Lithium battery thermal runaway detection device

Abstract

The invention relates to the technical field of lithium battery protection, and discloses a lithium battery thermal runaway detection device which comprises a matched power supply, a high-frequency alternating magnetic field coil, a thickness sensor, a broadband ultrasonic sensor, a high-frequency current transformer and an intelligent processing module. The high-frequency alternating magnetic field coil emits high-frequency electromagnetic waves to enable the lithium battery diaphragm to output test voltage, if the diaphragm has insulation weak points, partial discharge is caused, and liquefied water is formed and gasified due to small-area breakdown. The thickness sensor detects the bulging thickness of the battery caused by gasification of liquefied water, the broadband ultrasonic sensor receives an ultrasonic signal of disruptive discharge, and the high-frequency current transformer collects the high-frequency characteristic quantity of current. And the intelligent processing module receives the data, normalizes the data, calculates a fault value in a weighted manner, and judges that the diaphragm has an early fault point if the fault value exceeds a threshold value. The device performs multi-dimensional detection, accurately calculates a fault value, can timely find an early fault point of the lithium battery diaphragm, effectively prevents thermal runaway, and ensures the use safety of a lithium battery.

Description

Technical Field

[0001] The present application relates to the technical field of lithium battery protection, and in particular to a lithium battery thermal runaway detection device. Background Art

[0002] In today's energy field, lithium batteries have occupied an important position in many application scenarios due to their high energy density, long service life and good charge and discharge performance. Whether in energy storage systems, providing a stable supply of electricity, or in the field of electric vehicles, promoting the development of green travel, lithium batteries play a key role. However, with the increasing application of lithium batteries, the problem of thermal runaway fire has gradually become prominent, becoming a major obstacle to its further development and large-scale application.

[0003] The occurrence of thermal runaway fire of lithium battery is due to the runaway chemical reaction inside the battery, which leads to rapid accumulation of heat and eventually causes combustion. This fire is different from ordinary fires. It is a spontaneous combustion from the inside out. This characteristic makes the fire extinguishing process difficult. Traditional fire extinguishing methods, such as pouring water and using dry powder fire extinguishers, often cannot effectively curb the combustion of lithium batteries. Because the combustion reaction inside the lithium battery takes place in the closed battery shell, it is difficult for external fire extinguishing agents to directly act on the fire source. In addition, when lithium batteries burn, they release a large amount of heat and harmful gases, which further increases the difficulty and risk of fire extinguishing.

[0004] When a lithium battery experiences thermal runaway, on-site emergency response also faces great challenges. Due to the danger of explosions that are difficult to predict and control at any time, it is difficult for rescuers to approach the accident site for effective treatment. This not only poses a serious threat to the safety of rescuers' lives, but also makes it impossible for emergency response work to be carried out in a timely and smooth manner, resulting in the possibility of further spread of the fire and causing greater losses.

[0005] In terms of fire warning, the traditional fire warning technology based on measuring lithium battery current, voltage, capacity, temperature and other parameters, which is currently widely used, has exposed obvious limitations in practical applications. In the early stage of thermal runaway of lithium batteries, the internal chemical reactions and physical changes are relatively subtle, and the changes in these traditional measurement parameters are not significant, making it difficult to accurately reflect the potential dangers that already exist inside the battery. Therefore, it is often difficult to rely on these parameters to determine whether the lithium battery is in the early stage of thermal runaway and to issue an alarm signal in a timely manner. For example, in some cases, the diaphragm inside the battery may have been slightly damaged, but it cannot be detected by measuring parameters such as current and voltage. Only when the damage is further expanded, resulting in a significant decline in battery performance, can traditional measurement methods detect abnormalities, but at this time, the best time for prevention and treatment has often been missed. Summary of the invention

[0006] In order to be able to detect potential hazards in the early stage of lithium battery thermal runaway in a timely manner and take corresponding preventive measures, the present application provides a lithium battery thermal runaway detection device.

[0007] The present application provides a lithium battery thermal runaway detection device, adopting the following technical solutions: A lithium battery thermal runaway detection device includes a supporting power supply, a high-frequency alternating magnetic field coil, a thickness sensor, a broadband ultrasonic sensor, a high-frequency current transformer, and an intelligent processing module; The high-frequency alternating magnetic field coil is electrically connected to the supporting power supply through a high-frequency transformer and is used to emit high-frequency electromagnetic waves to output a test voltage at both ends of the lithium battery separator. If there is a weak insulation point in the lithium battery separator, partial discharge will occur under the action of the test voltage. If there is a small-area breakdown in the lithium battery separator, the water molecules oscillation heating effect is used to form liquefied water at the small-area breakdown of the lithium battery separator. When the heat accumulation at the liquefied water reaches the boiling point of the liquefied water, the liquefied water vaporizes; The thickness sensor is electrically connected to the supporting power supply and is used to detect the bulging thickness of the lithium battery caused by the vaporization of the dielectric liquefied water. The broadband ultrasonic sensor is electrically connected to the supporting power supply and is used to receive the ultrasonic signal generated by the breakdown discharge of the measured lithium battery separator under the action of the high-frequency electromagnetic wave. The high-frequency current transformer is electrically connected to the supporting power supply and is clamped to an output terminal of the measured lithium battery and is used to collect the high-frequency characteristic quantity in the battery current signal. The intelligent processing module is used to receive the bulging thickness, the ultrasonic signal, and the high-frequency characteristic, and after normalizing the bulging thickness, the ultrasonic signal, and the high-frequency characteristic quantity, calculate the fault value by weighted calculation. If the fault value exceeds the set fault threshold, it is determined that there is an early fault point in the measured lithium battery separator.

[0008] By adopting the above technical solutions, sensors based on multiple different principles are used to detect the lithium battery separator from multiple dimensions. Locally discharge is detected electrically, the bulging thickness is monitored physically, ultrasonic signals are captured acoustically, and analysis is also carried out from the characteristics of current changes. Information from multiple aspects complements and verifies each other. For example, when the broadband ultrasonic sensor detects a discharge signal, if the high-frequency current transformer also detects an abnormality in the high-frequency characteristic quantity and the thickness sensor finds a change in the bulging thickness, the early fault points of the separator can be judged more accurately. Compared with a single detection method, the detection accuracy is greatly improved. It can detect the early fault stages such as weak insulation points and small-area breakdowns in the lithium battery separator. By emitting high-frequency electromagnetic waves to generate a test voltage to detect local discharge, small changes in insulation performance can be found; by using the oscillation heating effect of water molecules to detect the liquefied water formed by small-area breakdown and subsequent vaporization bulging, subtle physical changes can be captured at the initial stage of the fault, providing the possibility to take measures in advance to prevent thermal runaway and effectively reducing the risk of safety accidents caused by thermal runaway of lithium batteries. The collected data is normalized and weighted to calculate the fault value, enabling different types of data to be comprehensively evaluated under the same standard. The normalization process eliminates the influence of data dimension and order of magnitude differences, ensures that each detection index has a reasonable weight in fault judgment, makes the fault judgment result more scientific and reliable, and can more accurately reflect the actual fault condition of the lithium battery separator.

