Intelligent high-safety lithium battery fault experiment device and prediction control system

Abstract

The invention discloses an intelligent high-safety lithium battery fault experiment device and a prediction control system. A four-layer composite protection cover and a multispectral monitoring window are integrated, and a three-dimensional safety barrier is formed from physical protection to real-time monitoring. The four layers of protective covers resist impact through the metal alloy layer, absorb energy through the honeycomb layer, insulate heat through the ceramic layer and uniformly dissipate heat through the graphene coating, so that thermal runaway diffusion is effectively inhibited; the multispectral monitoring window is combined with infrared thermal imaging and high-speed camera shooting, temperature abnormity and battery deformation are captured in real time, and accurate data support is provided for early warning. An LSTM model carried by the edge calculation unit predicts the thermal runaway risk 5-10 minutes ahead of time by fusing the temperature gradient, the gas concentration change and the voltage spectrum characteristics, triggers a grading response strategy, and forms a progressive active protection mechanism from sound-light alarm and liquid nitrogen directional cooling to millisecond power failure and aerosol fire extinguishing; and the device is suitable for electric vehicle battery packs and other complex scenes, and the scientific research practicability and industry universality of the device are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery testing, and in particular relates to an intelligent high-safety lithium battery failure test device and a prediction control system. Background Art

[0002] With the rapid development of new energy technologies, lithium batteries are widely used in electric vehicles, energy storage systems, and consumer electronics due to their high energy density and long cycle life. However, lithium batteries are prone to thermal runaway under extreme conditions such as overcharge, over-discharge, short circuit, and high temperature. This can release large amounts of toxic gases (such as HF and CO) and cause fires or explosions, posing a serious threat to human safety and environmental health.

[0003] Most existing lithium battery failure experimental devices improve experimental safety through multi-layer protective covers, aerosol fire extinguishing systems and sensor networks, but they still have the following technical bottlenecks:

[0004] Passive response mechanism: Relying on fixed thresholds to trigger protection actions (such as initiating fire extinguishing when the temperature exceeds 150°C), it is unable to predict potential risks through data trends in the early stages of a fault, resulting in delayed response.

[0005] Data silo problem: Sensor data is only used for real-time monitoring and is not integrated with historical data or electrochemical models, lacking a comprehensive assessment of the battery's state of health (SOH).

[0006] Insufficient scalability: Existing devices are mostly designed for a single battery type (such as 18650 cylindrical batteries), making it difficult to adapt to new battery types such as solid-state batteries and flexible batteries, and cannot support series / parallel testing of multiple battery packs;

[0007] Energy efficiency and environmental disadvantages: Traditional gas treatment relies on activated carbon adsorption, requiring frequent filter replacement;

[0008] Lack of remote collaboration capabilities: Experimental data is limited to local storage and lacks cloud synchronization, multi-terminal collaborative analysis, and remote emergency intervention functions.

[0009] To address the above issues, the present invention proposes an intelligent and highly secure lithium battery failure test device, which achieves a technological leap from "passive protection" to "active prevention" through artificial intelligence prediction models, multimodal data fusion, dynamic response strategies and cloud-based collaborative control. Summary of the Invention

[0010] The purpose of the present invention is to provide an intelligent and highly safe lithium battery failure test device and a predictive control system in order to solve the above-mentioned problems.

[0011] The technical solution adopted by the present invention is as follows: an intelligent high-safety lithium battery fault test device and a predictive control system, the fault test device comprising:

[0012] Enhanced four-layer composite protective cover:

[0013] Outer metal alloy layer: 6061-T6 aviation aluminum alloy, thickness 2mm, tensile strength ≥310MPa, surface anodized to improve corrosion resistance;

[0014] Middle honeycomb energy absorbing layer: Aluminum honeycomb core material density is 40kg / m 3 , honeycomb aperture is 3mm, which can absorb more than 80% of the impact energy;

[0015] Inner ceramic insulation layer: made of alumina ceramic (Al2O3 content ≥99%), thickness 5mm, temperature resistance ≥1600℃;

[0016] Newly added graphene coating: A graphene nano-coating (50 μm thick) is sprayed on the surface of the ceramic layer, with a thermal conductivity of 5300 W / (m·K), which can quickly and evenly disperse local high temperatures and shield external electromagnetic interference.

