Sodium-ion battery thermal runaway monitoring and early warning method and system for electric energy storage

Through the deep feature extraction model of multi-scale convolutional network, the thermoelectric data feature vector of sodium ion batteries during thermal runaway is extracted, the thermoelectric database is established, and the temperature rise safety boundary of the battery cell is predicted, which solves the lack of dynamics of early warning thresholds in the existing technology. The problem of limited adjustment and early warning accuracy achieves high-accuracy thermal runaway warning.

CN119481396APending Publication Date: 2025-02-18STATE GRID ANHUI ELECTRIC POWER CO LTD ELECTRIC POWER SCI RES INST

Patent Information

Application Number
CN202411774565.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-05
Publication Date
2025-02-18

AI Technical Summary

Technical Problem

In the prior art, the early warning threshold lacks dynamic adjustment, resulting in false alarm or misreport of the early warning signal, and the early warning accuracy is limited, making it difficult to effectively adapt to the dynamic changes in battery performance.

Method used

The deep feature extraction model of a multi-scale convolutional network is adopted to obtain the sampling data of the battery cell to generate the temperature change curve of the battery surface, voltage change curve, capacity change curve, and temperature change curve of the battery with different heating rates, and extract the sodium ion battery in the heat. Internal characteristic parameters and external thermoelectric data during the out-of-control process are established, thermoelectric database is predicted, the temperature rise safety boundary of the battery cell is predicted, and thermal runaway warning is performed.

Benefits of technology

It can timely monitor whether the battery has thermal runaway, improve the accuracy and reliability of early warning, avoid false alarms or misreports of early warning signals, and enhance the real-time monitoring and early warning capabilities of battery thermal runaway.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119481396A_ABST
    Figure CN119481396A_ABST
Patent Text Reader

Abstract

The invention provides a thermal runaway monitoring and early warning method and a thermal runaway monitoring and early warning system for a sodium-ion battery for electric energy storage. The method comprises the following steps: acquiring sampling data of a battery cell; according to the sampling data, generating a change curve of parameters such as battery surface temperature, voltage and capacity along with time and a change curve of battery temperature along with time at different heating rates; internal characteristic parameters and external thermoelectric data of the sodium-ion battery in the thermal runaway process are extracted according to the two change curves, a depth feature extraction model based on a multi-scale convolutional network is established, and thermoelectric data feature vectors of the sodium-ion battery in different thermal runaway stages are extracted; and generating a thermoelectric database according to the thermoelectric data feature vector, and predicting a temperature rise safety boundary of the battery cell to realize early warning of thermal runaway of the battery cell. According to the invention, the technical problems that the early warning threshold lacks dynamic adjustment and the early warning precision is limited are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of new energy monitoring, and in particular to a thermal runaway monitoring and early warning method and system for sodium ion batteries for electric energy storage. Background Art

[0002] In recent years, with the vigorous development of the global clean energy industry and the rapid growth of the electric vehicle market, sodium-ion batteries for energy storage, as an emerging and highly promising high-energy-density battery technology, have gradually attracted great attention from the industry. With their unique advantages, sodium-ion batteries are regarded as an important complementary technology to lithium-ion batteries. They can flexibly adapt to and meet the performance requirements of a variety of application scenarios, from large-scale energy storage systems to portable electronic devices. In addition, the use of sodium-ion batteries effectively alleviates the current problem of lithium-ion batteries' over-reliance on key raw materials, which helps maintain the stability of the supply chain and promote the diversified development of battery technology.

[0003] However, similar to lithium-ion batteries, sodium-ion batteries, while pursuing high efficiency and long life, also face the potential threat of thermal runaway. Battery thermal runaway is an extremely dangerous condition that usually occurs under certain extreme conditions, such as when the battery is subjected to external short circuit, overcharging, or excessively high ambient temperature. At this time, the chemical reaction rate inside the battery will increase abnormally rapidly, releasing huge amounts of energy, causing the battery temperature to rise sharply, far beyond its normal operating temperature range, and once this process is started, it is difficult to effectively control. Thermal runaway not only seriously threatens the safe and stable operation of the battery system, but may also cause catastrophic consequences such as fire and explosion.

