Online battery thermal runaway early warning method and system based on eis parameter extraction

Abstract

The application discloses an online battery thermal runaway early warning method and system based on electrochemical impedance spectrum (EIS) parameter extraction. The method comprises: an experimental device and method for EIS key parameter extraction for battery thermal runaway early warning, the experimental device comprises: a battery, a chamber capable of providing an adiabatic environment, a battery testing device, an EIS online acquisition device and a host computer, the parameter extraction method comprises a feature importance analysis algorithm and a sensitivity analysis algorithm; the thermal runaway early warning system comprises: a battery, an EIS online acquisition device and a host computer; the thermal runaway early warning method issues a first-level thermal runaway early warning when the battery temperature exceeds the upper limit of a thermal management threshold, issues a second-level thermal runaway early warning in a battery self-heating temperature interval, and issues a third-level thermal runaway early warning when the battery voltage drops according to the extracted thermal runaway early warning key parameters.

Description

Technical Field

[0001] This invention belongs to the field of lithium-ion battery safety and relates to an online battery thermal runaway early warning method and system based on electrochemical impedance spectroscopy (EIS) parameter extraction. Background Technology

[0002] With the continuous development of new energy technologies and the increasing maturity of battery manufacturing processes, batteries are beginning to be widely used in new energy vehicles and energy storage power stations. Battery safety has become a major bottleneck restricting the development of new energy technologies, and a large number of thermal runaway early warning technologies for power batteries and energy storage batteries are also constantly being developed. Traditional thermal runaway early warning technologies rely on monitoring the external physical parameters of the battery, such as voltage, current, and temperature, or detecting the organic and inorganic gases released after the safety valve opens before the battery thermal runaway. Due to limitations in size and economic cost, the number of sensors in the battery module box is limited, making it impossible to effectively monitor each cell. Moreover, due to the delayed effects of thermal and gas diffusion, the response time of temperature and gas sensors is relatively lagging. In order to improve the dimensionality of early warning parameters, increase the success rate and reliability of early warning, and advance the early warning time, it is crucial to introduce a parameter that can reflect the internal state of the battery in real time. EIS, as a means of reflecting the internal state of the battery, has been extensively studied. EIS can reflect the internal state of charge (SOC), state of aging (SOH), and internal temperature of the battery. Currently, some portable, high-precision, small-scale EIS monitoring modules are already being used in electric vehicles. These modules can monitor the electrochemical impedance parameters of the measured battery at different frequencies, including the real part Z' and the imaginary part Z” of the impedance, thereby obtaining the absolute value of the impedance |Z| and the phase angle θ. Extracting key parameters for thermal runaway early warning from a large number of parameters and utilizing these parameters for thermal runaway early warning is particularly important. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides an online battery thermal runaway early warning method and system based on EIS parameter extraction. An experimental setup for extracting key EIS parameters for thermal runaway early warning was constructed, successfully extracting effective parameters for early warning. Then, a battery thermal runaway early warning system was built using the experimentally extracted key EIS parameters. A battery thermal runaway early warning method suitable for different connection methods was developed, and successful early warning of battery thermal runaway was achieved.

[0004] To achieve the above objectives, the present invention employs the following technical solution:

[0005] An online battery thermal runaway early warning method based on EIS parameter extraction includes the following steps:

[0006] Step 1: Obtain the electrochemical impedance spectroscopy of the battery at different temperatures and under different charging states;

[0007] Step 2: Use the RReliefF algorithm and sensitivity analysis to extract key EIS parameters, which include: the real part Z' and imaginary part Z' of the impedance, the absolute value of the impedance |Z|, and the phase angle θ.

[0008] Step 3: Determine the thermal runaway early warning nodes corresponding to the key EIS parameters;

[0009] Step 4: Use the extracted EIS key parameters and nodes to perform thermal runaway early warning for batteries with different connection methods.

[0010] Furthermore, in step 2, the RReliefF algorithm is used to perform importance analysis on the large number of EIS key parameters obtained, and box plots are used to perform sensitivity analysis on the data features; based on the importance analysis, feature parameters with good temperature prediction effect are selected, and based on the sensitivity analysis, key parameters for thermal runaway early warning that are highly sensitive to temperature and low sensitive to SOC are found.