[0017] Modular expansion interface:

[0018] The test bench is equipped with a universal battery slot interface, supporting plug-and-play adaptation of 18650, 21700, square batteries (such as CATL CTP batteries) and solid-state batteries;

[0019] The interface expansion slots can be freely combined to support up to 16 groups of battery series / parallel testing to simulate electric vehicle battery pack scenarios.

[0020] Multi-spectral monitoring window: replaces traditional explosion-proof glass with a multi-layer composite transmission material: the outer layer is sapphire glass (thickness 3mm, Mohs hardness 9), the middle layer is an infrared filter (transmission band 8-14μm), and the inner layer is an anti-fog coating; the multi-spectral monitoring window is equipped with a dual camera module inside.

[0021] In a preferred embodiment, the modular expansion interface adopts a magnetic attraction + spring contact design to ensure that the battery is firmly installed and the contact resistance is less than 1mΩ.

[0022] In a preferred embodiment, the interior of the multi-spectral monitoring window is provided with:

[0023] Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, accuracy ±1℃;

[0024] Visible light high-speed camera: frame rate 1000fps, capturing battery deformation processes such as expansion and rupture.

[0025] In a preferred embodiment, a predictive control system for an intelligent, high-safety lithium battery failure test device includes: a data acquisition and sensing module, an edge computing and AI prediction module, a dynamic control and execution module, an environmental processing module, a cloud collaboration and data management module, and a modular expansion interface.

[0026] In a preferred embodiment, the submodules of the data acquisition and sensing module include:

[0027] Temperature monitoring unit:

[0028] Thermocouple sensor: collects battery surface and tab temperature with an accuracy of ±0.5°C.

[0029] Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, generates temperature thermogram.

[0030] Gas sensing unit:

[0031] CO / HF gas sensor: sampling frequency 100Hz, detection accuracy ±1ppm, monitoring gas concentration change rate (d[CO] / dt).

[0032] Electrochemical signal unit:

[0033] Hall current sensor: monitors charging and discharging current fluctuations in real time.

[0034] Voltage sensor: records voltage data and uses FFT analysis to extract abnormal harmonics (e.g., 1kHz harmonics indicate lithium plating risk).

[0035] Deformation monitoring unit:

[0036] Visible light high-speed camera: frame rate 1000fps, capturing the battery expansion and rupture process.

[0037] In a preferred embodiment, the edge computing and AI prediction module includes:

[0038] Hardware platform: NVIDIA Jetson Xavier NX: 21TOPS computing power, support for TensorFlow Lite and PyTorch frameworks, 15W power consumption.

[0039] LSTM neural network model:

[0040] Input parameters: temperature gradient (ΔT / Δt), CO concentration change rate (d[CO] / dt), and voltage fluctuation spectrum (FFT analysis results).

[0041] Output: Thermal runaway risk probability value (0% to 100%), based on 3000 thermal runaway case training, validation set AUC value is 0.93.

[0042] Multi-level dynamic response strategy:

[0043] Level 1 warning (risk <30%): triggers an audible and visual alarm (buzzer + LED), and the data is marked as "observation status".

[0044] Level 2 intervention (30% ≤ risk < 70%): the charge and discharge current is reduced to 50% of the rated value, and the liquid nitrogen micro-spray system is started (nozzle diameter 0.1mm, tab cooling rate ≥10℃ / s).

[0045] Level 3 protection (risk ≥ 70%): cut off the experimental circuit within 0.1s, start aerosol fire extinguishing (concentration ≥ 100g / m 3 ) and photocatalytic oxidation module.

