Graded early warning method for thermal runaway of battery

Abstract

The invention discloses a battery thermal runaway grading early warning method. The method comprises the following steps: collecting multi-dimensional data of battery operation through a multi-sensor system; processing the multi-dimensional data to perform data optimization and adaptive threshold generation; identifying thermal runaway stage features based on the optimized multi-dimensional data and an adaptive threshold; and triggering grading alarm according to a thermal runaway feature identification result. By constructing a full-chain early warning mechanism of data acquisition, dynamic optimization, feature decoupling and hierarchical response, three breakthroughs are realized: 1, a self-adaptive threshold model based on change rate compensation and coupling solves the problem of reliability attenuation caused by sensor attenuation and working condition fluctuation; secondly, fine tracking of the thermal runaway process is achieved through independent recognition logic of characteristics of the germination stage, the pre-occurrence stage and the occurrence time stage; and thirdly, the robustness and the response efficiency of the safety system are remarkably improved through accurate matching of graded alarm and prevention and control measures, and full-life-cycle safety protection is provided for the battery system.

Description

Technical Field

[0001] The present invention relates to the field of battery safety and protection, and in particular to a battery thermal runaway graded early warning method. Background Art

[0002] Traditional thermal runaway monitoring relies on fixed thresholds and single-parameter judgments, making it difficult to cope with the combined interference of battery aging, ambient temperature fluctuations, and sensor degradation. This method overcomes the limitations of static models by integrating multi-source heterogeneous data with adaptive thresholds. Summary of the Invention

[0003] Based on the problems raised in the background technology, the present invention proposes a battery thermal runaway graded warning method, comprising the following steps: S1, collect multi-dimensional data of battery operation through a multi-sensor system; S2, processing multi-dimensional data to perform data optimization and adaptive threshold generation; S3, identifying thermal runaway stage characteristics based on optimized multi-dimensional data and adaptive thresholds; S4. Trigger a graded alarm based on the thermal runaway feature identification results.

[0004] Furthermore, collecting multi-dimensional data of battery operation through the multi-sensor system specifically includes: collecting electrolyte leakage data through the VOC sensor in the multi-sensor system; collecting carbon monoxide gas data through the CO sensor in the multi-sensor system; collecting smoke concentration data through the smoke sensor in the multi-sensor system; collecting temperature data through the temperature sensor in the multi-sensor system; and outputting multi-dimensional data including electrolyte leakage data, carbon monoxide gas data, smoke concentration data and temperature data.

[0005] Furthermore, processing multi-dimensional data to perform data optimization and adaptive threshold generation specifically includes: executing a compensation strategy based on the rate of change on the multi-dimensional data to correct the sensor attenuation error, thereby obtaining optimized multi-dimensional data; and generating an adaptive threshold based on the optimized multi-dimensional data.

[0006] Furthermore, a rate-of-change compensation strategy is performed on the multi-dimensional data to correct the sensor attenuation error, thereby obtaining optimized multi-dimensional data, which specifically includes: performing rate-of-change compensation on the electrolyte leakage data to obtain compensated electrolyte leakage data; performing rate-of-change compensation on the carbon monoxide gas data to obtain compensated carbon monoxide gas data; performing rate-of-change compensation on the smoke concentration data to obtain compensated smoke concentration data; performing rate-of-change compensation on the temperature data to obtain compensated temperature data. The optimized multi-dimensional data specifically includes: compensated electrolyte leakage data, compensated carbon monoxide gas data, compensated smoke concentration data, and compensated temperature data.

[0007] Furthermore, dynamically generating an adaptive threshold based on the optimized multi-dimensional data specifically includes: establishing a multi-parameter dynamic model through an automatic coupling mechanism; generating a dynamic threshold based on a state evolution trajectory; verifying the threshold stability through parameter mutual feedback; and outputting an adaptive threshold optimized by dynamic coupling.