[0009] Optionally, the fault value = a1 × normalized bulging thickness value + a2 × normalized ultrasonic signal value + a3 × normalized high-frequency characteristic quantity value; where a1 is the weight coefficient of the normalized bulging thickness value, a2 is the weight coefficient of the normalized ultrasonic signal value, and a3 is the weight coefficient of the normalized high-frequency characteristic quantity value; among them, the normalized bulging thickness value, the normalized ultrasonic signal value, and the normalized high-frequency characteristic quantity value are calculated using the min-max normalization method.

[0010] By adopting the above technical solutions, the data scales of the bulging thickness, ultrasonic signal, and high-frequency characteristic quantity are unified, the dimension and order of magnitude differences are eliminated, and the accuracy of fault value calculation is improved; the weight coefficients can play a more accurate role, enhancing the reliability and stability of fault judgment; at the same time, it facilitates the calibration and optimization of the detection device. By adjusting the weight coefficients, it can quickly adapt to different lithium battery models and usage scenarios, improving the versatility and flexibility of the device.

[0011] Optionally, the intelligent processing module adjusts the size of a1 according to the usage time of the lithium battery; the shorter the usage time, the smaller a1; the longer the usage time, the larger a1.

[0012] By adopting the above technical solution, comprehensively consider the importance changes of various detection indexes in different stages of lithium battery use for fault judgment. Combining ultrasonic signals and high-frequency characteristic quantities, faults can be judged more accurately in different use stages of the battery. In the new battery stage, more emphasis is placed on judging whether there are early faults based on ultrasonic signals and high-frequency characteristic quantities; in the old battery stage, the weight of the bulging thickness information in fault judgment increases. The combination of multiple indexes and the reasonable distribution of weights can reduce misjudgment, improve the accuracy of fault judgment, discover potential thermal runaway hazards in advance, and ensure the safe use of lithium batteries.

[0013] Optionally, the intelligent processing module dynamically adjusts the weight coefficients according to the charge and discharge times N of the lithium battery; when N < the set charge and discharge times threshold N1, a1 = k1×N / N1; a2 = 1 - k2×N / N1, a3 = k3, where k1, k2, and k3 are constants.

[0014] By adopting the above technical solution, even tiny abnormalities in the initial stage of the battery can be accurately captured by reasonably distributing the weights of each detection index. During the process of gradually increasing charge and discharge times, pay timely attention to the change trends of the bulging thickness, ultrasonic signals, and high-frequency characteristic quantities, and predict the possible thermal runaway risks in advance. For example, in the initial stage of battery aging, if the high-frequency characteristic quantity shows abnormalities, even if the changes in the bulging thickness and ultrasonic signals are not obvious, due to the reasonable setting of the weight coefficients, the detection device can issue a warning to remind the user to take measures in time, effectively ensuring the safe use of lithium batteries and extending the battery life.

[0015] Optionally, it further includes an infrared sensor for acquiring an infrared image of the lithium battery and sending it to the intelligent processing module; The intelligent processing module identifies target features from the infrared image; Divide the target features into M feature regions, and calculate the heat value of each feature region; Calculate the coefficient of variation of the heat values of the M feature regions as the heat distribution value, and calculate the total heat value of the M feature regions; Normalize the heat distribution value and the total heat value and calculate the infrared heat value by weighted calculation; If the infrared heat value is higher than the set threshold, then normalize the bulging thickness, the ultrasonic signal, the high-frequency characteristic quantity, and the infrared temperature value and calculate the hidden danger value by weighted calculation; if the hidden danger value exceeds the set hidden danger threshold, it is judged that there is an early fault point in the separator of the measured lithium battery.

[0016] By adopting the above technical solution, an infrared sensor is introduced to obtain the infrared image of the lithium battery, adding information in the temperature dimension to the detection. The thermal runaway of a lithium battery is usually accompanied by abnormal temperature changes, and the infrared image can intuitively reflect the temperature distribution on the battery surface. The intelligent processing module identifies target features, divides feature regions and calculates the heat value from the infrared image, and then obtains the heat distribution value and the total heat value. These data can accurately present the heat state inside the battery. When the calculated infrared heat value is higher than the set threshold, it indicates that there may be potential problems such as local overheating in the battery. At this time, the infrared temperature value is incorporated into the calculation of the hidden danger value, integrating multi-dimensional information such as the bulging thickness, ultrasonic signal, and high-frequency feature quantity, making the judgment of the early fault point of the lithium battery diaphragm more accurate. For example, even if other detection indicators show normal, but the infrared heat value increases abnormally, the potential fault risk can be detected in time to avoid missed judgment.

[0017] Optionally, a pair of thickness sensors are provided and installed on both sides of the lithium battery; the thickness sensors are electrically connected to a thickness detection meter for displaying the change in the battery thickness detected by the thickness sensors.

[0018] By adopting the above technical solution, the two thickness sensors detect the battery thickness from different positions, and can obtain more comprehensive information about the battery bulging. If the internal fault of the battery causes uneven bulging, relying on a single sensor may miss the bulging situation in some areas, resulting in misjudgment. The sensors on both sides can respectively detect the thickness changes of different parts of the battery. By comprehensively analyzing these data, it is possible to more accurately judge whether there are problems inside the battery and the approximate location of the problem. For example, if the thickness change detected by one side sensor is significantly greater than that of the other side, it may indicate that the internal fault of the battery is biased towards this side, which helps to more accurately locate the fault point and improve the accuracy of the judgment of the early fault of the lithium battery diaphragm. The thickness sensors are electrically connected to the thickness detection meter, which can intuitively display the change in the battery thickness in real time. The operator can also quickly understand the bulging situation of the battery by directly observing the numerical change of the thickness detection meter, further verifying the accuracy of the detection result. During the detection process, abnormalities can be detected in time and responses can be made. For example, when the thickness change exceeds a certain range, further detection measures can be taken immediately or the use of the battery can be stopped to ensure the convenience and safety of the operation. At the same time, this intuitive display method also facilitates the operator to record data for subsequent analysis of the performance change trend of the battery, providing convenience for the maintenance and management of the battery.

