Early fire detection method for lithium battery and electric vehicle battery

Abstract

The present invention discloses a method for early detection of a lithium battery and an electric vehicle battery fire, comprising: a first step (S110) of generating time-series data for each off-gas component by detecting common off-gas components that occur in a first vent stage due to an initial fire of two or more types of electric vehicle batteries in a time-series manner through a composite off-gas detection sensor structure (100); a second step (S120) of determining whether a final fire has occurred by analyzing the time-series data in a pattern through a fire judgment algorithm (200) according to a pre-set minimum gas detection condition for each off-gas component and a fire confirmation condition in a parallel cross method; and a third step (S130) of generating and propagating warning information according to the determination of the final fire occurrence through a warning system (300), thereby increasing the detection accuracy of off-gases that occur before signs of fire in the lithium battery and electric vehicle battery.

Description

Technology Field

[0001] The present invention relates to a method for early detection of fire in lithium batteries and electric vehicle batteries, which improves the detection accuracy of off-gas generated in the primary vent before the secondary vent of the lithium battery and electric vehicle battery. Background Technology

[0003] Conventional electric vehicle fire detection technology primarily relies on detection technology that activates after visible phenomena, such as flames, smoke, or high temperatures, have occurred.

[0004] Therefore, there was a limitation in that it could not be properly detected at the first vent point, which is the pre-fire stage of an electric vehicle battery, that is, at the initial abnormal sign stage where invisible harmful off-gases are released as the internal pressure of the battery cell rises.

[0005] Meanwhile, commonly used flame detectors, smoke detectors, and heat detectors are not suitable for early response to electric vehicle battery fires because they all activate only after visible and physical signs of fire have occurred.

[0006] Furthermore, sensors or detectors attached to the underside of vehicles are easily exposed to external factors such as dust, moisture, and contamination due to the characteristics of the parking lot environment, and there is a high possibility that their detection performance will deteriorate over the long term.

[0007] Meanwhile, video-based smoke detection methods are susceptible to influences such as humidity, illumination, and background color, leading to a high probability of false positives or missed detections; furthermore, they have limitations, particularly during dry seasons, where the visibility of the smoke particles themselves is reduced.

[0008] Although early detection systems utilizing off-gas detection sensors have been applied in some energy storage (ESS) sectors, they are not optimized for electric vehicle parking environments such as open parking spaces and fail to account for the differences in thermal runaway characteristics among different types of electric vehicle batteries.

[0009] In addition, there is a problem in that the gas detection sensor itself reacts sensitively to changes in ambient temperature or humidity, resulting in insufficient automatic correction capabilities based on the environment.

[0010] As such, conventional fire detection technology has structural limitations in early warning and preemptive response because it is difficult to detect electric vehicle battery fires in their precursor stages and fails to account for various environmental changes and battery characteristics.

[0011] Therefore, there is a need for more precise, environment-adaptive, and integrated early detection technology that does not distinguish between battery types. Prior art literature

[0013] Korean Published Patent Application No. 10-2023-0123216 (Device and method for monitoring fire during charging at an electric vehicle charging station, Aug. 23, 2023) The problem to be solved

[0014] The technical problem that the concept of the present invention aims to solve is to provide a method for early detection of fires in lithium batteries and electric vehicle batteries, which can detect and predict fires caused by primary venting early before secondary venting by parallel cross-verifying common off-gases through a standalone composite off-gas detection sensor structure and a single AI-based fire judgment algorithm, even if the timing of occurrence and gas concentration of each off-gas differ depending on the battery type or parking space environment. means of solving the problem

[0016] To achieve the aforementioned objective, an embodiment of the present invention provides a method for early detection of a lithium battery and an electric vehicle battery fire, comprising: a first step of generating time-series data for each off-gas component by detecting a common off-gas component that occurs in a primary venting stage due to an initial fire of two or more types of electric vehicle batteries in a time-series manner through a composite off-gas detection sensor structure; a second step of determining whether a final fire has occurred by pattern-analyzing the time-series data through a fire judgment algorithm according to a preset minimum gas detection condition for each off-gas component and a fire confirmation condition of a parallel cross method; and a third step of generating and propagating warning information according to the determination of the final fire occurrence through a warning system.

[0018] Here, the electric vehicle battery may include a lithium-ion battery and a lithium iron phosphate battery.