[0008] Furthermore, identifying the characteristics of the thermal runaway stage based on the optimized multi-dimensional data and the adaptive threshold specifically includes: judging whether the characteristics of the thermal runaway incipient stage are met based on the adaptive electrolyte leakage threshold and the compensated electrolyte leakage data; judging whether the characteristics before the occurrence of thermal runaway are met based on the carbon monoxide gas adaptive threshold, the temperature adaptive threshold, the compensated carbon monoxide gas data and the compensated temperature data; judging whether the characteristics when thermal runaway occurs are met based on the compensated smoke concentration data, the compensated temperature data, the smoke concentration adaptive threshold and the temperature adaptive threshold; and outputting the thermal runaway characteristic identification result.

[0009] Furthermore, judging whether the thermal runaway incipient stage characteristics are met based on the adaptive electrolyte leakage threshold and the compensated electrolyte leakage data specifically includes: calculating the compensated electrolyte leakage change rate based on the compensated electrolyte leakage data, comparing the compensated electrolyte leakage change rate with the adaptive electrolyte leakage threshold, and when the electrolyte leakage change rate is greater than the adaptive leakage rate threshold, judging that the thermal runaway incipient stage characteristics are met; otherwise, the thermal runaway incipient stage characteristics are not met; wherein, the adaptive electrolyte leakage threshold can be adjusted lower as the electrolyte volatilization rate decays.

[0010] Furthermore, identifying the characteristics of the thermal runaway stage includes: comparing the change rate of the electrolyte leakage data with the first threshold of the adaptive threshold, and determining that the characteristics of the thermal runaway embryonic stage are met when the change rate of the electrolyte leakage data is greater than the adaptive leakage rate threshold; combining the change rates of the carbon monoxide gas data and the temperature data, and determining that the characteristics before the occurrence of thermal runaway are met when both the carbon monoxide gas data is greater than the gas concentration threshold and the temperature data change rate is greater than the temperature rise rate threshold; coupling the smoke concentration data and the temperature data, and determining that the characteristics of the thermal runaway occurrence are met when both the smoke concentration data is greater than the smoke concentration threshold and the temperature data is greater than the temperature critical threshold.

[0011] Furthermore, judging whether the characteristics before thermal runaway occurs are met based on the carbon monoxide gas adaptive threshold, the temperature adaptive threshold, the compensated carbon monoxide gas data and the compensated temperature data specifically includes: calculating the compensated temperature change rate based on the compensated temperature data; when the compensated carbon monoxide gas data is higher than the carbon monoxide gas adaptive threshold and the compensated temperature change rate is greater than the temperature adaptive threshold, it is determined that the characteristics before thermal runaway occurs are met; otherwise, the characteristics before thermal runaway occurs are not met.

[0012] Furthermore, judging whether the characteristics of thermal runaway occurrence are met based on the compensated smoke concentration data, the compensated temperature data, the smoke concentration adaptive threshold and the temperature adaptive threshold specifically includes: when the compensated smoke concentration data is higher than the smoke concentration adaptive threshold, and the compensated temperature data is higher than the temperature adaptive threshold, the characteristics of thermal runaway occurrence are met; otherwise, the characteristics of thermal runaway occurrence are not met, wherein, when the ambient temperature is ≥60°C, the temperature adaptive threshold automatically rises.

[0013] The present invention achieves three major breakthroughs by constructing a full-chain early warning mechanism of "data acquisition-dynamic optimization-feature decoupling-graded response": First, an adaptive threshold model based on change rate compensation and coupling solves the reliability degradation problem caused by sensor attenuation and operating condition fluctuations; second, through independent recognition logic of the three-stage characteristics of the germination period, before occurrence, and during occurrence, refined tracking of the thermal runaway process is achieved; third, the precise matching of graded alarms and prevention and control measures significantly improves the robustness and response efficiency of the safety system, providing full life cycle safety protection for the battery system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of a battery thermal runaway graded warning method proposed by the present invention; Figure 2 A partial flow chart of a battery thermal runaway graded warning method proposed by the present invention; Figure 3 A partial flow chart of a battery thermal runaway graded warning method proposed by the present invention; Figure 4 This is a partial flow chart of a battery thermal runaway graded warning method proposed by the present invention. DETAILED DESCRIPTION

[0015] refer to Figure 1 The present invention proposes a battery thermal runaway hierarchical warning method, which specifically includes: S1. Collect multi-dimensional data of battery operation through a multi-sensor system.