[0004] Currently, thermal runaway warning for batteries mainly relies on monitoring the temperature rise and voltage drop of the battery. Once the temperature or pressure inside the battery reaches a preset fixed threshold, the system triggers a warning signal. However, this method has the following problems:

[0005] (1) Lack of dynamic adjustment of warning thresholds: Traditional warning systems generally use static warning thresholds, which are often set based on the initial state of the battery when it leaves the factory. For example, the existing invention patent application document "Thermal runaway warning method and related equipment" with publication number CN119001515A, the existing method includes: obtaining thermal runaway warning parameters of the battery pack, the thermal runaway warning parameters include the internal resistance growth rate, pressure value, temperature value and temperature rise rate of each battery cell; the thermal runaway warning parameters also include the concentration of CO gas and dimethyl phosphate; based on the internal resistance growth rate, pressure value, temperature rise rate of each battery cell and the concentration of CO gas and dimethyl phosphate, respectively obtain the first warning value, the second warning value, the third warning value, the fourth warning value and the fifth warning value; when any one or more warning values ​​are greater than or equal to the corresponding preset threshold, a thermal runaway warning is issued; when any one or more warning values ​​are greater than or equal to the corresponding preset threshold and the temperature value is greater than the temperature threshold, a thermal runaway alarm is issued. However, in actual use, the performance of the battery in the above-mentioned existing solution will gradually change with the cumulative influence of multiple factors such as the number of charge and discharge cycles, ambient temperature fluctuations, and usage time. Therefore, the static warning threshold of the existing technology is difficult to accurately reflect the actual current state of the battery, which may lead to false alarms or missed alarms of the warning signal.

[0006] (2) Limited warning accuracy: For example, the existing invention patent application document "Lithium-ion Battery Safety Performance Grading Evaluation Method" with publication number CN119005681A, the existing method includes: obtaining thermal runaway warning parameters of the battery pack, the thermal runaway warning parameters include the internal resistance growth rate, pressure value, temperature value and temperature rise rate of each battery cell; the thermal runaway warning parameters also include the concentration of CO gas and dimethyl phosphate; based on the internal resistance growth rate, pressure value, temperature rise rate of each battery cell and the concentration of CO gas and dimethyl phosphate, respectively obtain a first warning value, a second warning value, a third warning value, a fourth warning value and a fifth warning value; when any one or more warning values ​​are greater than or equal to the corresponding preset threshold value, a thermal runaway warning is issued; when any one or more warning values ​​are greater than or equal to the corresponding preset threshold value and the temperature value is greater than the temperature threshold value, a thermal runaway alarm is issued. And the existing invention patent application document "Multi-time scale intelligent fire warning method and device for energy storage battery" with publication number CN119001481A, the existing method includes: charging and standing the battery at a first time scale, determining the first equilibrium potential and the second equilibrium potential of the battery, and determining the battery capacity of the battery; determining the degree of battery degradation according to the battery capacity; measuring the voltage of the battery during the charging and discharging process at a second time scale, determining the voltage sample data of the battery, and obtaining the isolation density of the voltage sample data, so as to determine the abnormal degree of battery operation; monitoring the environment in which the battery is located at a third time scale, determining the degree of thermal runaway risk of the environment in which the battery is located; determining the alarm strategy for fire warning according to the degree of degradation, abnormality and thermal runaway risk. Because in the aforementioned existing schemes, the warning threshold lacks the necessary dynamic adjustment mechanism, it is difficult for the traditional warning system to effectively adapt to the dynamic changes in battery performance, and thus it is difficult to provide accurate and reliable warning information.

[0007] In summary, the existing technology has technical problems such as lack of dynamic adjustment of warning threshold and limited warning accuracy. Summary of the invention

[0008] The technical problem to be solved by the present invention is: how to solve the technical problems of lack of dynamic adjustment of the warning threshold and limited warning accuracy in the prior art.