[0046] In a preferred embodiment, the dynamic control and execution module is divided into three submodules: current control, temperature control execution, and safety protection. The current control submodule dynamically adjusts the charge and discharge currents through a programmable power supply with an accuracy of 0.1 ampere, enabling precise intervention in the battery status. The temperature control execution submodule utilizes a liquid nitrogen micro-spray system, which sprays the battery tabs with a nozzle with a diameter of 0.1 mm, achieving a cooling rate exceeding 10 degrees Celsius per second. The safety protection submodule integrates a high-voltage relay and an aerosol fire extinguishing device. The relay disconnects the experimental circuit in milliseconds, and the fire extinguishing device sprays a fire extinguishing agent at a concentration of more than 100 grams per cubic meter to ensure full coverage and rapid suppression of thermal runaway.

[0047] In a preferred embodiment, the environmental treatment module consists of three submodules: a photocatalytic reactor, a chemical reaction pathway, and a performance monitoring module. The photocatalytic reactor submodule has a built-in ultraviolet LED array and a titanium dioxide nanocatalyst coating. The ultraviolet light source has a wavelength of 365 nanometers, the catalyst particle size is 10 to 20 nanometers, and the specific surface area exceeds 200 square meters per gram. The chemical reaction pathway submodule converts hydrogen fluoride into fluorine gas and water, and carbon monoxide into carbon dioxide, with conversion efficiencies exceeding 88 percent and 85 percent, respectively. The performance monitoring submodule ensures that the system resistance is less than 50 Pascals, is compatible with a gas flow rate of 10 to 20 cubic meters per hour, and has a catalyst life of up to five years without the need for filter replacement.

[0048] In a preferred embodiment, the cloud collaboration and data management module includes three sub-modules: cloud platform, core functions, and local storage. The cloud platform sub-module is based on the Alibaba Cloud IoT architecture and supports real-time data transmission using the MQTT protocol. The core function sub-module provides a remote monitoring panel with visualized temperature thermograms and gas concentration curves, supports multi-user permission hierarchical operations, and has built-in statistical tools to generate battery health status attenuation reports with an error of less than three percent. The local storage sub-module is equipped with a solid-state hard drive in the edge computing unit to temporarily store data when the network is disconnected and automatically resume transmission after the network is restored, ensuring the integrity and continuity of the experimental data.

[0049] In a preferred embodiment, the modular expansion interface consists of two submodules: a universal battery interface and an expansion slot. The universal battery interface submodule utilizes a magnetic structure and spring contact design, achieving a contact resistance of less than one milliohm. It supports plug-and-play operation with 18650 cylindrical batteries, 21700 batteries, prismatic batteries, and solid-state batteries. The expansion slot submodule accommodates up to sixteen battery groups for series or parallel testing, accurately simulating complex scenarios such as electric vehicle battery packs. It adapts to mainstream battery technologies over the next five years and expands its application range without hardware modifications.

[0050] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0051] 1. This invention integrates a four-layer composite protective cover and a multispectral monitoring window, creating a three-dimensional safety barrier from physical protection to real-time monitoring. The four-layer protective cover effectively suppresses the spread of thermal runaway through the impact resistance of the metal alloy layer, the energy absorption of the honeycomb layer, the thermal insulation of the ceramic layer, and the uniform heat dissipation of the graphene coating. The multispectral monitoring window combines infrared thermal imaging with high-speed video to capture temperature anomalies and battery deformation in real time, providing accurate data support for early warning. The LSTM model on the edge computing unit integrates temperature gradients, gas concentration changes, and voltage spectrum characteristics to predict thermal runaway risks 5-10 minutes in advance, triggering a tiered response strategy—from audible and visual alarms, targeted liquid nitrogen cooling, to millisecond-level power outages and aerosol fire extinguishing—forming a progressively progressive active protection mechanism. The modular design is compatible with cylindrical, prismatic, and solid-state batteries, supports testing of multiple battery packs in series and parallel, and is suitable for complex scenarios such as electric vehicle battery packs, significantly enhancing the device's scientific practicality and industry applicability.