[0020] In addition, in the first step above, H2, CO, and HF that are commonly generated in the lithium-ion battery and the lithium iron phosphate battery can be detected to generate the time series data.

[0022] In addition, in the second step above, the minimum gas detection conditions in the lithium-ion battery are CO ≥ 20 ppm and HF ≥ 6 ppm, and the minimum gas detection conditions in the lithium iron phosphate battery are H2 ≥ 50 ppm and HF ≥ 6 ppm, and the threshold value of the minimum gas detection conditions can be adjusted upward or downward by 10% to 20% depending on the humidity conditions.

[0024] In addition, in the second step above, the fire confirmation condition consists of a single detection condition of HF ≥ 6 ppm or H2 ≥ 100 ppm, and a simultaneous detection condition of H2 ≥ 50 ppm and CO ≥ 30 ppm, H2 ≥ 50 ppm and HF ≥ 6 ppm, or CO ≥ 30 ppm and HF ≥ 6 ppm, and the threshold value of the fire confirmation condition can be adjusted upward or downward by 10% to 20% depending on the humidity condition.

[0026] In addition, the above-described composite off-gas detection sensor structure may include a main body having a sealed structure formed by stacking an SMPS and a control board in an internal space and mounted on a raceway installed on the ceiling of a parking lot; an intake fan coupled to the bottom of the main body, which forcibly draws in air from the bottom and discharges it to the side under the control of the control board; a ring-shaped LED module coupled to the bottom of the intake fan; a cover made of a transparent or translucent material coupled to the bottom of the LED module, having an induction hole formed inside to guide external air from the bottom to the intake fan, and surrounding the induction hole and irradiating warning light from the LED module to the outside under the control of the control board; and a composite off-gas detection sensor coupled to the upper center of the induction hole, which detects off-gases of H2, CO, and HF generated during thermal runaway of the electric vehicle battery from the air forcibly drawn in through the induction hole by the operation of the intake fan and transmits them to the control board.

[0028] In addition, the SMPS and the control board can be cooled through air forcibly drawn in by the intake fan.

[0030] In addition, the control board can drive the intake fan when the composite off-gas detection sensor detects off-gas from an electric vehicle parked in an electric vehicle parking space.

[0032] In addition, when the control board detects an electric vehicle parked in the parking space, it drives the intake fan in the forward direction to draw in outside air, and when it does not detect an electric vehicle in the parking space, it drives the intake fan in the reverse direction for a certain period of time to discharge foreign substances attached to the induction hole, the blade of the intake fan, and the composite off-gas detection sensor to the outside. Effects of the invention

[0034] According to the present invention, common off-gases such as H₂, CO, and HF generated in the first vent stage of an electric vehicle battery are detected in a time-series manner, and the presence of fire is determined early through an AI-based fire judgment algorithm, thereby enabling a preemptive response in the first vent stage, that is, before the second explosion caused by the second vent. Furthermore, accurate fire prediction is possible with a single sensor and a single algorithm regardless of the battery type (NCM / LFP), false detections can be reduced and detection reliability increased through an automatic threshold correction function based on environmental conditions such as changes in humidity, and precise pattern analysis is performed by considering both the rate of increase and duration of gas concentration, thereby providing an early fire detection technology that can be extended and applied to various electric storage devices such as electric motorcycles and ESS as well as electric vehicles.

[0035] In addition, by forcibly drawing in air through an intake fan, it is possible to preemptively detect and warn of off-gases generated during thermal runaway before a fire in an electric vehicle battery, thereby visually propagating the toxic off-gas detection zone to block access and perform fire spread suppression work, and to notify access blockage and instruct the wearing of protective equipment. Furthermore, it is possible to ensure electrical stability by preventing overheating of the main body formed by a sealed structure through the intake fan, and to make maintenance easier by incorporating a composite off-gas detection sensor into the cover so that it can be replaced. Brief explanation of the drawing

[0037] FIG. 1 illustrates a flowchart of a method for early detection of fire in lithium batteries and electric vehicle batteries according to an embodiment of the present invention. Figure 2 illustrates an early detection method for lithium battery and electric vehicle battery fires according to Figure 1. Figure 3 illustrates a change in off-gas concentration applied to the early fire detection method for lithium batteries and electric vehicle batteries of Figure 1. Figure 4 illustrates the application of a composite off-gas detection sensor structure applied to the early fire detection method for lithium batteries and electric vehicle batteries of Figure 1. Figure 5 illustrates the composite off-gas detection sensor structure of Figure 4. Figure 6 illustrates a cross-sectional structure of the composite off-gas detection sensor structure of Figure 5. FIGS. 7 and FIGS. 8 illustrate exploded views of the composite off-gas detection sensor structure of FIG. 5, respectively. Figure 9 illustrates the airflow by the composite off-gas detection sensor structure of Figure 5. FIG. 10 illustrates light emission by the composite off-gas detection sensor structure of FIG. 5. Specific details for implementing the invention

[0038] Hereinafter, embodiments of the present invention having the aforementioned features will be described in more detail with reference to the attached drawings.