[0019] Optionally, the signal output end of the broadband ultrasonic sensor is electrically connected to a band-pass filter, and the band-pass filter is electrically connected to an oscilloscope. The band-pass filter is used to screen signals within a specific frequency range, and the oscilloscope is used to display the signal waveform detected by the broadband ultrasonic sensor after being screened by the band-pass filter.

[0020] By adopting the above technical solutions, the working environment of the lithium battery is complex, and various interference signals will be received during the detection by the broadband ultrasonic sensor. The band-pass filter can screen signals within a specific frequency range, remove interference from other frequencies, make the signals received by the oscilloscope purer, only retain the ultrasonic signals related to the breakdown discharge of the lithium battery diaphragm, and improve the signal quality. The signal waveform displayed by the oscilloscope contains rich information, such as signal amplitude, frequency, pulse width, etc. By analyzing the waveform, technicians can further verify the accuracy of the detection results, judge whether there is breakdown discharge and the discharge intensity, frequency, etc., and can also compare the normal and abnormal waveforms to more accurately judge the fault state of the lithium battery diaphragm and assist in fault judgment.

[0021] Optionally, a high-pass filter is electrically connected to the signal output end of the high-frequency current transformer, and the high-pass filter is electrically connected to a high-frequency ammeter; the high-pass filter is used to filter out low-frequency interference signals, and the high-frequency ammeter is used to display the high-frequency current characteristic quantity of the high-frequency current transformer after being filtered by the high-pass filter.

[0022] By adopting the above technical solutions, when the lithium battery is working, the current signal is complex and contains various frequency components. The low-frequency interference signals are likely to mask the high-frequency characteristic quantities generated by local discharge inside the battery. The high-pass filter can effectively filter out the low-frequency interference and only allow the high-frequency current characteristic quantities to pass through, accurately extracting the key information related to potential faults of the lithium battery and avoiding the misguidance of fault judgment caused by low-frequency signal interference. The high-frequency ammeter displays the filtered high-frequency current characteristic quantity in an intuitive numerical value. Technicians can directly understand the high-frequency characteristics of the battery current by observing the change in the reading of the high-frequency ammeter, and further verify the accuracy of the detection results. Once the reading shows abnormal fluctuations, it can be quickly judged that the battery may have potential faults such as local discharge, providing convenience for timely discovery and handling of potential lithium battery thermal runaway hazards.

[0023] In summary, the present application includes at least one of the following beneficial technical effects: Sensors based on multiple different principles are used to detect the lithium battery diaphragm from multiple dimensions, including electricity (detecting local discharge), physics (monitoring bulging thickness), acoustics (capturing ultrasonic signals), and analysis of current change characteristics. The information from multiple aspects complements and verifies each other. Compared with a single detection method, the accuracy of detection is greatly improved, and it can detect early faults such as weak insulation points and small-area breakdowns in the lithium battery diaphragm.

[0024] Normalize different types of data such as the collected bulging thickness, ultrasonic signals, and high-frequency characteristic quantities, and calculate the fault value by weighting, unifying the data scale, eliminating the influence of dimension and order of magnitude differences, ensuring that each detection index has a reasonable weight in fault judgment, making the fault judgment result more scientific and reliable, and being able to more accurately reflect the actual fault condition of the lithium battery diaphragm.

[0025] The intelligent processing module dynamically adjusts the weight coefficient according to the usage time and charge-discharge times of the lithium battery, and comprehensively considers the importance changes of each detection index in different stages of lithium battery use for fault judgment. In the new battery stage, it focuses more on judging whether there are early faults based on ultrasonic signals and high-frequency characteristic quantities; in the old battery stage, the weight of the bulging thickness information in fault judgment increases. The combination of multiple indicators and the reasonable distribution of weights can reduce misjudgment, improve the accuracy of fault judgment, discover potential thermal runaway hazards in advance, and ensure the safe use of lithium batteries. Description of the Drawings

[0026] Figure 1 It is a module diagram of the lithium battery thermal runaway detection device.

[0027] Reference Signs: 1. High-frequency excitation module; 2. Physical detection module; 3. Acoustic detection module; 4. Electrical detection module; 5. Intelligent processing module. Detailed Embodiments

[0028] The following details the embodiments of the present application, and the examples of the embodiments are shown in the drawings.

[0029] In the description of this specification, the description with reference to the terms "certain embodiments", "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0030] The embodiment of the present application discloses a lithium battery thermal runaway detection device, referring to Figure 1 , including the following modules: High-frequency excitation module: It includes a high-frequency alternating magnetic field coil and a supporting power supply, and can generate a high-frequency alternating magnetic field.

[0031] Physical detection module: It includes a double-thickness sensor group and a thickness display instrument.

[0032] Acoustic detection module: It includes a broadband ultrasonic sensor, a band-pass filtering system and a digital oscilloscope.

[0033] Electrical detection module: It includes a high-frequency current transformer, a high-pass filtering system and a high-frequency ammeter.

[0034] Intelligent processing module, including a normalization algorithm and a fuzzy decision-making system.

[0035] The form of a lithium battery thermal runaway detection device can be diverse. It can be a handheld device, a fixed device, or a movable device with wheels, etc. The overall structure of the detection device can accommodate a lithium battery, and limit positions can be set according to the structure of the lithium battery, etc., facilitating the detection sensor to quickly detect the lithium battery.

[0036] The high-frequency alternating magnetic field coil is electrically connected to a supporting power supply through a high-frequency transformer and is used to emit high-frequency electromagnetic waves to output a test voltage across both ends of the lithium battery separator. When in use, it is located on one side of the lithium battery. If there is a weak insulation point in the lithium battery separator, partial discharge will occur under the action of the test voltage; if there is a small-area breakdown in the lithium battery separator, the water molecule oscillation heating effect is utilized to form liquefied water at the small-area breakdown of the lithium battery separator; when the heat accumulation at the liquefied water reaches the boiling point of the liquefied water, the liquefied water vaporizes.