[0011] Furthermore, in step 4, the battery connection method includes individual battery cells, series modules, and parallel modules.

[0012] Furthermore, step 4 includes: issuing a first-level thermal runaway warning when the battery temperature exceeds the upper limit of the optimal thermal management temperature threshold, issuing a second-level thermal runaway warning within the battery's self-generated heat temperature range, and issuing a third-level thermal runaway warning when the battery voltage drops.

[0013] Furthermore, for ternary lithium-ion batteries, the upper limit of the optimal temperature threshold for thermal management is 45°C, and the range of battery self-generated heat temperature is 80°C to 105°C.

[0014] Furthermore, for series modules, when the absolute value of impedance |Z| or the phase angle θ exceeds the warning threshold, a first-level thermal runaway warning is issued; if it does not exceed the threshold, the risk of thermal runaway is low. Based on the first-level warning, if the slope of the absolute value of impedance |Z| changes from negative to positive and continues to increase, or if the battery phase angle θ begins to fluctuate and Δθ < 0, a second-level warning is issued. Based on the second-level warning, if the order of magnitude of the absolute value of impedance |Z| increases, a third-level thermal runaway warning is issued.

[0015] Furthermore, for parallel module batteries, if the absolute value of the impedance |Z| or the phase angle θ of the parallel module exceeds the warning threshold, a first-level thermal runaway warning is issued; if it does not exceed the threshold, the risk of thermal runaway is low. Based on the first-level warning, if the slope of the battery phase angle θ changes from positive to negative and continues to decrease, a second-level thermal runaway warning is issued. Based on the second-level warning, if the increase in the absolute value of the module impedance |Z| exceeds the warning threshold or the slope of the phase angle θ changes from negative to positive and begins to increase, a third-level thermal runaway warning is issued.

[0016] This invention also provides a thermal runaway early warning system based on an online battery thermal runaway early warning method using EIS parameter extraction. The system includes a battery, a chamber providing an adiabatic environment, a battery testing device, an online electrochemical impedance spectroscopy monitoring device, and a host computer. The system adjusts the battery temperature using the chamber to achieve thermal equilibrium within the battery. The battery testing device charges and discharges the battery to change its state of charge. The online electrochemical impedance spectroscopy monitoring device and the host computer collect and analyze the changes in battery temperature and electrochemical parameters when the temperature stabilizes and under different states of charge. The wall temperature of the chamber providing the adiabatic environment is controlled by programming on the host computer. The chamber includes a temperature acquisition device for monitoring the wall temperature and a temperature acquisition device for monitoring the sample temperature.

[0017] Furthermore, the chamber that provides an insulated environment is an adiabatic accelerating calorimeter, which contains a sealed insulated tank.

[0018] Furthermore, the online electrochemical impedance spectroscopy monitoring device collects data under the dynamic conditions of battery charging and discharging, and simultaneously collects data from multiple channels via a daisy-chain connection. It measures the real part Z' and imaginary part Z'” of the battery impedance at different frequencies, and then calculates the absolute impedance value |Z| and the phase angle θ. |Z| and θ are calculated using the following formulas:

[0019]

[0020] θ = arctan -1 (-Z" / Z').

[0021] Furthermore, the online electrochemical impedance spectroscopy monitoring device includes a battery management chip, a PC main control board, and an FPC slave control board.

[0022] The beneficial effects of this invention are:

[0023] This invention extracts key parameters for thermal runaway early warning based on battery EIS data. It can filter out EIS parameters that are sensitive to battery temperature but not to State of Charge (SOC), and then use these selected key thermal runaway early warning parameters to provide thermal runaway early warning for batteries with different connection methods. This method solves the problems of arbitrary selection of warning frequency and parameters in previous methods, exhibiting better robustness and specificity. It effectively compensates for the shortcomings of warning delays from voltage, temperature, and gas sensors, providing a new method for early warning in vehicle BMS. Attached Figure Description

[0024] Figure 1 A schematic diagram of the experimental setup for extracting key EIS parameters for battery thermal runaway early warning;

[0025] Figure 2This is a schematic diagram of a thermal runaway early warning system.