[0040] The method for early detection of a lithium battery and electric vehicle battery fire according to an embodiment of the present invention comprises: a first step (S110) of generating time-series data for each off-gas component by detecting common off-gas components that occur in a first vent stage due to an initial fire of two or more types of electric vehicle batteries in a time-series manner through a composite off-gas detection sensor structure (100); a second step (S120) of determining whether a final fire has occurred by analyzing the time-series data in a pattern through a fire judgment algorithm (200) according to a pre-set minimum gas detection condition for each off-gas component and a fire confirmation condition in a parallel cross method; and a third step (S130) of generating and propagating warning information according to the determination of a final fire occurrence through a warning system (300), thereby increasing the detection accuracy of off-gases that occur before signs of fire in the lithium battery and electric vehicle battery in the first vent before the second vent.

[0042] Hereinafter, with reference to FIGS. 1 to 3, the method for early detection of fire in a lithium battery and an electric vehicle battery of the above-described configuration will be specifically described as follows.

[0044] First, the first step (S110) is a step for detecting major off-gases, and through a standalone electrochemical composite off-gas detection sensor structure (100), common off-gas components generated in the primary vent stage due to an initial fire of two or more electric vehicle batteries are detected in a time series, and time series data for each off-gas component is generated and analyzed by a fire judgment algorithm (200).

[0045] That is, in the first step (S110), H2, CO, and HF, which are commonly generated in lithium-ion batteries and lithium iron phosphate batteries, are each detected to generate time series data consisting of the generation temperature, generation time, single gas concentration, combined gas concentration, and emission amount for each off-gas component that differs for each electric vehicle battery.

[0046] For example, representative electric vehicle batteries include lithium-ion batteries with a relatively high risk of thermal runaway, namely NCM (Nickel Cobalt Manganese) oxide lithium-ion batteries, and lithium iron phosphate (LFP; LiFePO₄, Lithium Iron Phosphate) batteries with a relatively low risk of thermal runaway.

[0047] Meanwhile, both batteries commonly generate colorless and odorless H2, colorless and odorless CO, and colorless HF with a pungent odor at high concentrations during the first vent, but the generation temperature, generation time, single gas concentration, combined gas concentration, and emission amount may differ for each off-gas component.

[0048] Here, the primary vent refers to an initial abnormal sign stage corresponding to a fire precursor, where the pressure release device (vent cap) of the battery cell opens when the internal pressure of the battery cell exceeds a certain level, and invisible off-gas is released, causing a risk to the human body due to harmful gases.

[0049] For example, in the case of a lithium-ion battery, the detection order based on the timing of the generation of major off-gases due to the primary vent is [CO ->, H2 -> VOC (Volatile Organic Compounds) -> HF], and in the case of a lithium iron phosphate battery, the detection order based on the timing of the generation of major off-gases due to the primary vent may be [H2 ->, CO -> HF].

[0051] Next, the second step (S120) is a step of determining the occurrence of a fire by performing AI-based time-series pattern analysis, wherein the time-series data for each off-gas component is pattern-analyzed through an AI-based fire judgment algorithm (200) embedded in a single composite off-gas detection sensor structure (100) or a separate external control server (not shown), and the final fire occurrence is predicted and determined early before the secondary vent that causes thermal runaway and explosion, according to the pre-set minimum gas detection conditions for each individual off-gas component corresponding to the actual fire occurrence and the fire confirmation conditions of the parallel cross method for two off-gas components.

[0052] For example, in the second stage (S120), through a single AI-based fire judgment algorithm (200) for time series data of each off-gas component from a single composite off-gas detection sensor structure (100), a primary vent accompanied by off-gas leakage from a different electric vehicle battery can be detected and predicted early, thereby identifying initial abnormal signs before a secondary vent that causes thermal runaway and explosion, rather than responding to the electric vehicle fire afterward, and thus enabling early fire detection at the only point in time when suppression is possible during a rapid response.