[0037] The specific principle of the high-frequency alternating magnetic field coil emitting high-frequency electromagnetic waves to conduct a withstand voltage test on the lithium battery separator is as follows: When an alternating current is passed through the high-frequency alternating magnetic field coil, an alternating magnetic field will be generated around it. According to the law of electromagnetic induction, an electromotive force will be induced across both ends of the dielectric separator inside the lithium battery in this alternating magnetic field. Since the high-frequency alternating magnetic field coil emits high-frequency electromagnetic waves and the alternating magnetic field changes rapidly, a relatively high test voltage can be induced across both ends of the separator. This voltage can simulate the high-voltage situation that the lithium battery may withstand during actual use to test the insulation performance of the separator. If there is a weak insulation point in the separator, partial discharge will occur when the induced voltage reaches a certain level. The lithium battery separator, as a dielectric material, has certain insulation performance and withstand voltage limit. Under the high voltage generated by the high-frequency electromagnetic waves, if the insulation performance of the separator is good, it can withstand the induced voltage without breakdown or leakage; on the contrary, when there are defects in the separator, such as tiny damage, impurity mixing, etc., its insulation performance will decline, and local current leakage or breakdown may occur at a lower induced voltage. By monitoring parameters such as whether there is a discharge phenomenon, the intensity of the discharge, and the voltage value when the discharge occurs, it can be determined whether the insulation withstand voltage performance of the separator meets the standard, and thus whether there is a potential hidden danger of thermal runaway in the internal separator of the lithium battery.

[0038] The thickness sensor is electrically connected to a supporting power supply and is used to detect the swelling thickness of the lithium battery caused by the vaporization of the dielectric liquefied water. There are a pair of thickness sensors, which are installed on both sides of the lithium battery; the thickness sensors are electrically connected to a thickness detection meter for displaying the change in the battery thickness detected by the thickness sensors.

[0039] Two thickness sensors detect the battery thickness from different positions, enabling more comprehensive acquisition of battery swelling information. If internal battery faults lead to uneven swelling, relying solely on a single sensor may miss the swelling in some areas, resulting in misjudgment. The sensors on both sides can respectively detect the thickness changes of different parts of the battery. By comprehensively analyzing this data, it is possible to more accurately determine whether there are problems inside the battery and the approximate location of the problems. For example, if the thickness change detected by one side sensor is significantly greater than that of the other side, it may imply that the internal battery fault is biased towards this side, which helps to more precisely locate the fault point and improve the accuracy of early fault judgment for lithium battery diaphragms. The thickness sensors are electrically connected to a thickness detection meter, which can intuitively display the battery thickness change in real time. Operators can also quickly understand the battery swelling situation by directly observing the numerical changes of the thickness detection meter, further verifying the accuracy of the detection results. During the detection process, abnormalities can be promptly discovered and responses can be made. For example, when the thickness change exceeds a certain range, further detection measures can be immediately taken or the battery use can be stopped to ensure the convenience and safety of the operation. At the same time, this intuitive display method also facilitates operators to record data for subsequent analysis of the battery performance change trend, providing convenience for battery maintenance and management.

[0040] Among them, detecting the liquefied water formed by the small-area breakdown of the lithium battery diaphragm by using the water molecule oscillation heating effect is based on the special physical properties of water molecules under the action of high-frequency electromagnetic waves and the internal structure characteristics of lithium batteries.

[0041] Interaction between high-frequency electromagnetic waves and water molecules: When a high-frequency coupling coil emits high-frequency electromagnetic waves and acts on a lithium battery, the water molecules inside the lithium battery will be affected by the electromagnetic waves. Water molecules are polar molecules. In a high-frequency alternating electric field, water molecules will vibrate and rotate rapidly along with the direction of the electric field. This rapid oscillating motion will cause friction between water molecules and between water molecules and other surrounding particles, thereby converting electromagnetic energy into heat energy, resulting in an increase in the temperature of water molecules. This is the water molecule oscillation heating effect.

[0042] Relationship between small-area breakdown of the diaphragm and liquefied water: During the normal operation of a lithium battery, the diaphragm plays a role in isolating the positive and negative electrodes and preventing short circuits. When a small-area breakdown occurs in the diaphragm, the electrochemical reaction between the positive and negative electrodes will be abnormal, and there may be local overheating phenomena, which will prompt the decomposition of the diaphragm material or the reaction of the electrolyte, and then produce liquefied water. These liquefied waters will accumulate near the area where the diaphragm is broken down.

[0043] Implementation process of detection: Due to the penetrability of high-frequency electromagnetic waves, they can reach the area where there is liquefied water in the internal separator of the lithium battery. In this area, the above-mentioned water molecule oscillation heating effect will further increase the temperature of the liquefied water, and the surrounding substances will also undergo some physical changes due to heat. For example, the heated liquefied water vaporizes to form tiny bubbles, resulting in a slight change in the internal pressure of the battery or causing the battery to bulge slightly. By detecting the slight bulge of the battery with a thickness gauge sensor, it is possible to indirectly determine whether there is liquefied water formed due to a small-area breakdown of the separator. If an abnormal bulge of the battery is detected, it indicates that there may be a small-area breakdown of the separator and liquefied water is generated, thus realizing the detection of liquefied water formed by a small-area breakdown of the lithium battery separator.

[0044] A broadband ultrasonic sensor, electrically connected to a supporting power supply, is used to receive ultrasonic signals generated by breakdown discharge of the measured lithium battery separator under the action of high-frequency electromagnetic waves. The signal output end of the broadband ultrasonic sensor is electrically connected to a band-pass filter, and the band-pass filter is electrically connected to an oscilloscope. The band-pass filter is used to screen signals within a specific frequency range, and the oscilloscope is used to display the signal waveform detected by the broadband ultrasonic sensor after being screened by the band-pass filter.

[0045] The working environment of lithium batteries is complex, and the broadband ultrasonic sensor will receive various interference signals during detection. The band-pass filter can screen signals within a specific frequency range, remove interference from other frequencies, make the signals received by the oscilloscope purer, only retain the ultrasonic signals related to the breakdown discharge of the lithium battery separator, and improve the signal quality. The signal waveform displayed by the oscilloscope contains rich information, such as signal amplitude, frequency, pulse width, etc. Technicians can further verify the accuracy of the detection results by analyzing the waveform, judge whether there is breakdown discharge and the discharge intensity, frequency, etc., and can also compare normal and abnormal waveforms to more accurately judge the fault state of the lithium battery separator and assist in fault judgment.

[0046] A high-frequency current transformer, electrically connected to a supporting power supply, is clamped to one of the lead terminals of the measured lithium battery and is used to collect high-frequency characteristic quantities in the battery current signal. The signal output end of the high-frequency current transformer is electrically connected to a high-pass filter, and the high-pass filter is electrically connected to a high-frequency ammeter; the high-pass filter is used to filter out low-frequency interference signals, and the high-frequency ammeter is used to display the high-frequency current characteristic quantities of the high-frequency current transformer after being filtered by the high-pass filter.

