A thermal runaway detection and alarm device and method for a lithium-ion battery energy storage system

Abstract

The application provides a thermal runaway detection and alarm device and method for a lithium ion battery energy storage system, and relates to the technical field of safety prevention and control of lithium ion battery energy storage systems.The application samples and monitors the gas in the battery pack through a vacuum pump and a gas suction pipe of a smoke collection and distribution module, and records the number of the gas suction pipe through an electromagnetic distribution valve of the smoke collection and distribution module;the gas collected by the smoke collection and distribution module is sequentially transmitted to a smoke pretreatment module and a multi-composite smoke sensing module, the sensor value of the gas sample is transmitted to a sensing data storage module and stored, and the sensing data storage module is used for transmitting the sensor value to a multi-information fusion early warning module for early warning logic judgment;the sensor value of the gas sample sampled for more than or equal to two times is transmitted to the multi-information fusion early warning module, and an early warning signal is sent through the early warning threshold value given by a thermal runaway detection and alarm algorithm;and the thermal runaway detection and alarm of the lithium ion battery energy storage system is realized.

Description

Technical Field

[0001] This invention relates to the field of safety control technology for lithium-ion battery energy storage systems, and in particular to a thermal runaway detection and alarm device and method for lithium-ion battery energy storage systems. Background Technology

[0002] In recent years, lithium-ion battery energy storage systems have been widely used in the field of power energy storage, serving the grid connection of new energy sources such as wind and solar power, and meeting the urgent needs of building new power systems. However, frequent and difficult-to-control fires in lithium-ion battery energy storage systems have become a major pain point restricting their development and application. As an energy-containing element, lithium-ion batteries are prone to thermal runaway and accidents such as fires and explosions due to reduced insulation and overheating caused by electrical abuse, thermal abuse, or mechanical abuse during use.

[0003] With the technological upgrade of lithium-ion battery energy storage systems, the integration of battery systems has increased significantly, from 1MW in traditional air-cooled systems to 5MW in liquid-cooled systems. Simultaneously, the protection level of battery modules has increased, making it difficult for early signs such as smoke and gas to diffuse outside the battery modules, posing new challenges to early warning detection. Based on the fire protection concept of "prevention first, combined with firefighting," this paper proposes a battery module-level detection and alarm solution to address the prevention and control challenges of lithium-ion battery energy storage systems. A single detection and alarm device cannot meet the detection zoning requirements of current fire detection and alarm technical specifications, and there are difficulties in verification and validation. False alarms from detectors can trigger fire suppression systems, leading to secondary disasters and economic losses. Installing multiple detectors presents challenges such as high cost and maintenance difficulties, and the enclosed space within the battery pack and the accumulation of interference factors increase the likelihood of false alarms. Traditional fire detection and alarm technologies and products cannot meet the detection needs of highly integrated and enclosed battery packs. Meanwhile, due to the high integration of battery modules and the large number and capacity of batteries themselves, external detection methods suffer from significant latency. Furthermore, the dense series arrangement of batteries within a lithium-ion battery pack means that once a cell experiences thermal runaway, the fire spreads rapidly. The time interval between the appearance of external signs and the entire pack bursting into flames or even exploding is short, making early warning and prevention difficult and inevitably leading to major fire accidents and severe economic losses. To overcome these problems caused by the high integration of battery energy storage systems and improve the safety level of lithium-ion energy storage batteries, there is an urgent need for a technical method and device that can accurately and reliably detect and alarm thermal runaway in lithium-ion battery energy storage systems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a thermal runaway detection and alarm device and method for lithium-ion battery energy storage systems; aiming to solve the problems of alarm lag, false alarms, missed alarms, and difficulty in early location in existing detection and alarm technologies for novel electrochemical energy storage systems.

[0005] On the one hand, a thermal runaway detection and alarm device for a lithium-ion battery energy storage system includes an intake pipe, a flue gas collection and distribution module, a flue gas pretreatment module, a multi-composite flue gas sensing module, a sensing data storage module, and a multi-information fusion early warning module.