[0053] Here, secondary vent may refer to a stage where the internal temperature continues to rise even after the pressure is released from the battery cell, causing visible flames, smoke, and explosions due to the decomposition and ignition of the electrolyte, leading to the spread of fire, damage to surrounding structures, and casualties.

[0055] Specifically, according to the definition of the change in off-gas concentration during the first vent for each battery exemplified in FIG. 3, in the second step (S120), the minimum gas detection conditions in the lithium-ion battery may be CO ≥ 20 ppm and HF ≥ 6 ppm, and the minimum gas detection conditions in the lithium iron phosphate battery may be H2 ≥ 50 ppm and HF ≥ 6 ppm.

[0056] In addition, in the second stage (S120), the fire confirmation condition may consist of a single detection condition of {HF ≥ 6 ppm} or {H2 ≥ 100 ppm}, and a simultaneous detection condition (gas combination pattern) of {H2 ≥ 50 ppm and CO ≥ 30 ppm}, {H2 ≥ 50 ppm and HF ≥ 6 ppm}, or {CO ≥ 30 ppm and HF ≥ 6 ppm}.

[0057] That is, the AI-based fire judgment algorithm (200) can analyze the time series data from the composite off-gas detection sensor structure (100) according to the gas detection minimum condition and fire confirmation condition according to the primary vent of the lithium-ion battery and lithium iron phosphate battery mentioned above, and can detect the fire corresponding to the primary vent that simultaneously satisfies the gas detection minimum condition, which is a threshold, and the parallel crossing condition for each off-gas component, regardless of the time of off-gas occurrence according to the type of electric vehicle battery.

[0058] Here, the AI-based fire judgment algorithm (200) is an AI-based analysis model built by pre-learning a dataset of time series data by different off-gas components for each battery built with big data and a dataset of whether fire is detected early based on the primary vent identification result according to the time series data. It can predict and output whether an initial fire occurs corresponding to a primary vent that is a fire precursor by taking time series data provided in real time from a composite off-gas detection sensor (100) as input.

[0060] Through this, even if the types of batteries are different, a single AI-based fire judgment algorithm (200) can analyze the pattern matrix for each off-gas combination to detect fire early based on whether a specific single condition or a specific simultaneous detection condition is satisfied.

[0062] Meanwhile, the threshold value of the aforementioned minimum gas detection condition can be adjusted by varying it up or down by 10% to 20% depending on the humidity condition, and the threshold value of the fire confirmation condition can be adjusted by varying it up or down by 10% to 20% depending on the humidity condition.

[0063] For example, the AI-based fire judgment algorithm (200) may ignore CO alone detection as a false detection condition and immediately generate warning information upon H2 (≥ 50 ppm) alone detection, and the corresponding threshold values ​​for the minimum gas detection condition and the fire confirmation condition for each off-gas component may vary according to humidity information in the area where the electric vehicle is parked, for example, when there is high humidity such as 90% relative humidity, each threshold value for the minimum gas detection condition and the fire confirmation condition may be adjusted upward by 10% to 20%, thereby calibrating the sensor sensitivity of the composite off-gas detection sensor structure (100) due to high humidity, and reducing the possibility of false detection due to gas diffusion and gas concentration distribution that are underestimated due to high humidity and increasing accuracy.

[0065] Meanwhile, the AI-based fire judgment algorithm (200) may further improve the accuracy of early fire detection and prediction by applying the concentration increase rate and time duration of each off-gas component as additional judgment conditions.

[0067] Next, the third step (S130) is a step for propagating the actual occurrence of a fire based on early detection. By generating and propagating warning information based on the final fire occurrence judgment by the AI-based fire judgment algorithm (200) through the warning system (300), it is possible to respond quickly and suppress the fire early before thermal runaway and explosion.

[0069] As illustrated in FIG. 2, the aforementioned method for early detection of lithium battery and electric vehicle battery fires can be applied in the same way to batteries other than electric vehicles (a), such as battery manufacturing facilities, electric bicycles, electric scooters, electric bikes, and facilities equipped with ESS (b).

[0071] Meanwhile, with reference to FIGS. 4 to 10, a composite off-gas detection sensor structure applied to the method for early detection of a lithium battery and electric vehicle battery fire according to an embodiment of the present invention is described in detail as follows.