[0047] When a lithium battery is working, the current signal is complex and contains various frequency components. The low-frequency interference signal is likely to mask the high-frequency characteristic quantities generated by partial discharge inside the battery. The high-pass filter can effectively filter out the low-frequency interference and only allow the high-frequency current characteristic quantities to pass through, accurately extracting the key information related to the potential faults of the lithium battery and avoiding the misguidance of fault judgment caused by low-frequency signal interference. The high-frequency ammeter displays the filtered high-frequency current characteristic quantities in an intuitive numerical value. Technicians can directly understand the high-frequency characteristics of the battery current by observing the change in the reading of the high-frequency ammeter, further verifying the accuracy of the detection result. Once the reading shows abnormal fluctuations, it can quickly determine that there may be potential fault hazards such as partial discharge in the battery, providing convenience for timely discovering and handling the hidden dangers of lithium battery thermal runaway.

[0048] The intelligent processing module is used to receive the bulging thickness, ultrasonic signal, and high-frequency characteristics, normalize the bulging thickness, ultrasonic signal, and high-frequency characteristic quantities, and then calculate the fault value through weighted calculation; if the fault value exceeds the set fault threshold, it is determined that there is an early fault point in the separator of the measured lithium battery.

[0049] Through the multi-dimensional collaborative detection technology, the accurate early warning of the early faults of the lithium battery separator is realized: the high-frequency excitation module simulates the high-voltage environment of the actual working condition through an alternating magnetic field, effectively exciting the partial discharge and the water molecule oscillation heating effect at the separator defect; the double thickness sensor group monitors the bulging change of the battery in real time, combining the detection on both sides to eliminate single-point misjudgment; the broadband ultrasonic sensor cooperates with the band-pass filtering system to accurately capture the ultrasonic signal generated by the discharge; the high-frequency current transformer and the high-pass filter are combined to extract the characteristic current of the partial discharge; the intelligent processing module fuses multi-source data through the normalization algorithm and outputs the fault value through the fuzzy decision-making system to realize the intelligent discrimination of the early fault point. This solution significantly improves the recognition sensitivity and positioning accuracy of the hidden dangers of thermal runaway through the multi-parameter fusion detection of physical deformation, acoustic signals, and electrical characteristics, providing a reliable guarantee for the safe operation of lithium batteries.

[0050] The calculation of the fault value is the key link in the entire lithium battery thermal runaway detection device to determine whether there is an early fault point in the lithium battery separator. Specifically, the calculation formula of the fault value is: fault value = a1 × normalized value of bulging thickness + a2 × normalized value of ultrasonic signal + a3 × normalized value of high-frequency characteristic quantity. In this formula, a1 is the weight coefficient of the normalized value of the bulging thickness, and its numerical size reflects the importance of the parameter of the bulging thickness in the comprehensive judgment. a2 is the weight coefficient of the normalized value of the ultrasonic signal, which measures the contribution ratio of the ultrasonic signal to the fault judgment. a3, as the weight coefficient of the normalized value of the high-frequency characteristic quantity, reflects the key role of the high-frequency characteristic quantity in the entire fault judgment system.

[0051] Among them, the normalized values of the bulging thickness, ultrasonic signal, and high-frequency feature quantity are all calculated using the min-max normalization method. The core of this normalization method lies in unifying the scale of the data, mapping the original data of different parameters into a fixed interval, thereby eliminating the influence brought by the differences in dimension and order of magnitude. In the detection of lithium battery thermal runaway, the bulging thickness, ultrasonic signal, and high-frequency feature quantity originally have different dimensions and orders of magnitude. Without normalization, these differences may cause some parameters to account for too large or too small a proportion in the calculation of the fault value, thereby affecting the accuracy of fault judgment.

[0052] For example, in the scenario of lithium battery detection in an energy storage power station: In order to ensure the safe and stable operation of the lithium battery system, a large-scale energy storage power station uses this thermal runaway detection device to regularly detect the batteries.

[0053] For a newly put into use battery pack, initially set the weight coefficients a1 = 0.1, a2 = 0.45, a3 = 0.45.

[0054] In a detection, the normalized value of the bulging thickness of a certain battery is 0.05, indicating that the battery has basically no bulging phenomenon. The normalized value of the ultrasonic signal is 0.2, and the normalized value of the high-frequency feature quantity is 0.15.

[0055] Calculate the fault value: Fault value = 0.1×0.05 + 0.45×0.2 + 0.45×0.15 = 0.005 + 0.09 + 0.0675 = 0.1625. Since the fault threshold is 0.4 and this fault value is less than the threshold, it is judged that the battery currently has no early fault points.

[0056] After five years of operation, the same battery pack is detected again for the battery pack that has been in operation for many years. At this time, adjust the weight coefficients to a1 = 0.7, a2 = 0.15, a3 = 0.15.

[0057] It is detected that the normalized value of the bulging thickness of a certain battery is 0.7, indicating that the battery is severely bulged.

[0058] The normalized value of the ultrasonic signal is 0.3, and the normalized value of the high-frequency feature quantity is 0.2.

[0059] Calculate the fault value: Fault value = 0.7×0.7 + 0.15×0.3 + 0.15×0.2 = 0.49 + 0.045 + 0.03 = 0.565. Because the fault value is greater than the fault threshold of 0.4, it is judged that there is an early fault point in the battery diaphragm, and the battery needs to be replaced in time to ensure the safe operation of the energy storage power station.

[0060] Through the min-max normalization method, the data of the bulging thickness, ultrasonic signal, and high-frequency characteristic quantity are brought to the same scale, providing a solid foundation for the subsequent calculation of the fault value. The advantages of this data processing method are not only reflected in improving the accuracy of the fault value calculation but also enabling the weight coefficient to play a more precise role. In the process of fault judgment, the weight coefficient is set according to the degree of influence of different parameters on the fault. The normalized data enables the weight coefficient to more accurately reflect the importance of each parameter, thereby enhancing the reliability and stability of the fault judgment.

[0061] In addition, the use of the min-max normalization method also facilitates the calibration and optimization of the detection device. In practical applications, different models of lithium batteries and different usage scenarios may have different requirements for thermal runaway detection. By adjusting the weight coefficient, the detection device can quickly adapt to these changes and meet different detection needs. For example, in some lithium battery models that are more sensitive to the bulging thickness, the value of a1 can be appropriately increased so that the bulging thickness occupies a greater proportion in the fault judgment; while in some scenarios where the ultrasonic signal can better reflect the fault situation, the weight of a2 can be correspondingly increased. This flexibility makes the detection device more versatile and can be widely used in the thermal runaway detection of various lithium batteries, providing a more reliable guarantee for the safe operation of lithium batteries.