[0006] The suction pipe is a multi-channel suction pipe, with multiple suction pipes arranged in parallel. The output end of each suction pipe is connected to a battery pack under test, and the output end is connected to a smoke collection and distribution module, which is used to collect the smoke inside the battery under test and transmit it to the smoke collection and distribution module. The smoke absorbed by any one of the suction pipes is set as the sample gas.

[0007] The output end of the flue gas acquisition and distribution module is connected to the flue gas pretreatment module, and the output end of the flue gas pretreatment module is connected to the multi-composite flue gas sensing module. The flue gas acquisition and distribution module includes a vacuum pump and an electromagnetic distribution valve. The vacuum pump provides a suction force of not less than -80Kpa to the suction pipe. The electromagnetic distribution valve selects multiple suction pipes and outputs address code signals for positioning the battery pack under test.

[0008] The flue gas pretreatment module specifically includes flue gas pretreatment module a and flue gas pretreatment module b. The flue gas acquisition and distribution module transmits the acquired flue gas to flue gas pretreatment module a and flue gas pretreatment module b respectively according to the control logic. The flue gas is used for drying, filtering, stabilizing and equalizing the received flue gas. Flue gas pretreatment module a and flue gas pretreatment module b are mutually verified to avoid deviations caused by filtering by a single flue gas pretreatment module. The flue gas passing through flue gas pretreatment module a is recorded as a-, and the flue gas passing through flue gas pretreatment module b is recorded as b-. Each flue gas pretreatment module is equipped with an anemometer and a first pressure sensor at the input end to monitor the airflow pressure and velocity flowing into and out of the flue gas pretreatment module.

[0009] The control logic is to divide the collected flue gas into frequencies according to a set ratio based on the sampling time.

[0010] The multi-composite flue gas sensing module is used to analyze the component type, concentration, and symptom values ​​of the flue gas transmitted by the flue gas pretreatment module. It includes a smoke probe, a gas sensing module, a second pressure sensor, two sets of temperature sensors, and a first gas flow meter. The gas sensing module consists of CO, H2, and CH4 gas sensors. The multi-composite flue gas sensing module includes multi-composite flue gas sensing module A and multi-composite flue gas sensing module B, which mutually verify each other to avoid deviations caused by drift of a single sensor. Flue gas passing through multi-composite flue gas sensing module A is recorded as A-, and flue gas passing through multi-composite flue gas sensing module B-. A second gas flow meter and a third pressure sensor are arranged at the inlet of each flue gas pretreatment module to monitor the airflow pressure and velocity flowing into and out of the multi-composite flue gas sensing module.

[0011] The sensor data storage module is used for local caching of data acquired by the flue gas acquisition and distribution module, the multi-composite flue gas sensing module, and the flue gas pretreatment module. Specifically, it includes address change signals of the sample gas intake pipe number, the flue gas pretreatment module number, and the multi-composite flue gas sensing module number. It also includes data information of the sensor modules of the multi-composite flue gas sensing module and the flue gas pretreatment module, i.e., locally cached sensor data, with a storage capacity of not less than 24 hours.

[0012] The multi-information fusion early warning module is a microprocessor that provides a hardware platform for the operation of the thermal runaway detection and alarm algorithm, and outputs graded early warning and alarm signals for battery thermal runaway disasters;

[0013] The thermal runaway detection and alarm algorithm includes a multi-dimensional information fusion algorithm, an early warning judgment and correction algorithm, and a positioning algorithm;

[0014] The multi-dimensional information fusion algorithm is a sequential Kalman filter fusion algorithm, which performs pixel-level fusion processing on smoke, gas, and temperature, and couples and superimposes the multi-dimensional information of smoke, gas, and temperature through a multi-level coupled Kalman filter.

[0015] The aforementioned early warning judgment correction algorithm combines gas flow rate, pressure, and wind speed with a rational gas state equation to correct the flue gas parameter values ​​collected by the sensor. The early warning judgment correction algorithm uses a BP neural network algorithm for self-correction, realizes hierarchical alarm logic, and improves the universality of the algorithm in different scenarios.

[0016] The hierarchical alarm logic includes three levels of alarms: Level 1 alarm response is to activate the fire extinguishing system; Level 2 alarm response is to send a trigger signal to the battery system to cut off the power supply; Level 3 alarm response is to send a fault warning signal to the battery system, without any specific response action.