[0072] That is, the composite off-gas detection sensor structure comprises: a main body (110) of a sealed structure formed by stacking an SMPS (111) and a control board (112) in an internal space and mounted on a raceway (10) installed on the ceiling of a parking lot; an intake fan (120) coupled to the bottom of the main body (110) and forcibly sucking in air from the bottom and discharging it to the side under the control of the control board (112); a ring-shaped LED module (130) coupled to the bottom of the intake fan (120); a cover (140) of a transparent or translucent material coupled to the bottom of the LED module (130), having an induction hole (141) formed inside to guide external air from the bottom to the intake fan (120), surrounding the induction hole (141) and irradiating warning light from the LED module (130) to the outside under the control of the control board (112); and a cover (140) coupled to the upper center of the induction hole (141) and driven by the intake fan (120). It includes a composite off-gas detection sensor (150) that detects off-gases of H2, CO, and HF generated during thermal runaway of the electric vehicle (20) battery from air forcibly sucked in through the induction hole (141) and transmits them to the control board (112), thereby enabling the preemptive detection and warning of off-gases generated during thermal runaway of the electric vehicle battery before a fire.

[0074] Specifically, first, the main body (110), with reference to FIGS. 4 to 8, is formed as a cylindrical sealed structure for waterproofing and anti-fouling purposes and is mounted on a raceway (10) installed on the ceiling of a parking lot. It is formed by stacking an SMPS (Switching Mode Power Supply) (111) that supplies power to a control board (112) in a sealed internal space, and a control board (112) that controls an intake fan (120), an LED module (130), and a composite off-gas detection sensor (150).

[0075] Here, as illustrated in FIGS. 4 to 6, the main body (110) may be fixed and mounted on the raceway (10) by means of a coupling frame (113) rotatably coupled to the top, or may be mounted on a guide rail (not shown) formed on the raceway (10) through a guide rail block (114) formed on the upper surface of the main body (110), and a power line branched from the raceway (10) may be connected to a gland cable connector (115) to supply power to the SMPS (111).

[0076] Additionally, as illustrated in FIG. 4, the main body (110) may be mounted on the top of an electric vehicle parking space (a) or between adjacent electric vehicle parking spaces (b), as needed.

[0077] Additionally, as illustrated in FIGS. 5 to 8, an arc-shaped exhaust guide (116) is arranged in a circular shape to cover the outer edge of the intake fan (120) at the bottom of the main body (110) to discharge air between the exhaust guides (116), and the exhaust guides (116) act as cooling fins and are formed in an arc shape to increase the contact area with air, thereby allowing heat conducted from the SMPS (111) and the control board (112) to the exhaust guides (116) to be cooled by air and efficiently dissipated.

[0079] Next, the intake fan (120) is equipped with a two-ball bearing fan, and with reference to FIGS. 6 and 8, it is coupled to the bottom of the main body (110), and as illustrated in FIG. 9, it is controlled by a control board (112) to forcibly draw in air around the electric vehicle (20) from the bottom and discharge it to the side, thereby minimizing off-gas concentration variation and shortening response time.

[0080] Here, the SMPS (111) and the control board (112) are cooled by air forcibly sucked in by the intake fan (120), thereby preventing overheating of the main body (110) formed with a sealed structure and ensuring electrical stability.

[0082] Meanwhile, the control board (112) can drive the intake fan (120) when the composite off-gas detection sensor (150) detects off-gas from the electric vehicle (20) parked in the electric vehicle parking space.

[0083] Alternatively, when the control board (112) detects an electric vehicle (20) that has entered and is parked in a parking space through an object recognition sensor (not shown), it drives the intake fan (120) in a forward direction to draw in external air and constantly detect whether off-gas is generated. When it does not detect an electric vehicle (20) in the parking space, it drives the intake fan (120) in a reverse direction for a certain period of time to discharge foreign substances attached to the induction hole (141), the blade of the intake fan (120), and the composite off-gas detection sensor (150) to the outside, thereby suppressing the inflow of foreign substances such as dust and improving sensing performance to minimize false detection of off-gas.