[0062] The intelligent processing module can dynamically adjust the size of the weight coefficient a1 according to the usage time of the lithium battery during the thermal runaway detection of the lithium battery. This adjustment mechanism fully reflects the in-depth understanding and precise grasp of the characteristics of different usage stages of the lithium battery. Specifically, when the usage time of the lithium battery is short, the intelligent processing module will set a1 to a smaller value; as the usage time of the lithium battery continues to increase, the value of a1 will increase accordingly.

[0063] During the entire usage cycle of the lithium battery, its performance and state will change significantly, and the fault information reflected by each detection index also varies at different stages. In the new battery stage, the overall structure and performance of the lithium battery are relatively stable, and the bulging phenomenon is usually not obvious. At this time, the ultrasonic signal and high-frequency characteristic quantity can more sensitively capture the potential early fault hidden dangers inside the battery. For example, a broadband ultrasonic sensor can detect the ultrasonic signal generated by the breakdown discharge of the lithium battery diaphragm under the action of high-frequency electromagnetic waves, and a high-frequency current transformer can collect the high-frequency characteristic quantity in the battery current signal. These signals and characteristic quantities can issue early warnings when the battery has minor faults. Therefore, in the new battery stage, the intelligent processing module will focus more on judging whether there are early faults in the battery based on the ultrasonic signal and high-frequency characteristic quantity, and appropriately reduce the weight of the bulging thickness in the fault judgment.

[0064] However, as the usage time of lithium batteries increases, the internal structure and performance of the batteries will gradually change. Problems such as aging and damage may occur to the separator, and the batteries may also experience bulging. In the stage of old batteries, the bulging thickness becomes a more important fault judgment index. The change in the bulging thickness can directly reflect the physical changes inside the battery, such as the bulging of the battery caused by the vaporization of liquefied water formed by the small-area breakdown of the separator. Therefore, the intelligent processing module will increase the value of a1 and enhance the weight of the bulging thickness information in fault judgment, enabling the bulging thickness to play a greater role in comprehensive fault judgment.

[0065] Through this method of combining multiple indicators with reasonable weight distribution, the intelligent processing module can more accurately judge faults at different usage stages of lithium batteries. In the stage of new batteries, focusing on ultrasonic signals and high-frequency characteristic quantities can timely detect early potential faults; in the stage of old batteries, increasing the weight of the bulging thickness information can more accurately grasp the aging and fault conditions of the batteries. This dynamic weight adjustment mechanism can effectively reduce the occurrence of misjudgments and greatly improve the accuracy of fault judgment. In practical applications, it can detect the hidden dangers of lithium battery thermal runaway in advance and ensure the safe use of lithium batteries.

[0066] For example, taking a new car as an example, it has just traveled 1000 kilometers, which is equivalent to a relatively short usage time of the lithium battery. At this time, the intelligent processing module will set a1 to a relatively small value, such as 0.2. This is because in the stage of new batteries, the internal structure and performance of the lithium battery are relatively stable, and the bulging phenomenon is usually not obvious. Just like the lithium battery of this new car, its internal separator is still new, without obvious aging and damage, and the change in the bulging thickness is very small. And in this stage, ultrasonic signals and high-frequency characteristic quantities can more sensitively capture the potential fault hazards that may exist inside the battery at an early stage. As time goes by, this car has traveled 50000 kilometers, and the lithium battery has entered the stage of old batteries with a longer usage time. At this time, the intelligent processing module will increase the value of a1, such as adjusting it to 0.6. Because as the usage time increases, the internal structure and performance of the battery gradually change. The separator of the lithium battery of this car has aged to a certain extent, and the electrochemical reaction between the positive and negative electrodes has become more complex, resulting in some local overheating phenomena inside the battery, which prompts the decomposition of the separator material or the reaction of the electrolyte, and then generates liquefied water. These liquefied waters accumulate near the area where the separator is broken down, causing the battery to show a slight bulge.

[0067] The intelligent processing module dynamically adjusts the weight coefficient according to the charge and discharge times N of the lithium battery during the detection of lithium battery thermal runaway. When the charge and discharge times N are less than the set charge and discharge times threshold N1, the weight coefficient will be adjusted according to specific rules. Specifically, a1 = k1×N / N1, a2 = 1 - k2×N / N1, a3 = k3, where k1, k2, and k3 are all constants.

[0068] During the continuous charge and discharge process of a lithium battery, its internal structure and performance will gradually change. The number of charge and discharge cycles can measure the degree of battery aging. Assume that the set charge and discharge cycle threshold N1 = 500 times, constants k1 = 0.5, k2 = 0.6, k3 = 0.2. For a lithium battery in an electric vehicle with a charge and discharge cycle number N = 50 times, at this time, calculated according to the formula, a1 = 0.5×50 / 500 = 0.05, a2 = 1 - 0.6×50 / 500 = 0.94, a3 = 0.2. At this stage, the battery is relatively new, and the aging degree of key components such as the separator is very low. Just like the lithium battery of this new car, its bulging thickness change is usually small, and this change may not be caused by serious faults. Therefore, it is very reasonable to set a1 as k1×N / N1 proportional to the number of charge and discharge cycles. This means that as the number of charge and discharge cycles increases, the weight of the bulging thickness in fault judgment will gradually increase, but in the initial stage, due to the small number of charge and discharge cycles, its weight is small. Doing so can avoid over - focusing on the bulging phenomenon that is not caused by faults, thereby reducing the possibility of misjudgment.

[0069] At the same time, set a2 as 1 - k2×N / N1, so that the weight of the ultrasonic signal gradually decreases as the number of charge and discharge cycles increases. In the new battery stage, the probability of serious problems such as breakdown discharge is very low, and the abnormality of the ultrasonic signal is often not the main fault feature. Therefore, in the initial stage, a higher weight is given to the ultrasonic signal to capture possible weak abnormalities. And a3 remains as the constant k3, which ensures that the high - frequency characteristic quantity has a stable weight in the initial stage to assist in judging weak early faults.

[0070] As the number of charge and discharge cycles continues to increase, each weight coefficient will change continuously according to the set rules, so that it can more accurately reflect the importance of each detection index for fault judgment at different aging stages, thereby greatly improving the accuracy of fault judgment. When the charge and discharge cycle number of the lithium battery in this electric vehicle reaches 200 times, a1 = 0.5×200 / 500 = 0.2, a2 = 1 - 0.6×200 / 500 = 0.76, a3 = 0.2. It can be seen that the weight of the bulging thickness has increased, while the weight of the ultrasonic signal has decreased accordingly.