[0026] Figure 3 Here is a flowchart of the thermal runaway early warning method;

[0027] Figure 4 The Nyquist plots are of the batteries at different temperatures and SOCs in Example 1;

[0028] Figure 5 The curves showing the variation of the key thermal runaway parameters |Z| and θ extracted in Example 1 are shown. Figure 5 (a) and (b) represent the curves showing the changes in the absolute value of the AC impedance |Z| and the phase angle θ of the battery at 1 kHz during the experiment to extract key parameters of thermal runaway.

[0029] Figure 6 This is a graph showing the voltage and temperature curves during the thermal runaway early warning experiment of the parallel module in Example 1;

[0030] Figure 7 Figure 1 shows the curves of |Z| and θ during the thermal runaway early warning experiment of the parallel module in Example 1; where Figure 1(a) shows the curve of the absolute value of the AC impedance |Z| of the battery at 1kHz during the heating process, and Figure 2(b) shows the curve of the phase angle θ.

[0031] Figure 8 This is a flowchart of the online battery thermal runaway early warning method based on EIS parameter extraction according to the present invention.

[0032] In the diagram, 1-battery testing device, 2-battery, 3-adiabatic accelerated calorimeter, 4-sealed insulating tank, 5-online electrochemical impedance spectroscopy monitoring equipment, 6-battery management chip, 7-host computer, 8-100W heating rod, 9-battery fixing mold. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0034] Figure 1 The diagram shows the experimental setup for extracting key EIS parameters for battery thermal runaway early warning. Figure 1As shown, the experimental setup includes a battery 2, a chamber that can provide a thermally insulated environment, a battery testing device 1, an online electrochemical impedance spectroscopy monitoring device 5, and a host computer 7. The battery temperature is adjusted by a chamber that provides an adiabatic environment, allowing the internal heat of the battery to reach equilibrium. The battery testing device 1 charges and discharges the battery 2, changing its state of charge (SOC). The online electrochemical impedance spectroscopy monitoring device 5 and the host computer 7 collect and analyze the temperature and electrochemical parameter changes of the battery 2 when the temperature stabilizes and at different SOCs. The wall temperature of the chamber, which provides an adiabatic environment, can be controlled by programming on the host computer 7. The chamber contains temperature acquisition devices for monitoring both the wall temperature and the sample temperature. The key EIS parameters include the real part Z' and the imaginary part Z'' of the impedance, the absolute value of the impedance |Z|, and the phase angle θ. The online electrochemical impedance spectroscopy monitoring device 5 can collect data under dynamic charging and discharging conditions. Through a daisy-chain connection, it can simultaneously collect data from multiple channels, measuring the real part Z' and the imaginary part Z'' of the battery impedance at different frequencies, and then calculating the absolute value of the impedance |Z| and the phase angle θ. |Z| and θ can be calculated using the following formulas:

[0035]

[0036] θ = arctan -1 (-Z" / Z')

[0037] The method for extracting key EIS parameters for thermal runaway early warning includes: a feature importance analysis algorithm and a sensitivity analysis algorithm; the feature importance analysis algorithm is used to select feature parameters that have a good effect on temperature prediction, and the sensitivity analysis algorithm is used to find key thermal runaway early warning parameters that are highly sensitive to temperature and low sensitive to SOC.

[0038] Figure 2 The image shows a thermal runaway early warning system. (As shown) Figure 2 As shown, the thermal runaway early warning system includes: a battery 2, an online electrochemical impedance spectroscopy monitoring device 5, and a host computer 7; the battery 2 includes: a single battery cell, a series module, and a parallel module; the online electrochemical impedance spectroscopy monitoring device 5 is the same as the experimental device for extracting key parameters of thermal runaway early warning EIS; the host computer 7 can receive, process, and visualize the electrochemical signals collected by the online electrochemical impedance spectroscopy monitoring device 5.

[0039] Preferably, the battery 2 is heated by a 100W heating rod to trigger thermal runaway, and the size of the heating rod is the same as that of a single battery in the battery 2.