[0017] The positioning algorithm uses the address change signals of the sample gas intake tube number, filter module number, and corresponding sensor number to locate the detected thermal runaway battery pack using an encoding method. It also introduces a time-series neural network to correct the early warning alarm parameters of the encoding bit. Specifically, it corrects the parameters and alarm threshold in the early warning judgment correction algorithm, thereby achieving the functions of encoding correction and algorithm efficiency improvement.

[0018] On the other hand, a thermal runaway detection and alarm method for a lithium-ion battery energy storage system, based on the aforementioned thermal runaway detection and alarm device for a lithium-ion battery energy storage system, includes the following steps:

[0019] Step 1: Assemble the air intake tube with a snap-fit ​​into the pre-drilled hole in the battery pack;

[0020] Step 2: The battery pack module corresponding to the intake pipe will be numbered through the flue gas collection and distribution module to achieve a one-to-one correspondence between the intake pipe and the battery pack in the lithium-ion battery energy storage system.

[0021] Step 3: Start the vacuum pump of the flue gas collection and distribution module to sample and monitor the gas in the battery pack of the lithium-ion battery energy storage system. The sampling period shall not exceed 60 seconds, and the suction pipe shall be recorded and numbered through the electromagnetic distribution valve of the flue gas collection and distribution module.

[0022] Step 4: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module a and the code is updated to a-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module a.

[0023] Step 5: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module b and the code is updated to b-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module b.

[0024] Step 6: Transmit the gas sample a-Sample1 processed by flue gas pretreatment module a to the multi-composite flue gas sensing module A, and update the code to Aa-Sample1. Process and obtain the four sets of sensor data results Aa-Sample1, Ab-Sample1, Ba-Sample1, and Bb-Sample1 of the sample1 gas sample in sequence, and record them as the sensor values ​​of the sample1 gas sample.

[0025] The four sets of sensor data are: the flue gas pretreatment module a and flue gas pretreatment module b corresponding to Sample1 gas sample, and the four sets of data formed by mutual verification between the multi-composite flue gas sensing module a and multi-composite flue gas sensing module b.

[0026] Step 7: Transmit the sensor values ​​of Sample1 gas sample to the sensor data storage module for storage, and then transmit them to the multi-information fusion early warning module for early warning logic judgment. Transmit the sensor values ​​of Sample1 gas sample with ≥2 sampling times to the multi-information fusion early warning module for time-division and branch-division verification of sensor values. The value obtained after exponential averaging of the verified values ​​is recorded as the sensor verification average value. The thermal runaway detection alarm algorithm judges the early warning threshold based on the given sensor verification average value. Once the judgment output of the thermal runaway detection alarm algorithm reaches the alarm threshold, an early warning signal is issued.

[0027] Step 8: Repeat steps 1-7 to sample different battery modules and realize thermal runaway detection and alarm for lithium-ion battery energy storage system.

[0028] The beneficial effects of adopting the above technical solution are as follows:

[0029] This invention provides a thermal runaway detection and alarm device and method for lithium-ion battery energy storage systems. By using a passive suction pipe arranged within the lithium-ion battery energy storage system, this invention avoids false alarms and missed alarms caused by interference from corresponding detection sensors. Simultaneously, it incorporates a flue gas pretreatment module to effectively filter moisture and dust within the highly integrated lithium-ion battery pack, improving the reliability of the detection and alarm. Through the mutually verifying flue gas pretreatment module and multiple composite flue gas sensing modules, environmental interference is avoided, enhancing alarm reliability. The thermal runaway detection and alarm algorithm, running through a multi-information fusion early warning module, achieves intelligent monitoring and reliable alarm for the lithium-ion battery energy storage system. No similar devices or equipment exist in this field. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the overall structure of the thermal runaway detection and alarm device for a lithium-ion battery energy storage system in an embodiment of the present invention;

[0031] Figure 2 This is an overall flowchart of the thermal runaway detection and alarm method for lithium-ion battery energy storage systems in this embodiment of the invention. Detailed Implementation

[0032] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0033] On the one hand, a thermal runaway detection and alarm device for lithium-ion battery energy storage systems, such as Figure 1 As shown, it includes an intake pipe, a flue gas collection and distribution module, a flue gas pretreatment module, a multi-composite flue gas sensing module, a sensor data storage module, and a multi-information fusion early warning module.