[0016] This step uses a multi-sensor system to synchronously collect the physical and chemical parameters of battery operation. VOC sensors capture real-time changes in volatile gas concentrations from electrolyte leakage, CO sensors continuously monitor carbon monoxide levels produced by battery decomposition, smoke sensors sense particulate matter concentrations from electrolyte combustion, and temperature sensors track thermal distribution on and within the battery. Through multi-channel data parallel acquisition and timestamp alignment, a multi-dimensional data base reflecting the battery's safety status is constructed, significantly enhancing the environmental coverage and timeliness of the initial data.

[0017] S1 specifically includes: The VOC sensor in the multi-sensor system collects electrolyte leakage data; the CO sensor in the multi-sensor system collects carbon monoxide gas data; the smoke sensor in the multi-sensor system collects smoke concentration data; the temperature sensor in the multi-sensor system collects temperature data; and the system outputs multi-dimensional data including electrolyte leakage data, carbon monoxide gas data, smoke concentration data and temperature data.

[0018] S2. Processing multi-dimensional data to perform data optimization and generate adaptive thresholds. The adaptive thresholds include: electrolyte leakage adaptive threshold, carbon monoxide gas adaptive threshold, smoke concentration adaptive threshold, and temperature adaptive threshold.

[0019] This step first applies rate-of-change compensation based on historical decay curves to electrolyte leakage data, carbon monoxide gas data, smoke concentration data, and temperature data. This dynamically corrects baseline errors caused by sensor sensitivity drift, generating high-fidelity compensated electrolyte leakage data, compensated carbon monoxide gas data, compensated smoke concentration data, and compensated temperature data. A multi-parameter dynamic correlation model is then established based on an automatic coupling mechanism. This model uses real-time synchronization of ambient temperature and battery aging coefficients to perform parameter feedback on the compensated data. The state evolution trajectory of the compensated data is calculated in the continuous time domain, and adaptive thresholds for electrolyte leakage, carbon monoxide gas, smoke concentration, and temperature are dynamically generated based on the trajectory curvature. This mechanism achieves self-organized optimization of threshold parameters through closed-loop coupling of environmental variables and device status, ultimately outputting a set of adaptive thresholds that precisely match the actual battery degradation state and external operating conditions.

[0020] S2 specifically includes: S21 . Execute a compensation strategy based on the rate of change on the multi-dimensional data to correct the sensor attenuation error, thereby obtaining optimized multi-dimensional data.

[0021] Specifically, it includes: performing change rate compensation on electrolyte leakage data to obtain compensated electrolyte leakage data; performing change rate compensation on carbon monoxide gas data to obtain compensated carbon monoxide gas data; performing change rate compensation on smoke concentration data to obtain compensated smoke concentration data; performing change rate compensation on temperature data to obtain compensated temperature data. The optimized multi-dimensional data specifically includes: compensated electrolyte leakage data, compensated carbon monoxide gas data, compensated smoke concentration data and compensated temperature data.

[0022] S22: Generate adaptive thresholds based on the optimized multi-dimensional data, wherein the adaptive thresholds include: an adaptive electrolyte leakage threshold, an adaptive carbon monoxide gas threshold, an adaptive smoke concentration threshold, and an adaptive temperature threshold.

[0023] S22 specifically includes: S221. Establish a multi-parameter dynamic model through an automatic coupling mechanism.

[0024] Specifically, it includes: real-time parameter mutual feedback of the compensated electrolyte leakage data, and construction of the dynamic evolution equation of the compensated electrolyte leakage data. ,in, For compensated electrolyte leakage data / compensated carbon monoxide gas data / compensated smoke concentration data / compensated temperature data, is the ambient temperature, is the aging coefficient; the data association weight matrix is automatically adjusted according to the parameter mutual feedback strength.