[0009] The present invention adopts the following technical solutions to solve the above technical problems: A thermal runaway monitoring and early warning method for sodium ion batteries for power energy storage includes:

[0010] S1. Under preset conditions, obtaining sampling data of the battery cell, wherein the sampling data includes: temperature data and operating condition data;

[0011] S2. Generate a battery surface temperature change curve, a voltage change curve, a capacity change curve, and a battery temperature change curve at different heating rates according to the sampling data;

[0012] S3. According to the battery surface temperature change curve, voltage change curve, capacity change curve, and battery temperature change curves at different heating rates, the internal characteristic parameters and external thermoelectric data of the sodium-ion battery during thermal runaway are extracted. Based on a multi-scale convolutional network, a deep feature extraction model is established to extract the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway.

[0013] S4. Generate a thermoelectric database based on the thermoelectric data feature vector, predict the temperature rise safety boundary of the battery cell based on the thermoelectric database, and perform thermal runaway warning of the battery cell based on it.

[0014] The present invention cooperates with the peak-shaving and frequency-regulating dispatching requirements of the power grid for the energy storage system, takes into account the state of the battery cells that are thermally runaway due to electrical abuse and thermal abuse under the corresponding energy storage conditions, and uses a deep feature extraction model of a multi-scale convolutional network to extract the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway. It can timely monitor whether the battery has thermal runaway, and thus can promptly alarm when the battery has thermal runaway, so that the user can take corresponding measures against the thermal runaway of the battery.

[0015] The present invention predicts the temperature rise safety boundary of the battery cell based on the thermoelectric database. Compared with the traditional static warning threshold, it can more accurately reflect the current actual state of the battery and avoid false alarms or missed alarms of the warning signal.

[0016] In a more specific technical solution, in S1, the temperature data includes: electrical abuse temperature data, thermal abuse temperature data; the operating condition data includes: electrical abuse operating condition data, thermal abuse operating condition data.

[0017] In a more specific technical solution, in S1, the preset conditions include: thermal runaway conditions caused by electrical abuse and thermal runaway conditions caused by thermal abuse.

[0018] In a more specific technical solution, in S2, based on the acquired temperature data, the temperature rise rate β(t) at different times t and different states of charge SOC is calculated; recorded as: β1, β2, β3, ..., βn;

[0019] The temperature rise rate β(t) is fitted to obtain a temperature rise rate fitting curve β(x), where x represents the SOC value.

[0020] In a more specific technical solution, the temperature rise rate β(t) is obtained using the following logic:

[0021]

[0022] Wherein, t1 and t2 represent different moments, and T1 and T2 represent the temperature of the battery cell at different SOCs.

[0023] In a more specific technical solution, in S3, the deep feature extraction model includes: no less than 2 difference scale parallel convolutions.

[0024] In a more specific technical solution, in S3, the sampled data is used as the input vector X, and a feature fusion operation is performed on the input vector to obtain a fused feature; the fused feature is processed in sequence using a pooling layer, a normalization layer, and an activation layer to obtain a feature Z.

[0025] In a more specific technical solution, the following logic is used to perform a maximum pooling operation to obtain feature Z:

[0026] Z=f(R(Xmax))

[0027] Where f is the activation function; Xma x is the maximum pooling of the input phase; R is the normalization processing.

[0028] In a more specific technical solution, in S3, the component weights are extracted using the fully connected layer and adaptive function, the components are multiplied and added with the convolution kernels of different scales, the weights of each convolution layer are dynamically selected, and the output phasor Y is processed:

[0029]

[0030] Traditional thermal runaway detection technology relies on the temperature rise and voltage drop generated when thermal runaway occurs, but by this stage, thermal runaway is completely uncontrollable and the signal can only be used as an alarm for evacuation. Although there are some neural network models that can predict thermal runaway, these existing methods can only make predictions under well-defined laboratory conditions and are less practical under complex battery operating conditions.