[0034] The suction pipe is a multi-channel suction pipe, which in this embodiment is referred to as suction pipe 1, suction pipe 2, suction pipe 3, ..., suction pipe N respectively; the multiple suction pipes are arranged in parallel, and the output end of each suction pipe is connected to a battery pack under test, and the output end is connected to the smoke collection and distribution module, which is used to collect the smoke inside the battery under test and transmit it to the smoke collection and distribution module. The smoke absorbed by any one of the suction pipes is set as the sample gas.

[0035] The output of the flue gas collection and distribution module is connected to the flue gas pretreatment module, and the output of the flue gas pretreatment module is connected to the multi-composite flue gas sensing module to distribute and process the collected gas. The flue gas collection and distribution module includes a vacuum pump and an electromagnetic distribution valve. The vacuum pump provides a suction force of not less than -80 kPa to the suction pipe, and the gas flow rate needs to be greater than or equal to 0.01. The electromagnetic distribution valve selects and outputs address code signals for the multiple suction pipes respectively, which are used to locate the battery pack under test. In this embodiment, the corresponding numbers are Sample1, Sample2, Sample3, ..., SampleN.

[0036] The flue gas pretreatment module specifically includes flue gas pretreatment module a and flue gas pretreatment module b. The flue gas acquisition and distribution module transmits the acquired flue gas to flue gas pretreatment module a and flue gas pretreatment module b respectively according to the control logic. This is used to dry, filter, stabilize, and equalize the received flue gas, avoiding interference from interfering gases, dust, and water vapor on the detection accuracy of the multi-composite flue gas sensing module. Flue gas pretreatment module a and flue gas pretreatment module b are mutually verified to avoid deviations caused by filtration by a single flue gas pretreatment module. The flue gas passing through flue gas pretreatment module a is recorded as a-, and the flue gas passing through flue gas pretreatment module b is recorded as b-. Each flue gas pretreatment module is equipped with an anemometer and a first pressure sensor at its input end to monitor the airflow pressure and velocity flowing into and out of the flue gas pretreatment module. The inlet and outlet pressure or flow rate of the flue gas pretreatment module needs to be greater than or equal to 1.5.

[0037] The control logic is to divide the collected flue gas into frequencies according to a set ratio based on the sampling time.

[0038] The control logic in this embodiment includes two steps: inspection control and allocation control. First, the gas collected from the battery modules Sample1, Sample2, Sample3, ..., SampleN by suction pipes 1, 2, 3, ..., N is sampled sequentially for inspection. The sampling time is 10 t s. Then, the sampled gas is divided into frequencies according to a 7:3 ratio, i.e., the gas is allocated to flue gas pretreatment module a for 7 t s and to flue gas pretreatment module b for 3 t s.

[0039] The multi-composite flue gas sensing module is used to analyze the component type, concentration, and symptom values ​​of the flue gas transmitted by the flue gas pretreatment module. It includes a smoke probe, a gas sensing module, a second pressure sensor, two sets of temperature sensors, and a first gas flow meter. The gas sensing module consists of CO, H2, and CH4 gas sensors. The multi-composite flue gas sensing module includes multi-composite flue gas sensing module A and multi-composite flue gas sensing module B, which mutually verify each other to avoid deviations caused by drift of a single sensor. Flue gas passing through multi-composite flue gas sensing module A is recorded as A-, and flue gas passing through multi-composite flue gas sensing module B-. A second gas flow meter and a third pressure sensor are arranged at the inlet of each flue gas pretreatment module to monitor the airflow pressure and velocity flowing into and out of the multi-composite flue gas sensing module. The inlet and outlet pressure or flow rate of the multi-composite flue gas sensing module must be greater than or equal to 1.5.

[0040] The sensor data storage module is used for local caching of data acquired by the flue gas acquisition and distribution module, the multi-composite flue gas sensing module, and the flue gas pretreatment module. Specifically, it includes address change signals of the sample gas intake pipe number, the flue gas pretreatment module number, and the multi-composite flue gas sensing module number. It also includes data information of the sensor modules of the multi-composite flue gas sensing module and the flue gas pretreatment module, i.e., locally cached sensor data, with a storage capacity of not less than 24 hours.