[0071] By dynamically adjusting the weight coefficients, the adaptability of the detection device to lithium batteries in different charge and discharge states is greatly enhanced. Different users have very different frequencies and ways of using lithium batteries, and the number of charge and discharge cycles also varies. For example, some users often drive electric vehicles on long trips, and the lithium batteries will have relatively frequent charge and discharge cycles; while some users only travel short distances within the city, and the lithium batteries have relatively few charge and discharge cycles. This technical solution enables the detection device to adjust the detection strategy in real time according to the actual number of charge and discharge cycles of the battery. Whether it is a battery with frequent charge and discharge or a battery with fewer charge and discharge cycles, a reasonable assessment can be obtained based on its degree of aging.

[0072] In practical applications, this way of dynamically adjusting the weight coefficients can avoid detection deviations caused by different usage habits, ensure the stable and efficient operation of the detection system in various usage scenarios, and optimize the overall detection performance. Moreover, dynamically adjusting the weight coefficients according to the number of charge and discharge cycles helps the detection device to detect potential fault hazards in the early stage of battery aging. By reasonably allocating the weights of each detection index, even the slightest abnormalities that occur in the initial stage of the battery can be accurately captured.

[0073] For example, when the number of charge and discharge cycles of the lithium battery in an electric vehicle reaches 100 times, although the changes in the bulging thickness and ultrasonic signal are not obvious, the high-frequency characteristic quantities show abnormalities. Due to the reasonable setting of the weight coefficients, the detection device issues a warning in a timely manner. After receiving the warning, the user sends the vehicle to a professional repair shop for inspection in a timely manner and finds that there are some minor problems inside the lithium battery. After timely treatment, the further deterioration of the problem is avoided, effectively ensuring the safe use of the lithium battery and at the same time extending the service life of the battery.

[0074] In other embodiments, it further includes an infrared sensor for acquiring an infrared image of the lithium battery and sending it to the intelligent processing module.

[0075] After receiving the infrared image, the intelligent processing module performs a series of image processing tasks: accurately identifying the target features from the complex infrared image. The target features are the surface of the lithium battery, which contains important information about the temperature distribution of the lithium battery. The intelligent processing module will divide the target features into M feature regions for a more in-depth analysis of each local situation. For each feature region, the intelligent processing module calculates its heat value.

[0076] Based on the heat values of these M feature regions, the intelligent processing module will further calculate two important data. One is to calculate the dispersion coefficient of the heat values of the M feature regions and use it as the heat distribution value. The dispersion coefficient can reflect the degree of dispersion of heat among the various feature regions, which means it can reflect the uniformity of the heat distribution on the battery surface. The other is to calculate the total heat value of the M feature regions and count the "total temperature" of all small regions on the entire map.

[0077] After obtaining the heat distribution value and the total heat value, the intelligent processing module will normalize them and then calculate the infrared heat value through weighted calculation. The purpose of normalization is to eliminate the dimensional differences between different data, enabling these data to be compared and calculated under a unified standard. Weighted calculation gives corresponding weights according to the importance of different data, thus obtaining a comprehensive infrared heat value.

[0078] If the calculated infrared heat value is higher than the set threshold, it indicates that there may be potential problems such as local overheating in the battery. At this time, the intelligent processing module will take further measures. After normalizing the bulging thickness, ultrasonic signal, high-frequency characteristic quantity, and infrared temperature value, it will calculate the hidden danger value through weighted calculation. If the hidden danger value exceeds the set hidden danger threshold, it can be determined that there is an early fault point in the measured lithium battery separator.

[0079] Suppose in an energy storage system of a large data center, a large number of lithium batteries are used for power storage. One of the lithium batteries numbered B-03 usually seems to operate normally. The infrared sensor in the detection system continuously obtains the infrared image of this lithium battery and sends it to the intelligent processing module.

[0080] After the intelligent processing module identifies the target features from the infrared image, it divides them into M = 10 feature regions. Through calculation, the heat values of each feature region are Q1 = 30°C, Q2 = 32°C, Q3 = 31°C, Q4 = 35°C, Q5 = 33°C, Q6 = 30°C, Q7 = 32°C, Q8 = 38°C, Q9 = 31°C, Q10 = 32°C respectively. Through calculation, the total heat value is 334°C, and the heat distribution value is 0.12 after calculation by the coefficient of variation; among them, the calculation method of the coefficient of variation is the standard deviation divided by the average value. After normalization and weighted calculation, the infrared heat value is 0.7, and the set threshold is 0.5, indicating that the infrared heat value is higher than the set threshold.

[0081] At this time, the intelligent processing module normalizes the bulging thickness, ultrasonic signal, high-frequency characteristic quantity, and infrared temperature value and then calculates the hidden danger value through weighted calculation. Suppose the normalized value of the bulging thickness is 0.2, the normalized value of the ultrasonic signal is 0.3, the normalized value of the high-frequency characteristic quantity is 0.2, and the normalized value of the infrared temperature value is 0.7, and their respective weights are 0.2, 0.2, 0.2, 0.4 respectively. Then the hidden danger value is 0.2×0.2 + 0.2×0.3 + 0.2×0.2 + 0.4×0.7 = 0.42, and the set hidden danger threshold is 0.4. The hidden danger value exceeds the set hidden danger threshold. Therefore, it is determined that there is an early fault point in the separator of this lithium battery numbered B-03.

[0082] By introducing infrared sensors to obtain infrared images of lithium batteries, the temperature dimension is added to the detection, which is of great significance. Thermal runaway of lithium batteries is usually accompanied by abnormal temperature changes. Infrared images can intuitively reflect the temperature distribution on the battery surface. Just like the lithium battery B-03 in the above example, although there may be no obvious abnormalities in terms of bulging thickness, ultrasonic signals and high-frequency features, local overheating was discovered through infrared images. The intelligent processing module identifies target features from infrared images, divides feature areas and calculates heat values, and then obtains heat distribution values ​​and total heat values. These data can accurately present the heat status inside the battery.

[0083] In the process of processing infrared images, the target features are divided into M feature areas, the heat values ​​are calculated separately, and then the discrete coefficient is calculated to obtain the heat distribution value. This refined analysis method can capture the tiny abnormal heat distribution on the battery surface. The heat distribution value reflects the uniformity of heat on the battery surface. Even if the total heat value does not reach an obviously abnormal level, if the heat distribution value is abnormal, it may also indicate that there are local thermal hazards inside the battery, such as local heating caused by a local short circuit of the diaphragm. In the above example, the temperature of the feature area Q8 reaches 38°C, which is significantly higher than other areas. This may be the local heating phenomenon caused by a local short circuit of the diaphragm. Compared with only monitoring the overall temperature, this method can more keenly detect early and subtle thermal anomalies, warn of potential thermal runaway risks in advance, and buy time for timely measures to prevent the deterioration of faults.