[0025] S222: Generate a dynamic threshold based on the state evolution trajectory. Specifically including: Calculate the trajectory curvature of compensated electrolyte leakage data / compensated carbon monoxide gas data / compensated smoke concentration data / compensated temperature data in a multi-parameter dynamic model , ; When the curvature exceeds the preset critical value, the threshold update is triggered: ,in: is the current parameter reference value, is the parameter fluctuation intensity, The value is dynamically adjusted by the curvature change rate. is the ambient temperature compensation factor, which is a calibration value and has no unit. is the measured value of the ambient temperature, which comes from the temperature sensor. is the rate of change of ambient temperature; is the aging coefficient compensation factor, which is a calibrated value and has no unit; The battery aging level comes from the BMS output.

[0026] S223. Verify the threshold stability through parameter mutual feedback.

[0027] For example, cross-validation of the generated thresholds is performed to ensure that the compensated electrolyte leakage threshold and the compensated temperature threshold satisfy the thermodynamic constraint relationship.

[0028] S224. Output the adaptive threshold value optimized by dynamic coupling.

[0029] S3. Identify the characteristics of the thermal runaway stage based on the optimized multi-dimensional data and adaptive thresholds.

[0030] This step identifies thermal runaway characteristics through a parallel comparison mechanism. For incipient characteristics, the rate of change of the compensated electrolyte leakage data is compared in real time with the adaptive electrolyte leakage threshold. A trigger is triggered when the rate of change exceeds the threshold, and the threshold is automatically adjusted downward based on the electrolyte's volatility, thereby improving the ability to detect early trace leaks. For pre-onset characteristics, the compensated carbon monoxide gas data is simultaneously verified to see if it exceeds the adaptive carbon monoxide gas threshold. The compensated temperature change rate is then compared with the adaptive temperature threshold. The characteristic is confirmed only when both conditions are met, thus avoiding interference from single parameter mutations. For onset characteristics, the compensated smoke concentration data must exceed the adaptive smoke concentration threshold and the compensated temperature data must simultaneously exceed the adaptive temperature threshold. The temperature threshold is dynamically increased with the measured temperature, thereby suppressing false triggering under normal high-temperature conditions. Through the dynamic coordination of multi-dimensional logic thresholds and thresholds, high-precision separation of the three stages of thermal runaway is achieved.

[0031] S3 specifically includes: S31. Determine whether the characteristics of the incipient stage of thermal runaway are met based on the adaptive electrolyte leakage threshold and the compensated electrolyte leakage data.

[0032] Specifically, the compensated electrolyte leakage change rate is calculated based on the compensated electrolyte leakage data and compared with the adaptive electrolyte leakage threshold. When the electrolyte leakage change rate exceeds the adaptive leakage rate threshold, it is determined that the thermal runaway incipient stage characteristics are met; otherwise, the thermal runaway incipient stage characteristics are not met. The adaptive electrolyte leakage threshold can be adjusted lower as the electrolyte volatilization rate decays to improve the sensitivity of trace leak detection.

[0033] S32: Determine whether the characteristics before thermal runaway occur based on the CO gas adaptive threshold, the temperature adaptive threshold, the compensated CO gas data, and the compensated temperature data. Specifically, it includes: The compensated temperature change rate is calculated based on the compensated temperature data. When the compensated carbon monoxide gas data is higher than the carbon monoxide gas adaptive threshold and the compensated temperature change rate is greater than the temperature adaptive threshold, it is determined that the characteristics before thermal runaway occur are met; otherwise, the characteristics before thermal runaway occur are not met.

[0034] S33: Determine whether the characteristics of thermal runaway are met based on the compensated smoke density data, the compensated temperature data, the smoke density adaptive threshold, and the temperature adaptive threshold. Specifically: When the compensated smoke density data exceeds the smoke density adaptive threshold, and the compensated temperature data exceeds the temperature adaptive threshold, the thermal runaway characteristic is met; otherwise, the thermal runaway characteristic is not met. As the temperature rises, the adaptive threshold will automatically increase to prevent false triggering.