[0041] The multi-information fusion early warning module is a microprocessor that provides a hardware platform for the operation of the thermal runaway detection and alarm algorithm, and outputs graded early warning and alarm signals for battery thermal runaway disasters;

[0042] The thermal runaway detection and alarm algorithm includes a multi-dimensional information fusion algorithm, an early warning judgment and correction algorithm, and a positioning algorithm;

[0043] The multi-dimensional information fusion algorithm is a sequential Kalman filter fusion algorithm, which performs pixel-level fusion processing on smoke, gas, and temperature, and couples and superimposes the multi-dimensional information of smoke, gas, and temperature through a multi-level coupled Kalman filter.

[0044] The aforementioned early warning judgment correction algorithm combines gas flow rate, pressure, and wind speed with a rational gas state equation to correct the flue gas parameter values ​​collected by the sensor, avoiding false alarms and missed alarms caused by environmental climate changes; the early warning judgment correction algorithm uses a BP neural network algorithm for self-correction, realizes hierarchical alarm logic, and improves the universality of the algorithm in different scenarios.

[0045] The hierarchical alarm logic includes three levels of alarms: Level 1 alarm response is to activate the fire extinguishing system; Level 2 alarm response is to send a trigger signal to the battery system to cut off the power supply; Level 3 alarm response is to send a fault warning signal to the battery system, without any specific response action.

[0046] The positioning algorithm uses the address change signals of the sample gas intake tube number, filter module number, and corresponding sensor number to locate the detected thermal runaway battery pack using an encoding method. It also introduces a time-series neural network to correct the early warning alarm parameters of the encoding bit. Specifically, it corrects the parameters and alarm threshold in the early warning judgment correction algorithm, thereby achieving the functions of encoding correction and algorithm efficiency improvement.

[0047] On the other hand, a thermal runaway detection and alarm method for a lithium-ion battery energy storage system is implemented based on the aforementioned thermal runaway detection and alarm device for a lithium-ion battery energy storage system, such as... Figure 2 As shown, it includes the following steps:

[0048] Step 1: Assemble the air intake tube with a snap-fit ​​into the pre-drilled hole in the battery pack;

[0049] Step 2: The battery pack module corresponding to the intake pipe will be numbered through the flue gas collection and distribution module to achieve a one-to-one correspondence between the intake pipe and the battery pack in the lithium-ion battery energy storage system.

[0050] Step 3: Start the vacuum pump of the flue gas collection and distribution module to sample and monitor the gas in the battery pack of the lithium-ion battery energy storage system. The sampling period shall not exceed 60 seconds, and the suction pipe shall be recorded and numbered through the electromagnetic distribution valve of the flue gas collection and distribution module.

[0051] Step 4: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module a and the code is updated to a-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module a.

[0052] Step 5: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module b and the code is updated to b-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module b.

[0053] Step 6: Transmit the gas sample a-Sample1 processed by flue gas pretreatment module a to the multi-composite flue gas sensing module A, and update the code to Aa-Sample1. Process and obtain the four sets of sensor data results Aa-Sample1, Ab-Sample1, Ba-Sample1, and Bb-Sample1 of the sample1 gas sample in sequence, and record them as the sensor values ​​of the sample1 gas sample.

[0054] The four sets of sensor data are: the flue gas pretreatment module a and flue gas pretreatment module b corresponding to Sample1 gas sample, and the four sets of data formed by mutual verification between the multi-composite flue gas sensing module a and multi-composite flue gas sensing module b.

[0055] Step 7: Transmit the sensor values ​​of Sample1 gas sample to the sensor data storage module for storage, and then transmit them to the multi-information fusion early warning module for early warning logic judgment. Transmit the sensor values ​​of Sample1 gas sample with ≥2 sampling times to the multi-information fusion early warning module for time-division and branch-division verification of the sensor values. The verified values ​​are then exponentially averaged and recorded as the sensor verification average value. The thermal runaway detection alarm algorithm uses the given sensor verification average value to determine the early warning threshold. Once the judgment output of the thermal runaway detection alarm algorithm reaches the alarm threshold, an early warning signal is issued. The standard deviation of the sensor values ​​must be less than or equal to 0.05.

