Battery thermal runaway early warning method based on multi-sensing information fusion

Abstract

The invention discloses a battery thermal runaway early warning method based on multi-sensing information fusion, which is suitable for safety management of a lithium ion battery. A fiber bragg grating sensor is embedded into a battery, temperature and strain multi-dimensional physical parameters of the battery are monitored in real time, external electrochemical parameters are combined, and a multi-source sensing information fusion system is constructed. And by analyzing a square curve of an internal strain derivative, temperature gradient change and voltage drop characteristics, a thermal runaway early-stage signal is extracted, and graded early warning is realized. According to the method, the hysteresis of traditional surface parameter monitoring is broken through, an early warning signal is provided before thermal runaway, and the safety and reliability of a battery system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium-ion batteries, and particularly to a method for predicting battery thermal runaway by fusing multi-sensor information. Background Art

[0002] Due to their high energy density and long cycle life, lithium-ion batteries have become key energy carriers in fields such as electric vehicles and energy storage systems. However, they are prone to thermal runaway under extreme conditions such as overcharging, mechanical extrusion, or high-temperature abuse, resulting in intense heat release, gas production, and even explosion, seriously threatening personal and property safety.

[0003] Current battery management systems mainly judge abnormalities by monitoring parameters such as voltage, current, and surface temperature, but there are significant defects: traditional methods rely on external parameter monitoring and cannot capture early pre-thermal runaway characteristics such as internal material decomposition and microscopic structural damage in the battery; the alarm threshold is set broadly, prone to false alarms or missed alarms, and often misses the best intervention time when an abnormality is detected. In the prior art, early warning is triggered by voltage difference or surface temperature threshold, but key signals such as internal gas production and strain mutation in the battery are not effectively utilized, and early warning of thermal runaway cannot be achieved in a timely manner. Therefore, there is an urgent need for a method that can deeply fuse multi-dimensional physical and chemical parameters to achieve early warning of thermal runaway, so as to improve the safety and reliability of the battery system. Summary of the Invention

[0004] To address the deficiencies in current battery safety detection, this paper proposes a method for predicting battery thermal runaway by fusing multi-sensor information.

[0005] The method for predicting battery thermal runaway by fusing multi-sensor information proposed by the present invention includes the following steps:

[0006] S1. Embedding a fiber Bragg grating sensor into the lithium-ion battery to real-time monitor temperature and strain parameters, and synchronously collect external voltage signals;

[0007] S2. Constructing a multi-source sensor information acquisition system, analyzing the characteristics of the square curve of the internal strain derivative, temperature gradient, and voltage drop, and extracting pre-thermal runaway signals;

[0008] S3. Establishing multi-scenario warning thresholds based on mechanical, electrical, and thermal abuse experimental data, and combining real-time data to judge the risk level of thermal runaway;

[0009] S4. Triggering a hierarchical warning signal according to the risk level and outputting a warning message before thermal runaway.

[0010] Preferably, in step S1, a dual-grating fiber Bragg grating sensor is used to decouple the temperature and strain signals, and combined with the external voltage data, the internal microscopic changes of the battery are directly sensed. The fiber Bragg grating sensor includes a sensor for directly monitoring the electrolyte environment and a sensor encapsulated in a glass tube, which shields mechanical interference and decouples the temperature signal to ensure data reliability.

[0011] Preferably, in step S2, the reflected wavelength data is collected by a high-speed spectrometer, the temperature and strain sensitivity coefficients are calibrated, the peak characteristics of the strain derivative square curve are extracted, and the thermal runaway stage is divided in combination with the voltage drop rule.

[0012] The steps of the data acquisition method include:

[0013] S21. Use a dual-grating fiber Bragg grating sensor to decouple the internal temperature and strain signals of the battery;

[0014] S22. Synchronously collect the reflected wavelength data through a high-speed spectrometer and record the external voltage signal;

[0015] S23. Determine the temperature and strain sensitivity coefficients through a calibration experiment to ensure accurate data;

[0016] S24. Process the data in real time and extract the strain derivative square curve and temperature change characteristics.

[0017] Preferably, in S2, the fiber sensor and voltage data are correlated and analyzed through a multi-channel synchronous acquisition system, and the sliding window algorithm is used to eliminate the fluctuation interference, improving the real-time performance and accuracy of multi-cell monitoring.

[0018] Preferably, in S3, according to the strain mutation characteristics of the mechanical abuse experiment, the lithium plating and gas generation characteristics of the electrical abuse experiment, and the temperature sudden increase and strain sudden drop phenomena of the thermal abuse experiment, the mechanical damage, overcharge, and thermal runaway trigger thresholds are set respectively to improve the multi-scenario adaptability.

[0019] Preferably, in S4, the hierarchical warning signals include the abnormal characteristic peaks of the internal strain derivative square curve caused by gas production or material decomposition, the temperature sudden increase exceeding the preset threshold, and the voltage drop during the internal short circuit or thermal runaway stage.

[0020] The present invention can provide early warning signals before thermal runaway occurs, significantly improving the safety and reliability of the battery system, and is applicable to scenarios such as electric vehicles and energy storage systems. Description of the Drawings

[0021] Figure 1 It is a flowchart of a battery thermal runaway warning method based on multi-sensor information fusion proposed by the present invention. Detailed Embodiment

[0022] Refer to Figure 1, a battery thermal runaway warning method based on multi-sensor information fusion proposed by the present invention includes the following steps:

[0023] S1. Embed a fiber Bragg grating sensor into the lithium-ion battery to monitor the internal temperature and strain parameters of the battery in real time, and synchronously collect external voltage signals;

[0024] In order to ensure the accuracy and real-time nature of the monitoring, in this embodiment, in S1, a dual-grating fiber Bragg grating sensor is used to decouple the internal temperature and strain signals of the battery, and the reflected wavelength data is synchronously collected by a high-speed spectrometer, while recording the external voltage signal.