[0040] Preferably, the battery 2 is fixed by a battery fixing device 9, and the mold consists of two perforated cover plates and stud nuts, which serve to fix the battery 2.

[0041] Figure 3 The diagram shown is a flowchart of a thermal runaway early warning method. Figure 3 As shown, the key EIS parameters required for thermal runaway early warning are the real part of impedance Z', the imaginary part of impedance Z'", the absolute value of impedance |Z|, and the phase angle θ. The thermal runaway early warning method includes: using the extracted key EIS parameters for thermal runaway early warning. Based on the extracted key parameters, for batteries with different connection methods, a Level 1 thermal runaway early warning is issued when the battery temperature exceeds the upper limit of the optimal thermal management temperature threshold; a Level 2 thermal runaway early warning is issued within the battery's self-generating heat temperature range; and a Level 3 thermal runaway early warning is issued when the battery voltage drops.

[0042] Preferably, the detailed steps for thermal runaway early warning are as follows: First, determine the connection method of the battery module. If battery 2 is a single cell or connected in series, when the absolute value of impedance |Z| or the phase angle θ exceeds the warning threshold, a first-level thermal runaway warning is issued. If it does not exceed the threshold, the risk of thermal runaway is low. Based on the first-level warning, if the slope of the absolute value of impedance |Z| changes from negative to positive and continues to increase, or if the battery phase angle θ starts to fluctuate and Δθ<0, a second-level thermal runaway warning is issued. Based on the second-level warning, if the order of magnitude of the absolute value of impedance |Z| increases, a third-level thermal runaway warning is issued.

[0043] Preferably, if batteries 2 are connected in parallel, a first-level thermal runaway warning is issued when the absolute value of the impedance |Z| or the phase angle θ of the parallel module exceeds the warning threshold; if it does not exceed the threshold, the risk of thermal runaway is low. Based on the first-level warning, if the slope of the battery phase angle θ changes from positive to negative and continues to decrease, a second-level thermal runaway warning is issued. Based on the second-level warning, if the increase in the absolute value of the module's impedance |Z| exceeds the warning threshold or the slope of the phase angle θ changes from negative to positive and begins to increase, a third-level thermal runaway warning is issued.

[0044] Preferably, the chamber that provides an adiabatic environment is an adiabatic accelerating calorimeter 3, which is equipped with a sealed adiabatic tank 4, with a temperature rise rate detection range of 0.02 K / min, a reaction initiation temperature of 30°C, and a step temperature rise of 5°C.

[0045] Preferably, the battery testing device 1 has a voltage range of 0-5V, a current range of 0-12A, an accuracy of 0.05%, and a sampling time of 1s.

[0046] Preferably, the online electrochemical impedance spectroscopy monitoring device 5 includes a battery management chip 6, a PC main control board and an FPC slave control board, with a voltage measurement accuracy of ±2mV, a cell measurement voltage range of 1.9V to 5.5V, and should be able to monitor the electrochemical impedance spectra of low-impedance cells. The daisy-chain communication rate is 1Mbps, the operating temperature range is -40℃ to 105℃, and the impedance monitoring frequency range is 7.5mHz to 7.8kHz.

[0047] Preferably, the positive and negative electrodes of the battery 2 and the online electrochemical impedance spectroscopy monitoring device 5 are connected by welding, and the length of the wire does not exceed 10cm.

[0048] Preferably, the RReliefF algorithm is used to perform importance analysis on the large amount of EIS data obtained, and box plots are used to perform sensitivity analysis on the data features.

[0049] Preferably, the RReliefF algorithm is a feature selection algorithm for regression problems. It extracts the key features most relevant to the dependent variable from a large number of features in the sample and ranks them by importance. RReliefF penalizes giving different predictor values ​​to neighbors with the same response value and rewards giving different predictor values ​​to neighbors with different response values. The principle of the RReliefF algorithm is as follows:

[0050] The algorithm's inputs are: feature matrix X and response vector y; the algorithm's output is: the predicted feature F for different feature vector pairs. j The contribution ranking W is determined by the following steps:

[0051] Step 1: Set the weight W dy W dj W dy∧dj and W j Set to 0, where W dy W represents the weights for different response values ​​y. dj W represents the weights of different predicted values ​​Fj. dy∧dj These are the weights of different response values ​​y and different predicted values ​​Fj;

[0052] Step 2: For i:=1 to m, randomly select an observation x. r , where i is the number of iterations and m is the number of rows of data;

[0053] Step 3: Locate the distance observation x r The k nearest neighbors;

[0054] Step 4: For the nearest neighbor x q Update all intermediate weights:

[0055]

[0056]

[0057] W dy∧dj i =W dy∧dj i-1 +Δ y (x r ,x q )·Δ j (x r ,x q )·d rq ;

[0058] in

[0059] Where d rq Represents the distance between two sets of data. Represents the normalized distance, rank(r,q) represents the distance ranking of the q-th feature vector among the r feature vectors, sigma represents the normalization factor, and Δy(x r ,x q ) represents the observation x under the continuous variable y. r and observation x q The difference between them, Δ j (x r ,x q ) represents the predictor variable F j Next x r and x q The difference between them;

[0060] Step 5: Calculate the weights W of the prediction vector. j ,

[0061] Preferably, for ternary lithium-ion batteries, the upper limit of the optimal temperature threshold for thermal management is 45°C, and the battery self-heating temperature range is 80°C to 105°C.

[0062] like Figure 8 As shown, the online battery thermal runaway early warning method based on EIS parameter extraction of the present invention specifically includes the following steps:

[0063] Step 1: Obtain the electrochemical impedance spectroscopy of the battery at different temperatures and under different charging states;

[0064] Step 2: Use the RReliefF algorithm and sensitivity analysis to extract key EIS parameters, which include: the real part Z' and imaginary part Z' of the impedance, the absolute value of the impedance |Z|, and the phase angle θ.

[0065] Step 3: Determine the thermal runaway early warning nodes corresponding to the key EIS parameters;

[0066] Step 4: Use the extracted EIS key parameters and nodes to perform thermal runaway early warning for batteries with different connection methods.

[0067] Example 1

[0068] Taking a cylindrical ternary lithium-ion battery as an example, firstly, using Figure 1 The experimental setup shown extracts key EIS parameters for thermal runaway early warning of the lithium-ion battery; then, the 1×3 parallel lithium-ion battery module is heated with a heating power of 100W, and thermal runaway early warning is performed. The invention is described in detail below. The method is established in twelve steps:

[0069] Step 1: Solder the positive and negative electrodes of the 100% SOC lithium-ion battery to the four pins of the online electrochemical impedance spectroscopy monitoring device using wires, and connect the online electrochemical impedance spectroscopy monitoring device to the host computer using a data transmission line.

[0070] Step 2: Place the thermocouple in the middle of the lithium-ion battery and fix it in place with high-temperature adhesive insulating tape.

[0071] Preferably, the thermocouple is a type K thermocouple with a temperature measurement range of 0 to 1000℃ and an error of ±2.5℃.

[0072] Step 3, insert battery 2 Figure 1 In the sealed and insulated container 4 of the adiabatic accelerated calorimeter 3 shown, the experimental parameters are set on the host computer 7, the initial temperature of the battery 2 is controlled to be 25°C, the temperature rise step is 5°C, the waiting time is 30 minutes, and then the experiment is started.

[0073] Step 4: Open the host computer data visualization software interface. Whenever the battery temperature rises by 5°C, use the online electrochemical impedance spectroscopy monitoring device 5 to record the AC impedance spectrum of the battery in the range of 1Hz-7.8kHz. Figure 4 The left figure shows the changes in electrochemical impedance spectroscopy under different temperature gradients before the voltage drop of battery 2. The experimental data shows that the AC impedance of the battery first decreases and then increases with increasing temperature.

[0074] Step 5: Solder the positive and negative electrodes of another 100% SOC lithium-ion battery of the same model to the four pins of the online electrochemical impedance spectroscopy monitoring device 5 using wires, and connect the online electrochemical impedance spectroscopy monitoring device 5 to the host computer 7 using a data transmission line.

[0075] Step six: Connect the lithium-ion battery from step five to the battery testing device 1, set the discharge steps in the host computer 7, control the SOC decrease step size of the battery to 5%, and then start the experiment.