[0085] Next, the LED module (130), with reference to FIGS. 6 to 8, is coupled in a ring shape to the bottom of the intake fan (120), and a plurality of RGB LED elements (131) are arranged on the bottom surface, so that when off-gas is detected by the composite off-gas detection sensor (150) as exemplified in FIG. 10 (b), it emits a red warning light to visually propagate the toxic off-gas detection zone to block access and perform fire spread suppression work, and can also propagate a warning voice through a speaker (not shown) to inform access blockage and inform to wear protective gear.

[0087] Next, the cover (140), with reference to FIGS. 5 to 10, is formed in a cylindrical shape at the bottom of the LED module (130) and coupled thereto, and a guide hole (141) is formed inside to guide external air from the bottom to the intake fan (120). The guide hole (141) is surrounded and the warning light of the LED module (130) is directed outward by the control of the control board (112), thereby allowing the light from the LED module (130) to be diffused outward and the external air to be guided to the intake fan (120).

[0088] Here, as illustrated in FIG. 9, the induction hole (141) is formed with a tapered cross-sectional structure that narrows toward the intake fan (120) to implement the Bernoulli effect, thereby increasing the flow velocity to maximize air suction power and smoothly forcibly sucking in and detecting highly toxic HF gas that is heavier than air and sinks to the bottom, and is formed with a structure that expands in a tapered shape toward the cross-section of the main body part (110) formed on the outside of the intake fan (120) to increase the discharge pressure.

[0090] Next, the composite off-gas detection sensor (150), with reference to FIGS. 5 to 9, is formed at the top of the induction hole (141) and is coupled to the central bottom surface of the mounting plate (151) which is open to communicate with the induction hole (141). This allows the off-gas of H2, CO, and HF to be detected early through the electrochemical composite off-gas detection sensor (150) and transmitted to the control board (112) before the explosion caused by the secondary vent and thermal runaway, which occurs when the battery, such as the lithium-ion battery of the electric vehicle (20), is forcibly sucked in through the induction hole (141) by the driving of the intake fan (120). This enables the control board (112) to drive the LED module (130) to emit a red warning light or propagate a warning sound.

[0091] Accordingly, the composite off-gas detection sensor (150) is coupled to the cover (140) so as to be replaceable, making maintenance easier, and may be replaced with a sensor that detects other gases other than H2, CO, and HF gases generated from the lithium-ion battery, such as CO2.

[0093] Meanwhile, the aforementioned composite off-gas detection sensor structure can be applied to battery manufacturing facilities or facilities equipped with an Energy Storage System (ESS) to preemptively detect off-gas in such facilities and fundamentally prevent it from escalating into a large-scale fire.

[0095] In addition, this embodiment may detect abnormal temperatures corresponding to off-gas generation and electric vehicle fires after double cross-monitoring through the detection of H2, CO, and HF gases by an off-gas detection sensor structure and image analysis by a thermal imaging camera (not shown) that photographs the parking surface, and may also implement a response to prevent fire spread by using an electric vehicle fire transfer barrier.

[0097] Accordingly, by configuring the method for early detection of lithium battery and electric vehicle battery fires as described above, even if the timing of occurrence and gas concentration of each off-gas differ depending on the battery type or parking space environment, common off-gases can be verified in parallel cross-verified through a standalone composite off-gas detection sensor structure and a single AI-based fire judgment algorithm to detect and predict fires caused by primary venting before secondary venting, thereby reducing the possibility of false detection and increasing prediction accuracy. Furthermore, by forcibly drawing in air through an intake fan, off-gases generated during thermal runaway before a fire in the electric vehicle battery can be preemptively detected and warned of, allowing the toxic off-gas detection zone to be visually propagated to block access and perform fire spread suppression work. Additionally, it is possible to notify access blockage and inform the wearing of protective equipment. Furthermore, electrical stability can be ensured by preventing overheating of the main body formed as a sealed structure by the intake fan, and maintenance can be made easier by combining the composite off-gas detection sensor to the cover so that it can be replaced.