[0056] Step 8: Repeat steps 1-7 to sample different battery modules and realize thermal runaway detection and alarm for lithium-ion battery energy storage system.

[0057] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A thermal runaway detection and alarm device for a lithium-ion battery energy storage system, characterized in that, It includes an intake pipe, a flue gas collection and distribution module, a flue gas pretreatment module, a multi-composite flue gas sensing module, a sensor data storage module, and a multi-information fusion early warning module; The suction pipe is a multi-channel suction pipe, which is set in parallel. The input end of each suction pipe is connected to a battery pack under test, and the output end is connected to the smoke collection and distribution module. It is used to collect the smoke inside the battery under test and transmit it to the smoke collection and distribution module. The smoke absorbed by any one suction pipe is recorded as the sample gas. The output of the flue gas collection and distribution module is connected to the flue gas pretreatment module, and the output of the flue gas pretreatment module is connected to the multi-composite flue gas sensing module. The flue gas pretreatment module specifically includes flue gas pretreatment module a and flue gas pretreatment module b. The flue gas acquisition and distribution module transmits the acquired flue gas to flue gas pretreatment module a and flue gas pretreatment module b respectively according to the control logic. The flue gas pretreatment module a and flue gas pretreatment module b are used to dry, filter, stabilize and equalize the received flue gas. The flue gas pretreatment module a and flue gas pretreatment module b verify each other to avoid deviations caused by filtering by a single flue gas pretreatment module. The flue gas that passes through flue gas pretreatment module a is recorded as a-, and the flue gas that passes through flue gas pretreatment module b is recorded as b-. Each flue gas pretreatment module is equipped with an anemometer and a first pressure sensor at its input end to monitor the airflow pressure and velocity flowing into and out of the flue gas pretreatment module. The multi-composite flue gas sensing module is used to analyze the component type, concentration, and symptom values ​​of the flue gas transmitted by the flue gas pretreatment module. It includes a smoke probe, a gas sensing module, a second pressure sensor, two sets of temperature sensors, and a first gas flow meter. The gas sensing module is a CO, H2, and CH4 gas sensor. The multi-composite flue gas sensing module includes multi-composite flue gas sensing module A and multi-composite flue gas sensing module B. The two sets are mutually verified to avoid deviations caused by drift of a single sensor. The flue gas passing through multi-composite flue gas sensing module A is recorded as A-, and the flue gas passing through multi-composite flue gas sensing module B is recorded as B-. A second gas flow meter and a third pressure sensor are arranged at the inlet of each set of flue gas pretreatment modules to monitor the airflow pressure and velocity flowing into and out of the multi-composite flue gas sensing module. The multi-information fusion early warning module is a microprocessor that provides a hardware platform for the operation of the thermal runaway detection and alarm algorithm, and outputs graded early warning and alarm signals for battery thermal runaway disasters.

2. The thermal runaway detection and alarm device for a lithium-ion battery energy storage system according to claim 1, characterized in that, The control logic is to sample the collected flue gas in a time-division manner according to a set ratio.

3. The thermal runaway detection and alarm device for a lithium-ion battery energy storage system according to claim 1, characterized in that, The sensor data storage module is used for local caching of data acquired by the flue gas acquisition and distribution module, the multi-composite flue gas sensing module, and the flue gas pretreatment module. Specifically, it includes address change signals of the sample gas intake pipe number, the flue gas pretreatment module number, and the multi-composite flue gas sensing module number. It also includes data information from the sensor modules of the multi-composite flue gas sensing module and the flue gas pretreatment module, i.e., locally cached sensor data, with a storage capacity of not less than 24 hours.

4. The thermal runaway detection and alarm device for a lithium-ion battery energy storage system according to claim 1, characterized in that, The flue gas collection and distribution module includes a vacuum pump and an electromagnetic distribution valve. The vacuum pump provides a suction force with an absolute value of not less than -80 kPa to the suction pipe. The electromagnetic distribution valve selects multiple suction pipes and outputs address code signals for positioning the battery pack under test.