[0035] Specifically, when the ambient temperature reaches or exceeds 60°C, the temperature adaptive threshold will automatically increase according to the preset algorithm to avoid false triggering under high temperature conditions. The preset algorithm is as follows: , where R is the updated temperature adaptive threshold; is the current temperature adaptive threshold; is the ambient temperature compensation factor, which is a calibration value and has no unit. is the measured value of the ambient temperature, which comes from the temperature sensor. is the rate of change of ambient temperature.

[0036] S34. Outputting the thermal runaway feature identification result. The identification result includes: meeting the thermal runaway incipient stage characteristics, meeting the thermal runaway pre-occurrence characteristics, meeting the thermal runaway occurrence characteristics, or no thermal runaway.

[0037] S4. Trigger a graded alarm based on the thermal runaway feature identification results.

[0038] This step activates a hierarchical response based on the identification results: when incipient characteristics are met, a pre-alarm signal is generated and a self-check process is initiated; when pre-occurrence characteristics are met, an emergency power reduction command and ventilation system linkage are triggered; and when onset characteristics are met, the fire extinguishing device is activated and the power circuit is cut off. By mapping alarm signals to emergency measures at different levels, a progressive defense system is established, significantly reducing response delays in extreme conditions and optimizing the efficiency of security resource scheduling.

[0039] S4 specifically includes: mapping the comprehensive risk value into a three-level alarm signal, generating an incipient alarm signal when the identification result meets the characteristics of the thermal runaway embryonic stage; generating a pre-occurrence alarm signal when the identification result meets the characteristics before the thermal runaway occurs; generating an on-occurrence alarm signal when the identification result meets the characteristics when the thermal runaway occurs; and outputting the final alarm signal.

[0040] The above are only preferred specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A battery thermal runaway graded warning method, characterized in that: The following steps are involved: S1, collect multi-dimensional data of battery operation through a multi-sensor system; S2, processing multi-dimensional data to perform data optimization and adaptive threshold generation; S3, identifying thermal runaway stage characteristics based on optimized multi-dimensional data and adaptive thresholds; S4. Trigger a graded alarm based on the thermal runaway feature identification results.

2. The battery thermal runaway graded warning method according to claim 1, characterized in that: The multi-dimensional data of battery operation collected through the multi-sensor system specifically includes: collecting electrolyte leakage data through the VOC sensor in the multi-sensor system; collecting carbon monoxide gas data through the CO sensor in the multi-sensor system; collecting smoke concentration data through the smoke sensor in the multi-sensor system; collecting temperature data through the temperature sensor in the multi-sensor system; and outputting multi-dimensional data including electrolyte leakage data, carbon monoxide gas data, smoke concentration data and temperature data.

3. The battery thermal runaway graded warning method according to claim 1 or 2, characterized in that: Processing multi-dimensional data to perform data optimization and adaptive threshold generation specifically includes: executing a compensation strategy based on a rate of change on the multi-dimensional data to correct a sensor attenuation error, thereby obtaining optimized multi-dimensional data; and generating an adaptive threshold based on the optimized multi-dimensional data.

4. The battery thermal runaway graded warning method according to claim 3, characterized in that: A rate-of-change compensation strategy is performed on the multi-dimensional data to correct the sensor attenuation error, thereby obtaining optimized multi-dimensional data, specifically including: performing rate-of-change compensation on the electrolyte leakage data to obtain compensated electrolyte leakage data; performing rate-of-change compensation on the carbon monoxide gas data to obtain compensated carbon monoxide gas data; performing rate-of-change compensation on the smoke concentration data to obtain compensated smoke concentration data; performing rate-of-change compensation on the temperature data to obtain compensated temperature data. The optimized multi-dimensional data specifically includes: compensated electrolyte leakage data, compensated carbon monoxide gas data, compensated smoke concentration data, and compensated temperature data.