[0099] The embodiments described in this specification and the configurations illustrated in the drawings are merely the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention; therefore, it should be understood that various equivalents and modifications that can replace them may exist at the time of filing this application. Explanation of the symbols

[0101] S110: Main off-gas detection step S120: Fire occurrence determination stage S130: Fire spread stage 10 : Raceway 20 : Electric Vehicle 100 : Combined off-gas detection sensor 110 : Main body 111 : SMPS 112: Control board 113: Connecting frame 114 : Guide rail block 115 : Gland cable connector 116 : Exhaust guide section 120 : Intake fan 130 : LED module 131 : RGB LED element 140 : Cover 141 : Guide hole 150 : Combined off-gas detection sensor 151 : Mounting plate 200: Fire Detection Algorithm 300: Warning System

Claims

Claim 1 A first step of generating time-series data by each off-gas component by detecting common off-gas components occurring in the primary venting stage due to an initial fire of two or more types of electric vehicle batteries in a time-series manner through a composite off-gas detection sensor structure; a second step of determining whether a final fire has occurred by pattern-analyzing the time-series data through a fire judgment algorithm, based on a preset minimum gas detection condition for each off-gas component and a fire confirmation condition using a parallel intersection method; A method for early detection of lithium battery and electric vehicle battery fire, comprising: a third step of generating and disseminating warning information based on the final determination of fire occurrence through a warning system; wherein the electric vehicle battery includes a lithium-ion battery and a lithium iron phosphate battery; wherein H2, CO, and HF commonly occurring in the lithium-ion battery and the lithium iron phosphate battery are detected to generate the time series data; wherein the minimum gas detection condition is above a specific concentration of CO and HF and above a specific concentration of H2 and HF; and wherein the fire confirmation condition is formed by a parallel cross method of a single detection condition of HF or H2 and a simultaneous detection condition of H2 and CO, H2 and HF, or CO and HF, thereby early detecting and predicting a primary vent accompanied by off-gas leakage from different electric vehicle batteries, and identifying initial abnormal signs before a secondary vent causing thermal runaway and explosion, rather than a post-fire response to the electric vehicle fire. Claim 2 delete Claim 3 delete Claim 4 A method for early detection of fire in lithium batteries and electric vehicle batteries according to claim 1, wherein in the second step, the minimum gas detection condition in the lithium-ion battery is CO ≥ 20 ppm and HF ≥ 6 ppm, the minimum gas detection condition in the lithium iron phosphate battery is H2 ≥ 50 ppm and HF ≥ 6 ppm, and the threshold value of the minimum gas detection condition is corrected upward or downward by 10% to 20% depending on the humidity condition. Claim 5 A method for early detection of fire in lithium batteries and electric vehicle batteries according to claim 1, wherein in the second step, the fire confirmation condition comprises a single detection condition of HF ≥ 6 ppm or H2 ≥ 100 ppm, and a simultaneous detection condition of H2 ≥ 50 ppm and CO ≥ 30 ppm, H2 ≥ 50 ppm and HF ≥ 6 ppm, or CO ≥ 30 ppm and HF ≥ 6 ppm, and the threshold value of the fire confirmation condition is corrected upward or downward by 10% to 20% depending on the humidity condition. Claim 6 In claim 1, the composite off-gas detection sensor structure comprises: a main body portion having a sealed structure formed by stacking an SMPS and a control board in an internal space and mounted on a raceway installed on the ceiling of a parking lot; an intake fan coupled to the bottom of the main body portion, which forcibly draws in air from the bottom and discharges it to the side under the control of the control board; a ring-shaped LED module coupled to the bottom of the intake fan; a cover made of a transparent or translucent material coupled to the bottom of the LED module, having an induction hole formed inside to guide external air from the bottom to the intake fan, and surrounding the induction hole and irradiating warning light from the LED module to the outside under the control of the control board; and a composite off-gas detection sensor coupled to the upper center of the induction hole, which detects off-gases of H2, CO, and HF generated during thermal runaway of the electric vehicle battery from air forcibly drawn in through the induction hole by the operation of the intake fan and transmits them to the control board. Claim 7 A method for early detection of lithium battery and electric vehicle battery fire according to claim 6, characterized by cooling the SMPS and the control board through air forcibly sucked in by the intake fan. Claim 8 A method for early detection of lithium battery and electric vehicle battery fire according to claim 6, wherein the control board drives the intake fan when the composite off-gas detection sensor detects off-gas from an electric vehicle parked in an electric vehicle parking space. Claim 9 A method for early detection of lithium battery and electric vehicle battery fire according to claim 6, characterized in that when the control board detects an electric vehicle parked in a parking space, the intake fan is driven in a forward rotation to draw in external air, and when the electric vehicle in the parking space is not detected, the intake fan is driven in a reverse rotation for a certain period of time to discharge foreign substances attached to the induction hole, the blade of the intake fan, and the composite off-gas detection sensor to the outside.

Images