5. A thermal runaway detection and alarm device for a lithium-ion battery energy storage system according to claim 1, characterized in that, The thermal runaway detection and alarm algorithm includes a multi-dimensional information fusion algorithm, an early warning judgment and correction algorithm, and a positioning algorithm; The multi-dimensional information fusion algorithm is a sequential Kalman filter fusion algorithm, which performs pixel-level fusion processing on smoke, gas and temperature. The multi-dimensional information of smoke, gas and temperature collected by multiple sensors is coupled and superimposed sequentially through a multi-level coupled Kalman filter. Distributed fusion logic is adopted, that is, the data of a single sensor is fused and then the fused data of different sensors are fused. The warning judgment correction algorithm combines gas flow rate, pressure, and wind speed to correct the flue gas parameter values ​​collected by the sensor using the ideal gas state equation. The warning judgment correction algorithm uses a BP neural network algorithm for self-correction, realizes hierarchical alarm logic, and improves the universality of the algorithm in different scenarios. The positioning algorithm uses the address change signals of the sample gas intake tube number, filter module number, and corresponding sensor number to locate the detected thermal runaway battery pack using an encoding method. It also introduces a temporal neural network to correct the early warning alarm parameters of the encoding bit. Specifically, it corrects the parameters and alarm thresholds in the early warning judgment correction algorithm, thereby achieving the functions of encoding correction and algorithm efficiency improvement.

6. A thermal runaway detection and alarm device for a lithium-ion battery energy storage system according to claim 5, characterized in that, The hierarchical alarm logic includes three levels of alarms: Level 1 alarm response activates the fire extinguishing system of the battery energy storage system; Level 2 alarm response sends a trigger signal to the battery system to cut off the power supply; Level 3 alarm response sends a fault warning signal to the battery system without any specific response action.

7. A thermal runaway detection and alarm method for a lithium-ion battery energy storage system, implemented based on the thermal runaway detection and alarm device for a lithium-ion battery energy storage system as described in claim 1, characterized in that, Includes the following steps: Step 1: Assemble the air intake tube with a snap-fit ​​into the pre-drilled hole in the battery pack; Step 2: The battery pack module corresponding to the intake pipe will be numbered through the flue gas collection and distribution module to achieve a one-to-one correspondence between the intake pipe and the battery pack in the lithium-ion battery energy storage system. Step 3: Start the vacuum pump of the flue gas collection and distribution module to sample and monitor the gas in the battery pack of the lithium-ion battery energy storage system. The sampling period shall not exceed 60 seconds, and the suction pipe shall be recorded and numbered through the electromagnetic distribution valve of the flue gas collection and distribution module. Step 4: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module a and the code is updated to a-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module a. Step 5: The gas sample of Sample1 collected by the flue gas collection and distribution module is transmitted to the flue gas pretreatment module b and the code is updated to b-Sample1. The gas sample of Sample1 is dried and filtered by the flue gas pretreatment module b. Step 6: Transmit the gas sample a-Sample1 processed by flue gas pretreatment module a to the multi-composite flue gas sensing module A, and update the code to Aa-Sample1. Process and obtain the four sets of sensor data results Aa-Sample1, Ab-Sample1, Ba-Sample1, and Bb-Sample1 of the sample1 gas sample in sequence, and record them as the sensor values ​​of the sample1 gas sample. The four sets of sensor data are: the flue gas pretreatment module a and flue gas pretreatment module b corresponding to Sample1 gas sample, and the four sets of data formed by mutual verification between the multi-composite flue gas sensing module A and multi-composite flue gas sensing module B. Step 7: Transmit the sensor values ​​of Sample1 gas sample to the sensor data storage module for storage, and then transmit them to the multi-information fusion early warning module for early warning logic judgment. Transmit the sensor values ​​of Sample1 gas sample with ≥2 sampling times to the multi-information fusion early warning module for time-division and branch-division verification of the sensor values. The value obtained after exponential averaging of the verified values ​​is recorded as the sensor verification average value. The thermal runaway detection alarm algorithm judges the early warning threshold based on the given sensor verification average value. Once the judgment output of the thermal runaway detection alarm algorithm reaches the alarm threshold, an early warning signal is issued. Step 8: Repeat steps 1-7 to sample different battery modules to achieve thermal runaway detection and alarm for lithium-ion battery energy storage systems.

Images