[0084] The heat distribution value and the total heat value are normalized and then weighted to calculate the infrared heat value. Then, the infrared heat value, bulging thickness, ultrasonic signal, and high-frequency feature are normalized and weighted to calculate the hidden danger value, thus realizing the fusion judgment of multiple parameters. Different detection indicators reflect the status of lithium batteries from different angles. The bulging thickness reflects the physical deformation, the ultrasonic signal and high-frequency feature reflect the change of electrical performance, and the infrared temperature value reflects the thermal state. Multi-parameter fusion can comprehensively and comprehensively evaluate the status of lithium battery separators, make full use of the advantages of each indicator, make up for the limitations of single indicator detection, and enhance the comprehensiveness and reliability of detection. Whether it is thermal anomalies caused by electrical faults or physical or electrical changes caused by thermal problems, they can be detected and judged more effectively.

[0085] Whether to include the infrared temperature value in the calculation of the hidden danger value is determined based on whether the infrared heat value is higher than the set threshold. This adaptive judgment strategy improves the reliability of detection. When the battery temperature is normal, the original detection indicators are mainly used for fault judgment, reducing unnecessary calculations and interference; while when the infrared heat value is abnormal, the infrared temperature value is timely taken into consideration and the judgment strategy is adjusted to more accurately evaluate the fault risk. This flexible adjustment method according to the actual situation avoids misjudgment or missed judgment that may occur due to a fixed judgment mode, making the detection result more in line with the actual operating state of the lithium battery. In the above example, if the infrared heat value does not exceed the set threshold, the infrared temperature value will not be included in the calculation of the hidden danger value, but the state of the lithium battery will continue to be judged by the bulging thickness, ultrasonic signal and high-frequency characteristic quantity, which can improve the detection efficiency and ensure the accuracy of detection at the same time.

[0086] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present application.

Claims

1. A lithium battery thermal runaway detection device, characterized in that, It includes a supporting power supply, a high-frequency alternating magnetic field coil, a thickness sensor, a broadband ultrasonic sensor, a high-frequency current transformer and an intelligent processing module; The high-frequency alternating magnetic field coil is electrically connected to the supporting power supply through a high-frequency transformer and is used to emit high-frequency electromagnetic waves to output a test voltage at both ends of the lithium battery separator; The thickness sensor is electrically connected to the supporting power supply and is used to detect the bulging thickness of the lithium battery due to the vaporization of dielectric liquefied water; The broadband ultrasonic sensor is electrically connected to the supporting power supply and is used to receive the ultrasonic signal generated by the breakdown discharge of the measured lithium battery separator under the action of the high-frequency electromagnetic wave; The high-frequency current transformer is electrically connected to the supporting power supply and is clamped to an output terminal of the measured lithium battery for collecting high-frequency characteristic quantities in the battery current signal; The intelligent processing module is used to receive the bulging thickness, the ultrasonic signal and the high-frequency characteristics, and calculate a fault value by weighted calculation after normalizing the bulging thickness, the ultrasonic signal and the high-frequency characteristic quantities; if the fault value exceeds the set fault threshold, it is determined that there is an early fault point in the measured lithium battery separator.

2. The lithium battery thermal runaway detection device according to claim 1, wherein, The fault value = a1 × normalized value of bulging thickness + a2 × normalized value of ultrasonic signal + a3 × normalized value of high-frequency characteristic quantity; where a1 is the weight coefficient of the normalized value of bulging thickness, a2 is the weight coefficient of the normalized value of ultrasonic signal, and a3 is the weight coefficient of the normalized value of high-frequency characteristic quantity; among them, the normalized value of bulging thickness, the normalized value of ultrasonic signal and the normalized value of high-frequency characteristic quantity are calculated by the minimum-maximum normalization method.

3. The lithium battery thermal runaway detection device according to claim 2, wherein The intelligent processing module adjusts the size of a1 according to the usage time of the lithium battery; the shorter the usage time, the smaller a1; the longer the usage time, the larger a1.

4. The lithium battery thermal runaway detection device according to claim 2, wherein, The intelligent processing module dynamically adjusts the weight coefficient according to the charge and discharge times N of the lithium battery; when N < the set charge and discharge times threshold N1, a1 = k1 × N / N1; a2 = 1 - k2 × N / N1, a3 = k3, where k1, k2 and k3 are constants.

5. The lithium battery thermal runaway detection device according to claim 2, characterized in that, It also includes an infrared sensor for obtaining an infrared image of the lithium battery and sending it to the intelligent processing module; The intelligent processing module identifies target features from the infrared image; The target features are divided into M feature regions, and the heat value of each feature region is calculated; Calculating the discrete coefficient of the heat values of the M feature regions as the heat distribution value, and calculating the total heat value of the M feature regions; Calculating an infrared heat value by weighted calculation after normalizing the heat distribution value and the total heat value; If the infrared heat value is higher than the set threshold, then calculate a hidden danger value by weighted calculation after normalizing the bulging thickness, the ultrasonic signal, the high-frequency characteristic quantity and the infrared temperature value; if the hidden danger value exceeds the set hidden danger threshold, it is determined that there is an early fault point in the measured lithium battery separator.

6. The lithium battery thermal runaway detection device according to claim 1, characterized in that There are a pair of thickness sensors, which are installed on both sides of the lithium battery; the thickness sensors are electrically connected to a thickness detection meter for displaying the change in the battery thickness detected by the thickness sensors.

7. The lithium battery thermal runaway detection device according to claim 1, characterized in that, The signal output terminal of the broadband ultrasonic sensor is electrically connected to a band-pass filter, and the band-pass filter is electrically connected to an oscilloscope. The band-pass filter is used to screen signals within a specific frequency range, and the oscilloscope is used to display the signal waveform detected by the broadband ultrasonic sensor after being screened by the band-pass filter.

8. The lithium battery thermal runaway detection device according to claim 1, wherein, The signal output terminal of the high-frequency current transformer is electrically connected to a high-pass filter, and the high-pass filter is electrically connected to a high-frequency ammeter; the high-pass filter is used to filter out low-frequency interference signals, and the high-frequency ammeter is used to display the high-frequency current characteristic quantity of the high-frequency current transformer after being filtered by the high-pass filter.

Images