[0052] 2. In the present invention, the photocatalytic oxidation module uses ultraviolet light to excite nanocatalysts to efficiently convert the HF and CO produced in the experiment into harmless fluorine gas, water and carbon dioxide, with conversion rates exceeding 88% and 85% respectively. The catalyst life is up to 5 years, and no consumables need to be replaced, which greatly reduces operation and maintenance costs. The cloud-based collaborative platform uses the Internet of Things architecture to achieve remote data synchronization and multi-terminal collaboration. Administrators can view temperature thermograms, gas concentration curves and risk levels in real time, and generate battery health status reports based on statistical tools with an error of less than 3%. The complementary data storage mechanism of local edge computing and the cloud ensures that experimental data can be fully saved and retransmitted even in the event of a network outage, ensuring scientific research continuity and data reliability. The integration of these technologies not only promotes the transformation of lithium battery safety testing from passive response to active prevention, but also provides the industry with efficient, economical and environmentally friendly solutions through green and intelligent design. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the overall structure of the experimental device of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] Example:

[0056] Reference Figure 1 , an intelligent high-safety lithium battery fault test device, the fault test device includes:

[0057] Enhanced four-layer composite protective cover:

[0058] Outer metal alloy layer: 6061-T6 aviation aluminum alloy, thickness 2mm, tensile strength ≥310MPa, surface anodized to improve corrosion resistance;

[0059] Middle honeycomb energy absorbing layer: Aluminum honeycomb core material density is 40kg / m 3 , honeycomb aperture is 3mm, which can absorb more than 80% of the impact energy;

[0060] Inner ceramic insulation layer: made of alumina ceramic (Al2O3 content ≥99%), thickness 5mm, temperature resistance ≥1600℃;

[0061] Newly added graphene coating: A graphene nano-coating (50 μm thick) is sprayed on the surface of the ceramic layer, with a thermal conductivity of 5300 W / (m·K), which can quickly and evenly disperse local high temperatures and shield external electromagnetic interference.

[0062] Modular expansion interface:

[0063] The test bench is equipped with a universal battery slot interface, supporting plug-and-play adaptation of 18650, 21700, square batteries (such as CATL CTP batteries) and solid-state batteries;

[0064] The interface expansion slots can be freely combined to support up to 16 groups of battery series / parallel testing to simulate electric vehicle battery pack scenarios.

[0065] Multi-spectral monitoring window: replaces traditional explosion-proof glass with a multi-layer composite transmission material: the outer layer is sapphire glass (thickness 3mm, Mohs hardness 9), the middle layer is an infrared filter (transmission band 8-14μm), and the inner layer is an anti-fog coating; the multi-spectral monitoring window is equipped with a dual camera module inside.

[0066] The modular expansion interface uses a magnetic + spring contact design to ensure that the battery is firmly installed and the contact resistance is less than 1mΩ.

[0067] The internal settings of the multi-spectral monitoring window are:

[0068] Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, accuracy ±1℃;

[0069] Visible light high-speed camera: frame rate 1000fps, capturing battery deformation processes such as expansion and rupture.

[0070] A predictive control system for an intelligent, high-safety lithium battery failure test device. The predictive control system includes: a data acquisition and sensing module, an edge computing and AI prediction module, a dynamic control and execution module, an environmental processing module, a cloud collaboration and data management module, and a modular expansion interface.

[0071] The sub-modules of the data acquisition and sensing module include:

[0072] Temperature monitoring unit:

[0073] Thermocouple sensor: collects battery surface and tab temperature with an accuracy of ±0.5°C.

[0074] Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, generates temperature thermogram.