[0076] Step 7: Open the host computer data visualization software interface. Whenever the battery SOC drops by 5%, use the electrochemical impedance spectroscopy online monitoring device 5 to record the AC impedance spectrum of the battery in the range of 1Hz-7.8kHz. Figure 4 The right figure shows the changes in the electrochemical impedance spectrum of battery 2 under different SOC gradients. The experimental data show that the AC impedance of the battery first decreases and then increases as the SOC decreases.

[0077] Step 8: Use the RReliefF algorithm to perform importance analysis on the large amount of EIS data obtained, and use box plots to perform sensitivity analysis on the data features.

[0078] Preferably, the key parameters for thermal runaway early warning obtained through the above experiments and analysis are the absolute value of AC impedance |Z| and the phase angle θ at a frequency of 1 kHz.

[0079] Step 9: Using the thermal runaway parameters of the individual cells obtained by the adiabatic accelerated calorimeter 3, determine the key points where the battery temperature reaches the upper limit of the optimal temperature threshold for thermal management, the temperature range of battery self-generated heat, and the order of magnitude increase in battery impedance. Figure 5 The changes in key parameters |Z| and θ for early warning of battery thermal runaway when the adiabatic accelerated calorimeter 3 (ARC) and the online electrochemical impedance spectroscopy monitoring device 5 are used together.

[0080] Step 10: Weld a 1×3 parallel module using this type of battery. The module uses... Figure 2 The mold shown is fixed, and the positive and negative terminals of the parallel modules are connected to the EIS online monitoring device and the host computer in sequence with wires. A K-type thermocouple is placed in the middle of each battery, and a 100W heating rod 8 is used to heat the battery module.

[0081] Preferably, the 100W heating rod is a cylindrical heating rod with a diameter × height (18mm × 65mm). The lithium-ion battery module and the 100W heating rod 8 are fixed together using a battery mold.

[0082] Preferably, the host computer is a computer with a Windows 10 operating system, a Core i5-6300HQ CPU with 4 cores and 4 threads, and 8GB of memory.

[0083] Step 11: While starting the heating, simultaneously open the online electrochemical impedance spectroscopy monitoring device 5 and the host computer data visualization software interface, and follow the instructions... Figure 3 The thermal runaway early warning process shown is used to issue an early warning. Figure 6 These are the temperature and voltage change curves during the heating process of battery 2. Figure 7 This describes the changes in the absolute value of the AC impedance |Z| and the phase angle θ at 1kHz during the heating process of battery 2.

[0084] Preferably, for parallel battery modules, if the absolute value of the AC impedance |Z| or the change in the phase angle θ exceeds the corresponding value at the upper limit of the optimal thermal management temperature of 45°C, a first-level thermal runaway warning is issued. Figure 6 At 115 seconds; after the first-level warning is issued, if the slope of the phase angle θ begins to change from positive to negative and continues to decrease, a second-level thermal runaway warning is issued, corresponding to... Figure 6 At 241 seconds; after the second-level warning is issued, if the absolute value of the AC impedance |Z| changes to a certain threshold or the slope of the phase angle θ begins to change from negative to positive and continues to increase, a first-level thermal runaway warning will be issued, corresponding to... Figure 6 The moment at 272 seconds.

[0085] Step 12: Disconnect the power supply to the 100W heating rod when thermal runaway occurs in the first battery.