[0031] The multi-scale convolutional neural network adopted in the present invention adds multiple scale feature extraction on the basis of the traditional convolutional neural network, that is, by using convolution kernels of different sizes to extract the multi-scale features of the input data; and multiple parallel convolution paths are set in the network, each path uses a convolution kernel of a different size; the features of different paths are fused by methods such as splicing or weighted summation to integrate information of multiple scales, and the fused features can be better used for subsequent classification or regression tasks. Considering the state of the battery cell in thermal runaway due to electrical abuse and thermal abuse under the corresponding energy storage conditions, multiple parallel convolution paths are constructed, each path uses a different convolution kernel size and step size, which is more in line with the peak-shaving and frequency-regulating scheduling requirements of the power grid for the energy storage system, so as to capture the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway.

[0032] The multi-scale convolutional network deep feature extraction model adopted by the present invention can enhance the expression ability of the model by using multiple parallel convolutions of different scales, effectively improving the ability of multi-scale feature extraction, calculation speed and detection accuracy. In addition, the model also has a certain flexibility and can be combined with other model architectures to adapt to different warning tasks.

[0033] In a more specific technical solution, the thermal runaway monitoring and early warning system for sodium-ion batteries for power energy storage includes:

[0034] The battery cell data acquisition module is used to obtain the sampling data of the battery cell under preset conditions, wherein the sampling data includes: temperature data and operating condition data;

[0035] A curve fitting module is used to generate a battery surface temperature change curve, a voltage change curve, a capacity change curve, and a battery temperature change curve at different heating rates according to the sampled data. The curve fitting module is connected to the battery cell data acquisition module.

[0036] A feature extraction module is used to extract internal characteristic parameters and external thermoelectric data of the sodium-ion battery during thermal runaway according to the battery surface temperature change curve, voltage change curve, capacity change curve, and battery temperature change curves at different heating rates. A deep feature extraction model is established based on a multi-scale convolutional network to extract thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway. The feature extraction module is connected to the curve fitting module.

[0037] The thermal runaway warning module is used to generate a thermoelectric database based on the thermoelectric data feature vector, predict the temperature rise safety boundary of the battery cell based on the thermoelectric database, and perform thermal runaway warning of the battery cell based on this. The thermal runaway warning module is connected to the feature extraction module.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] The present invention cooperates with the peak-shaving and frequency-regulating dispatching requirements of the power grid for the energy storage system, takes into account the state of the battery cells that are thermally runaway due to electrical abuse and thermal abuse under the corresponding energy storage conditions, and uses a deep feature extraction model of a multi-scale convolutional network to extract the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway. It can timely monitor whether the battery has thermal runaway, and thus can promptly alarm when the battery has thermal runaway, so that the user can take corresponding measures against the thermal runaway of the battery.

[0040] The present invention predicts the temperature rise safety boundary of the battery cell based on the thermoelectric database. Compared with the traditional static warning threshold, it can more accurately reflect the current actual state of the battery and avoid false alarms or missed alarms of the warning signal.

[0041] The multi-scale convolutional network deep feature extraction model adopted in the present invention uses multiple parallel convolutions of different scales, which effectively improves the ability and calculation speed of multi-scale feature extraction.

[0042] The present invention solves the technical problems existing in the prior art that the warning threshold lacks dynamic adjustment and the warning accuracy is limited. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a schematic diagram of the basic steps of the thermal runaway monitoring and early warning method for sodium-ion batteries for electric energy storage according to Example 1 of the present invention;

[0044] Figure 2 Schematic diagram of the deep feature extraction model of the multi-scale convolutional network of Example 1 of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Example 1

[0047] like Figure 1 As shown, the thermal runaway monitoring and early warning method for sodium ion batteries for power energy storage provided by the present invention includes the following basic steps:

[0048] S1. Obtaining the temperature data and electrical abuse condition data of the battery cell;

[0049] In this embodiment, the sampled data include but are not limited to: battery surface temperature, voltage, and capacity under thermal runaway caused by electrical abuse, and heating rate, battery temperature, and voltage under thermal runaway caused by thermal abuse;

[0050] In the present embodiment, the conditions for thermal runaway caused by electrical abuse include but are not limited to: thermal runaway of the battery cell caused by triggering factors, and the triggering factors include but are not limited to: overcharging, over-discharging and short circuit; in the present embodiment, the preset operations of the conditions for thermal runaway caused by thermal abuse include: heating experiment, adiabatic thermal runaway experiment; for example: an adiabatic accelerating calorimeter can be used to conduct heating experiments and adiabatic thermal runaway experiments.