[0075] Gas sensing unit:

[0076] CO / HF gas sensor: sampling frequency 100Hz, detection accuracy ±1ppm, monitoring gas concentration change rate (d[CO] / dt).

[0077] Electrochemical signal unit:

[0078] Hall current sensor: monitors charging and discharging current fluctuations in real time.

[0079] Voltage sensor: records voltage data and uses FFT analysis to extract abnormal harmonics (e.g., 1kHz harmonics indicate lithium plating risk).

[0080] Deformation monitoring unit:

[0081] Visible light high-speed camera: frame rate 1000fps, capturing the battery expansion and rupture process.

[0082] Edge computing and AI prediction modules include:

[0083] Hardware platform: NVIDIA Jetson Xavier NX: 21TOPS computing power, support for TensorFlow Lite and PyTorch frameworks, 15W power consumption.

[0084] LSTM neural network model:

[0085] Input parameters: temperature gradient (ΔT / Δt), CO concentration change rate (d[CO] / dt), and voltage fluctuation spectrum (FFT analysis results).

[0086] Output: Thermal runaway risk probability value (0% to 100%), based on 3000 thermal runaway case training, validation set AUC value is 0.93.

[0087] Multi-level dynamic response strategy:

[0088] Level 1 warning (risk <30%): triggers an audible and visual alarm (buzzer + LED), and the data is marked as "observation status".

[0089] Level 2 intervention (30% ≤ risk < 70%): the charge and discharge current is reduced to 50% of the rated value, and the liquid nitrogen micro-spray system is started (nozzle diameter 0.1mm, tab cooling rate ≥10℃ / s).

[0090] Level 3 protection (risk ≥ 70%): cut off the experimental circuit within 0.1s, start aerosol fire extinguishing (concentration ≥ 100g / m 3 ) and photocatalytic oxidation module.

[0091] The dynamic control and execution module is divided into three submodules: current control, temperature control execution, and safety protection. The current control submodule dynamically adjusts the charge and discharge currents through a programmable power supply with an accuracy of 0.1 ampere, enabling precise intervention in the battery status. The temperature control execution submodule utilizes a liquid nitrogen micro-spray system, which sprays liquid nitrogen onto the battery tabs through a nozzle with a diameter of 0.1 mm, achieving a cooling rate exceeding 10 degrees Celsius per second. The safety protection submodule integrates a high-voltage relay and an aerosol fire extinguishing device. The relay disconnects the experimental circuit in milliseconds, while the fire extinguishing device sprays fire extinguishing agent at a concentration of over 100 grams per cubic meter to ensure rapid suppression of thermal runaway.

[0092] The environmental treatment module consists of three submodules: a photocatalytic reactor, a chemical reaction pathway, and performance monitoring. The photocatalytic reactor submodule houses a built-in UV LED array and a titanium dioxide nanocatalyst coating. The UV light source has a wavelength of 365 nanometers, the catalyst particle size ranges from 10 to 20 nanometers, and the specific surface area exceeds 200 square meters per gram. The chemical reaction pathway submodule converts hydrogen fluoride into fluorine gas and water, and carbon monoxide into carbon dioxide, with conversion efficiencies exceeding 88 percent and 85 percent, respectively. The performance monitoring submodule ensures system resistance is below 50 Pascals, compatible with gas flow rates of 10 to 20 cubic meters per hour, and a catalyst lifespan of up to five years without the need for filter replacement.

[0093] The cloud collaboration and data management module consists of three submodules: cloud platform, core functions, and local storage. The cloud platform submodule is based on the Alibaba Cloud IoT architecture and supports real-time data transmission via the MQTT protocol. The core functions submodule provides a remote monitoring panel with visualized temperature heat maps and gas concentration curves, supports multi-user permission levels, and includes built-in statistical tools to generate battery health degradation reports with an error of less than 3%. The local storage submodule uses a solid-state drive in the edge computing unit to temporarily store data during network outages and automatically resume transmission after the network is restored, ensuring the integrity and continuity of experimental data.