[0086] It should be noted that the specific embodiments described above enable those skilled in the art to gain a more comprehensive understanding of the present invention, but do not limit the invention in any way. Therefore, although the present invention has been described in detail with reference to the accompanying drawings and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention. In short, all technical solutions and improvements that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. An online battery thermal runaway early warning method based on EIS parameter extraction, characterized in that, Includes the following steps: Step 1: Under an adiabatic environment, charge and discharge the battery using a battery testing device to change the battery's state of charge. Electrochemical impedance spectroscopy (EIS) spectra of batteries at different temperatures and under different charging states were collected using an online monitoring device. Step 2: Extract key EIS parameters using the RReliefF algorithm and sensitivity analysis. These key EIS parameters include: the real part of the impedance. Z’ and the virtual part Z” The absolute value of impedance |Z| and phase angle θ; The RReliefF algorithm was used to perform importance analysis on a large number of EIS key parameters, and box plots were used to perform sensitivity analysis on data features. Based on the importance analysis, feature parameters with better temperature prediction effect were selected, and based on the sensitivity analysis, key parameters for thermal runaway early warning that are highly sensitive to temperature and low sensitive to SOC were found. Step 3: Determine the thermal runaway early warning nodes corresponding to the key EIS parameters; Step 4: Use the extracted EIS key parameters and nodes to perform thermal runaway early warning for batteries with different connection methods; for single cells or series battery packs, when the absolute value of impedance |Z| or phase angle... θ If the warning threshold is exceeded, a Level 1 thermal runaway warning is issued; if it is not exceeded, the risk of thermal runaway is low. Based on the Level 1 warning, if the slope of the absolute value of impedance |Z| changes from negative to positive and continues to increase, or if the phase angle... θ Start fluctuating and Δ θ If the value is less than 0, a Level 2 thermal runaway warning is issued. Based on the Level 2 warning, if the absolute value of the impedance |Z| increases by an order of magnitude, a Level 3 thermal runaway warning is issued. For parallel battery packs, the absolute value of the impedance |Z| or the phase angle of the parallel battery pack... θ If the warning threshold is exceeded, a Level 1 thermal runaway warning is issued; if it is not exceeded, the risk of thermal runaway is low. Based on the Level 1 warning, if the phase angle... θ If the slope changes from positive to negative and continues to decrease, a second-level thermal runaway warning is issued. Based on the second-level warning, if the absolute value of the impedance |Z| of the parallel battery pack increases beyond the warning threshold or phase angle... θ If the slope changes from negative to positive and begins to increase, a level three warning for thermal runaway will be issued.

2. The online battery thermal runaway early warning method based on EIS parameter extraction according to claim 1, characterized in that: Step 4 includes: issuing a Level 1 thermal runaway warning when the battery temperature exceeds the upper limit of the optimal thermal management temperature threshold, issuing a Level 2 thermal runaway warning when the battery self-generated heat temperature range, and issuing a Level 3 thermal runaway warning when the battery voltage drops.

3. The online battery thermal runaway early warning method based on EIS parameter extraction according to claim 2, characterized in that: The upper limit of the optimal temperature threshold for thermal management is 45℃, and the temperature range for battery self-generated heat is 80℃~105℃.

4. The thermal runaway early warning system based on EIS parameter extraction for online battery thermal runaway early warning according to any one of claims 1-3, characterized in that: The system includes a battery, a chamber providing an insulated environment, a battery testing device, an online electrochemical impedance spectroscopy (EIS) monitoring device, and a host computer. The temperature of the battery is adjusted using the insulated chamber to achieve thermal equilibrium within the battery. The battery is charged and discharged using the battery testing device to change its state of charge (SOC). The online EIS monitoring device and host computer collect and analyze the battery temperature and EIS key parameter changes when the temperature stabilizes and at different SOCs. The wall temperature of the insulated chamber is controlled by programming on the host computer. The chamber contains a temperature acquisition device for monitoring the wall temperature and a temperature acquisition device for monitoring the sample temperature.

5. The thermal runaway early warning system according to claim 4, characterized in that: The chamber that provides an insulated environment is an adiabatic accelerating calorimeter, which contains a sealed insulated tank.

6. The thermal runaway early warning system according to claim 4, characterized in that: The online electrochemical impedance spectroscopy monitoring device collects data under dynamic battery charging and discharging conditions. It simultaneously collects data from multiple channels via a daisy-chain connection, measuring the real part of the battery's impedance at different frequencies. Z’ and the virtual part Z” Then the absolute value of the impedance can be calculated. |Z| and phase angle θ , |Z| and θ Calculated using the following formula: ； 。 7. The thermal runaway early warning system according to claim 4, characterized in that: The aforementioned online electrochemical impedance spectroscopy monitoring device includes a battery management chip, a PC main control board, and an FPC slave control board.

Images