing system, and performs data analysis and image processing.

6. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fiber as described in any one of claims 1 to 4, characterized in that, The data processing unit identifies areas without temperature readings and fills these areas with preset room temperature values.

7. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers as described in claim 6, characterized in that, The preset room temperature value is 22℃.

8. The in-situ lithium metal battery based on distributed optical fiber as described in any one of claims 1 to 4 Spatiotemporal thermal monitoring system Its features are, Apply lithium anode protection to the monitored lithium metal battery The strategy, the lithium anode protection strategy includes: Pyramid-patterned lithium anodes increase surface area and improve ion transport pathways by forming pyramid structures on the surface of lithium anodes. Copper mesh patterned lithium anodes utilize micropatterns created on the surface of lithium anodes to guide the uniform distribution and deposition of lithium ions. PLA-coated lithium anodes enhance thermal stability by applying a protective coating to the surface of the lithium anode, taking advantage of its chemical and mechanical barrier properties. Monitoring was conducted using the OST-SRTM system to evaluate the effectiveness of different lithium anode protection strategies.

9. The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fiber as described in claim 8, characterized in that, The pyramid-patterned lithium anode is created by imprinting a microstructure pattern onto the lithium anode using a pyramid-shaped mold made with high-precision 3D printing technology. The mold size is 50mm×50mm×0.5mm, the height of a single pyramid is 50μm, the width is 50μm, and the spacing is 40μm. A pressure of 2MPa is applied during the imprinting process. The copper mesh patterned lithium anode uses a 250-mesh copper mesh to apply a pressure of 1 MPa to the lithium surface and performs two imprinting processes to improve the uniformity of surface modification. The first imprinting is done in the direct direction, and the second imprinting is done after rotating the mesh by 45 degrees. The PLA-coated lithium anode is prepared by dissolving PLA in DMSO solution, stirring at 130°C for 5 hours and continuously stirring at 80°C for 24 hours to obtain a uniform milky white solution, preparing a PLA solution with a mass ratio of 1.2%, and then scraping the solution onto the lithium metal surface to form a PLA coating.

10. A method for in-situ spatiotemporal thermal monitoring of lithium metal batteries, characterized in that, The in-situ spatiotemporal thermal monitoring system for lithium metal batteries based on distributed optical fibers as described in any one of claims 1 to 9 is used.

Images