[0051] S2. Generate curves of battery surface temperature, voltage, capacity and other parameters changing with time and curves of battery temperature changing with time at different heating rates based on the acquired sampling data;

[0052] In this embodiment, based on the acquired temperature data, the temperature rise rate β(t) at different times t and different SOCs is calculated; the temperature rise rate of the battery cell at different SOCs is calculated and marked as: β1, β2, β3, ..., βn, and the temperature rise rate fitting curve β(x) is fitted, where x represents the SOC value.

[0053] In this embodiment, the calculation formula of the temperature rise rate β(t) is as follows:

[0054]

[0055] Wherein, t1 and t2 represent different moments respectively, and T1 and T2 represent the temperature of the battery cell corresponding to different SOCs respectively.

[0056] S3. Extract the internal characteristic parameters and external thermoelectric data of the sodium-ion battery during thermal runaway according to the two change curves, establish a deep feature extraction model based on a multi-scale convolutional network, and extract the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway;

[0057] In this embodiment, the deep feature extraction model of the multi-scale convolutional network uses multiple parallel convolutions of different scales, effectively improving the ability and calculation speed of multi-scale feature extraction.

[0058] like Figure 2 As shown, in this embodiment, the sampled data is used as the input vector X feature fusion, and then passes through the pooling layer, normalization layer, and activation layer to obtain the feature Z. The calculation formula is as follows:

[0059] Z=f(R(Xmax))

[0060] Among them, f is the activation function; Xma x is the maximum pooling of the input phase; R is the normalization processing.

[0061] After the fully connected layer and adaptive function, the four component weights are extracted, the four components are multiplied by the convolution kernels of different scales and then added, and the weights of each convolution layer are dynamically selected. The output phasor Y expression is as follows:

[0062]

[0063] S4. Generate a thermoelectric database based on the thermoelectric data feature vector, and predict the temperature rise safety boundary of the battery cell to achieve early warning of thermal runaway of the battery cell.

[0064] In summary, the present invention cooperates with the peak-shaving and frequency-regulating dispatching requirements of the power grid for the energy storage system, takes into account the state of the battery cell that is thermally runaway due to electrical abuse and thermal abuse under the corresponding energy storage conditions, and uses a deep feature extraction model of a multi-scale convolutional network to extract the thermoelectric data feature vectors of the sodium-ion battery at different stages of thermal runaway. It can timely monitor whether the battery has thermal runaway, and thus can promptly alarm when the battery has thermal runaway, so that the user can take corresponding measures against the thermal runaway of the battery.

[0065] The present invention predicts the temperature rise safety boundary of the battery cell based on the thermoelectric database. Compared with the traditional static warning threshold, it can more accurately reflect the current actual state of the battery and avoid false alarms or missed alarms of the warning signal.

[0066] The multi-scale convolutional network deep feature extraction model adopted in the present invention uses multiple parallel convolutions of different scales, which effectively improves the ability and calculation speed of multi-scale feature extraction.

[0067] The present invention solves the technical problems existing in the prior art that the warning threshold lacks dynamic adjustment and the warning accuracy is limited.

[0068] 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 the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. 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 embodiments of the present invention.