[0094] The modular expansion interface consists of two submodules: a universal battery interface and an expansion slot. The universal battery interface submodule utilizes a magnetic structure and spring contact design, achieving a contact resistance of less than one milliohm. It supports plug-and-play operation with 18650 cylindrical, 21700, prismatic, and solid-state batteries. The expansion slot submodule accommodates up to sixteen battery groups for series or parallel testing, accurately simulating complex scenarios such as electric vehicle battery packs. It adapts to mainstream battery technologies over the next five years and expands its application range without hardware modifications.

[0095] From the above, it can be seen that: in the present invention, a four-layer composite protective cover and a multi-spectral monitoring window are integrated to form a three-dimensional safety barrier from physical protection to real-time monitoring. The four-layer protective cover effectively suppresses the spread of thermal runaway through the impact resistance of the metal alloy layer, the energy absorption of the honeycomb layer, the heat insulation of the ceramic layer and the uniform heat dissipation of the graphene coating; the multi-spectral monitoring window combines infrared thermal imaging and high-speed camera to capture temperature anomalies and battery deformation in real time, providing accurate data support for early warning. The LSTM model carried by the edge computing unit predicts the risk of thermal runaway 5-10 minutes in advance by fusing temperature gradients, gas concentration changes and voltage spectrum characteristics, triggering a graded response strategy - from sound and light alarms, liquid nitrogen directional cooling to millisecond-level power outages and aerosol fire extinguishing, forming a step-by-step active protection mechanism. The modular design is compatible with cylindrical, square and solid-state batteries, supports series / parallel testing of multiple battery packs, and is suitable for complex scenarios such as electric vehicle battery packs, significantly improving the scientific research practicality and industry universality of the device.

[0096] In the present invention, the photocatalytic oxidation module uses ultraviolet light to excite nanocatalysts to efficiently convert the HF and CO produced in the experiment into harmless fluorine gas, water and carbon dioxide, with conversion rates exceeding 88% and 85% respectively. The catalyst life is up to 5 years, and no consumables need to be replaced, which greatly reduces operation and maintenance costs. The cloud-based collaborative platform uses the Internet of Things architecture to achieve remote data synchronization and multi-terminal collaboration. Administrators can view temperature thermograms, gas concentration curves and risk levels in real time, and generate battery health status reports based on statistical tools with an error of less than 3%. The complementary data storage mechanism of local edge computing and the cloud ensures that experimental data can be fully saved and retransmitted even in the event of a network outage, ensuring scientific research continuity and data reliability. The integration of these technologies not only promotes the transformation of lithium battery safety testing from passive response to active prevention, but also provides the industry with efficient, economical and environmentally friendly solutions through green and intelligent design.

[0097] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further limitations, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0098] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An intelligent and highly safe lithium battery failure test device and predictive control system, characterized by: The fault experiment device comprises: Enhanced four-layer composite protective cover: Outer metal alloy layer: 6061-T6 aviation aluminum alloy, thickness 2mm, tensile strength ≥310MPa, surface anodized to improve corrosion resistance; Middle honeycomb energy absorbing layer: Aluminum honeycomb core material density is 40kg / m 3 , honeycomb aperture is 3mm, which can absorb more than 80% of the impact energy; Inner ceramic insulation layer: made of alumina ceramic, 5mm thick, temperature resistant ≥1600℃; New graphene coating: A graphene nano-coating is sprayed on the surface of the ceramic layer, with a thermal conductivity of 5300W / (m·K), which can quickly and evenly disperse local high temperatures and shield external electromagnetic interference; Modular expansion interface: The test bench is equipped with a universal battery slot interface, supporting plug-and-play adaptation of 18650, 21700, square batteries and solid-state batteries; The interface expansion slots can be freely combined to support up to 16 groups of battery series / parallel testing to simulate electric vehicle battery pack scenarios; Multi-spectral monitoring window: replaces traditional explosion-proof glass with a multi-layer composite transmission material: the outer layer is sapphire glass, the middle layer is an infrared filter, and the inner layer is an anti-fog coating; the multi-spectral monitoring window is equipped with a dual camera module inside.

2. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The modular expansion interface adopts a magnetic + spring contact design to ensure that the battery is firmly installed and the contact resistance is less than 1mΩ.

3. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The interior of the multi-spectral monitoring window is provided with: Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, accuracy ±1℃; Visible light high-speed camera: frame rate 1000fps, capturing battery deformation processes such as expansion and rupture.

4. The predictive control system of an intelligent high-safety lithium battery failure test device according to claim 1, characterized in that: The predictive control system includes: a data acquisition and sensing module, an edge computing and AI prediction module, a dynamic control and execution module, an environmental protection processing module, a cloud collaboration and data management module, and a modular expansion interface.

5. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The submodules of the data acquisition and sensing module include: Temperature monitoring unit: Thermocouple sensor: collects battery surface and tab temperature with an accuracy of ±0.5°C; Infrared thermal imaging camera: resolution 640×480, temperature measurement range -20℃~1500℃, generating temperature thermogram; Gas sensing unit: CO / HF gas sensor: sampling frequency 100Hz, detection accuracy ±1ppm, monitoring gas concentration change rate; Electrochemical signal unit: Hall current sensor: real-time monitoring of charging and discharging current fluctuations; Voltage sensor: records voltage data and extracts abnormal harmonics through FFT analysis; Deformation monitoring unit: Visible light high-speed camera: frame rate 1000fps, capturing the battery expansion and rupture process.

6. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The edge computing and AI prediction module includes: Hardware platform: NVIDIA Jetson Xavier NX: 21TOPS computing power, support for TensorFlow Lite and PyTorch frameworks, 15W power consumption; LSTM neural network model: Input parameters: temperature gradient, CO concentration change rate, voltage fluctuation spectrum; Output: Thermal runaway risk probability value, based on 3000 thermal runaway case training, validation set AUC value of 0.93; Multi-level dynamic response strategy: Level 1 warning: triggers an audible and visual alarm, and the data is marked as "observation status"; Secondary intervention: When the charge and discharge current drops to 50% of the rated value, the liquid nitrogen micro-spray system is started; Level 3 protection: Cut off the experimental circuit within 0.1s, and start the aerosol fire extinguishing and photocatalytic oxidation modules.

7. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The dynamic control and execution module is divided into three submodules: current control, temperature control execution and safety protection; The current control submodule dynamically adjusts the charge and discharge current through a programmable power supply. The temperature control execution submodule adopts a liquid nitrogen micro-spray system to spray the battery tabs in a targeted manner through a nozzle with a diameter of 0.1 mm. The safety protection submodule integrates a high-voltage relay and an aerosol fire extinguishing device.

8. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The environmental treatment module consists of three sub-modules: a photocatalytic reactor, a chemical reaction path, and performance monitoring. The photocatalytic reactor sub-module has a built-in ultraviolet LED array and a titanium dioxide nanocatalyst coating. The ultraviolet light source has a wavelength of 365 nanometers, the catalyst particle size is 10 to 20 nanometers, and the specific surface area exceeds 200 square meters per gram. The chemical reaction path sub-module converts hydrogen fluoride into fluorine gas and water, and carbon monoxide into carbon dioxide. The performance monitoring sub-module ensures that the system resistance is less than 50 Pascals.

9. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The cloud collaboration and data management module includes three sub-modules: cloud platform, core functions and local storage.

10. The intelligent high-safety lithium battery failure test device and predictive control system according to claim 1, characterized in that: The modular expansion interface consists of two submodules: a universal battery interface and an expansion slot. The expansion slot submodule can accommodate up to sixteen groups of batteries for series or parallel testing.

Images