5. The battery thermal runaway graded warning method according to claim 3, characterized in that: Dynamically generating adaptive thresholds based on optimized multi-dimensional data specifically includes: establishing a multi-parameter dynamic model through an automatic coupling mechanism; generating dynamic thresholds based on state evolution trajectories; verifying threshold stability through parameter mutual feedback; and outputting adaptive thresholds optimized by dynamic coupling.

6. The battery thermal runaway graded warning method according to claim 3, characterized in that: Identifying thermal runaway stage characteristics based on optimized multi-dimensional data and adaptive thresholds specifically includes: judging whether the thermal runaway incipient stage characteristics are met based on the adaptive electrolyte leakage threshold and the compensated electrolyte leakage data; judging whether the thermal runaway pre-occurrence characteristics are met based on the carbon monoxide gas adaptive threshold, the temperature adaptive threshold, the compensated carbon monoxide gas data, and the compensated temperature data; judging whether the thermal runaway occurrence characteristics are met based on the compensated smoke concentration data, the compensated temperature data, the smoke concentration adaptive threshold, and the temperature adaptive threshold; and outputting the thermal runaway characteristic identification results.

7. The battery thermal runaway graded warning method according to claim 6, characterized in that: Determining whether the thermal runaway incipient stage characteristics are met based on the adaptive electrolyte leakage threshold and the compensated electrolyte leakage data specifically includes: calculating the compensated electrolyte leakage change rate based on the compensated electrolyte leakage data, comparing the compensated electrolyte leakage change rate with the adaptive electrolyte leakage threshold, and when the electrolyte leakage change rate is greater than the adaptive leakage rate threshold, determining that the thermal runaway incipient stage characteristics are met; otherwise, the thermal runaway incipient stage characteristics are not met; wherein the adaptive electrolyte leakage threshold can be adjusted lower as the electrolyte volatilization rate decays.

8. The battery thermal runaway graded warning method according to claim 5, characterized in that: Identifying the characteristics of the thermal runaway stage includes: comparing the change rate of the electrolyte leakage data with the first threshold of the adaptive threshold, and determining that the characteristics of the thermal runaway embryonic stage are met when the change rate of the electrolyte leakage data is greater than the adaptive leakage rate threshold; combining the change rates of the carbon monoxide gas data and the temperature data, and determining that the characteristics before the thermal runaway occurs are met when both the carbon monoxide gas data is greater than the gas concentration threshold and the temperature data change rate is greater than the temperature rise rate threshold; coupling the smoke concentration data and the temperature data, and determining that the characteristics of the thermal runaway occurrence are met when both the smoke concentration data is greater than the smoke concentration threshold and the temperature data is greater than the temperature critical threshold.

9. The battery thermal runaway graded warning method according to claim 5, characterized in that: Judging whether the characteristics before thermal runaway occur are met based on the carbon monoxide gas adaptive threshold, the temperature adaptive threshold, the compensated carbon monoxide gas data and the compensated temperature data specifically includes: calculating the compensated temperature change rate based on the compensated temperature data; when the compensated carbon monoxide gas data is higher than the carbon monoxide gas adaptive threshold and the compensated temperature change rate is greater than the temperature adaptive threshold, it is determined that the characteristics before thermal runaway occur are met; otherwise, the characteristics before thermal runaway occur are not met.

10. The battery thermal runaway graded warning method according to claim 5, characterized in that: Whether the characteristics of thermal runaway occurrence are met is judged based on the compensated smoke concentration data, the compensated temperature data, the smoke concentration adaptive threshold and the temperature adaptive threshold. Specifically, when the compensated smoke concentration data is higher than the smoke concentration adaptive threshold and the compensated temperature data is higher than the temperature adaptive threshold, the characteristics of thermal runaway occurrence are met; otherwise, the characteristics of thermal runaway occurrence are not met. Among them, when the ambient temperature is ≥60℃, the temperature adaptive threshold automatically rises.

Images