Claims

1. A thermal runaway monitoring and early warning method for sodium-ion batteries for power energy storage, characterized in that: The method comprises: S1. Under preset conditions, obtaining sampling data of the battery cell, wherein the sampling data includes: temperature data and operating condition data; S2. Generate a battery surface temperature change curve, a voltage change curve, a capacity change curve, and a battery temperature change curve at different heating rates according to the sampled data; S3, extracting internal characteristic parameters and external thermoelectric data of the sodium ion battery during thermal runaway according to the battery surface temperature change curve, the voltage change curve, the capacity change curve, and the battery temperature change curves at different heating rates, and establishing a deep feature extraction model based on a multi-scale convolutional network to extract thermoelectric data feature vectors of the sodium ion battery at different stages of thermal runaway; S4. Generate a thermoelectric database according to the thermoelectric data feature vector, predict the temperature rise safety boundary of the battery cell according to the thermoelectric database, and perform a thermal runaway warning of the battery cell accordingly.

2. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S1, the temperature data includes: electrical abuse temperature data and thermal abuse temperature data; the operating condition data includes: electrical abuse operating condition data and thermal abuse operating condition data.

3. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S1, the preset conditions include: thermal runaway conditions caused by electrical abuse and thermal runaway conditions caused by thermal abuse.

4. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S2, the temperature rise rate β(t) at different times t and different states of charge SOC is calculated based on the acquired temperature data; it is recorded as: β1, β2, β3, ..., β n ; The temperature rise rate β(t) is fitted to obtain a temperature rise rate fitting curve β(x), wherein x represents the SOC value.

5. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 4 is characterized in that: The temperature rise rate β(t) is obtained using the following logic: In the formula, t1 and t2 represent different moments, and T1 and T2 represent the temperature of the battery cell corresponding to different SOCs.

6. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S3, the deep feature extraction model includes: no less than 2 difference scale parallel convolutions.

7. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S3, the sampled data is used as an input vector X, and a feature fusion operation is performed on the input vector to obtain a fused feature; the fused feature is processed in sequence using a pooling layer, a normalization layer, and an activation layer to obtain a feature Z.

8. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 7 is characterized in that: Using the following logic, perform the maximum pooling operation to obtain feature Z: Z=f(R(XXmax)) Where f is the activation function; Xma x is the maximum pooling of the input phase; R is the normalization processing.

9. The thermal runaway monitoring and early warning method for sodium ion batteries for electric energy storage according to claim 1 is characterized in that: In S3, the component weights are extracted using the fully connected layer and adaptive function, the components are multiplied and added with the convolution kernels of different scales, the weights of each convolution layer are dynamically selected, and the output phasor Y is obtained by processing:

10. A thermal runaway monitoring and early warning system for sodium-ion batteries for power storage, characterized in that: The system comprises: A cell data acquisition module is used to obtain sampled data of the cell under preset conditions, wherein the sampled data includes: temperature data and operating condition data; A curve fitting module, used for generating a battery surface temperature change curve, a voltage change curve, a capacity change curve, and a battery temperature change curve at different heating rates according to the sampled data, wherein the curve fitting module is connected to the battery cell data acquisition module; A feature extraction module is used to extract internal characteristic parameters and external thermoelectric data of the sodium ion battery in the thermal runaway process according to the battery surface temperature change curve, the voltage change curve, the capacity change curve, and the battery temperature change curves at different heating rates, and to establish a deep feature extraction model based on a multi-scale convolutional network to extract the thermoelectric data feature vector of the sodium ion battery at different stages of thermal runaway. The feature extraction module is connected to the curve fitting module; A thermal runaway warning module is used to generate a thermoelectric database according to the thermoelectric data feature vector, predict the temperature rise safety boundary of the battery cell according to the thermoelectric database, and perform thermal runaway warning of the battery cell accordingly. The thermal runaway warning module is connected to the feature extraction module.

Citation Information

Patent Citations

  • Multi-time-scale intelligent fire early warning method and device for energy storage battery

    CN119001481A

  • Thermal runaway early warning method and related equipment thereof

    CN119001515A

  • Lithium ion battery safety performance grading evaluation method

    CN119005681A

Cited By

  • Method, device and system for testing thermal diffusion state of battery module and medium

    CN120629983A

  • A method, device, system and medium for testing a thermal diffusion state of a battery module

    CN120629983B