[0025] S2. Construct a multi-source sensing information fusion model for the monitoring data, and extract the pre-thermal runaway signals by analyzing the square curve of the internal strain derivative, the temperature gradient change, and the voltage drop characteristics;

[0026] In order to accurately extract the pre-thermal runaway signals, in this embodiment, in S2, by extracting the peak characteristics of the square curve of the internal strain derivative, and combining the temperature change and voltage drop data, a multi-source sensing information fusion model is constructed, and the sliding window algorithm is used to eliminate the interference of data fluctuations.

[0027] The method for obtaining multi-source sensing data includes:

[0028] S21. Use a dual-grating fiber Bragg grating sensor to decouple the internal temperature and strain signals of the battery;

[0029] S22. Synchronously collect the reflected wavelength data by a high-speed spectrometer and record the external voltage signal;

[0030] S23. Determine the sensitivity coefficients of temperature and strain through a calibration experiment to ensure the accuracy of the decoupled data;

[0031] S24. Process the data in real time and extract the characteristics of the square curve of the strain derivative and the temperature gradient change.

[0032] Among them, during the data acquisition process, in this embodiment, in S23, through the pre-conducted temperature and strain calibration experiments, the sensitivity coefficients of the sensors are determined to correct the actual monitoring data and ensure the decoupling accuracy of the temperature and strain signals.

[0033] S3. Based on the experimental data of mechanical abuse, electrical abuse, and thermal abuse, establish multi-scenario warning thresholds, and combine the real-time monitoring data with the thresholds to judge the thermal runaway risk level;

[0034] S4. Trigger a hierarchical warning signal according to the risk level and output a warning message before thermal runaway occurs inside the battery.

[0035] The present invention can provide early warning signals before thermal runaway occurs, improving the safety and reliability of the battery system.

[0036] Embodiment 1

[0037] In this battery thermal runaway warning method based on multi-sensing information fusion, the specific steps are as follows:

[0038] (1) Embed a dual-gate fiber Bragg grating sensor inside the battery to monitor temperature and strain signals in real time and synchronously collect external voltage data;

[0039] (2) Use a high-speed spectrometer to collect reflected wavelength data at a frequency of once per second, decouple the temperature and strain signals through calibration coefficients, and calculate the square curve of the internal strain derivative;

[0040] (3) Based on the pre-established mechanical, electrical, and thermal abuse experiment thresholds, set the mechanical damage, overcharge, and thermal runaway trigger thresholds; (4) Analyze the peak characteristics of the strain derivative square curve, sudden temperature increase, and voltage drop in real time to judge the thermal runaway risk level; (5) When an abnormal peak in the strain derivative square curve is detected, the system issues a first-level alarm; if the temperature suddenly increases beyond the threshold or the voltage drops suddenly, trigger a second-level or third-level warning, prompting immediate emergency measures such as cooling down and power-off.

[0041] The above is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A battery thermal runaway warning method based on multi-sensor information fusion, characterized in that, It includes the following steps: S1. Embed the fiber Bragg grating sensor into the lithium-ion battery to monitor the temperature and strain parameters inside the battery in real time, and synchronously collect the external voltage signal; S2. Construct a multi-dimensional sensing information early warning system for the monitored data. Extract the early warning signals before thermal runaway by analyzing the square curve of the internal strain derivative, the change of the temperature gradient, and the voltage drop characteristics; S3. Based on the experimental data of mechanical abuse, electrical abuse, and thermal abuse, establish multi-scenario early warning thresholds, and combine the real-time monitoring data with the thresholds to judge the thermal runaway risk level; S4. Trigger hierarchical early warning signals according to the risk level, and output early warning information before thermal runaway occurs inside the battery.

2. The method for warning of battery thermal runaway by multi-sensor information fusion according to claim 1, characterized in that In S1, the fiber optic sensor is used to directly monitor the temperature and strain in the electrolyte environment; the sensor is encapsulated with a glass tube to shield mechanical strain interference and decouple the temperature signal.

3. A method for battery thermal runaway warning with multi-sensor information fusion according to claim 1, characterized in that, The construction method of the multi-dimensional sensing early warning in S2 is as follows: S21. Collect the reflected wavelength data of the fiber Bragg grating through a high-speed spectrometer, and calibrate the sensitivity coefficients of temperature and strain; S22. Extract the peak characteristics of the square curve of the internal strain derivative; S23. Combine the voltage drop characteristics to establish the rules for dividing the thermal runaway stage.

4. A method for battery thermal runaway early warning with multi-sensor information fusion according to claim 3, characterized in that In S2, it also includes a multi-channel synchronous acquisition system for correlation analysis of the fiber optic sensor and voltage data to ensure the safety and real-time performance of multiple battery cells.

5. A method for battery thermal runaway early warning with multi-sensor information fusion according to claim 1, characterized in that, The analysis method in S2 uses a sliding window algorithm to eliminate data fluctuation interference.

6. A method for battery thermal runaway warning by multi-sensor information fusion according to claim 1, characterized in that The determination method of the multi-scenario early warning threshold in S3 is constructed based on the characteristic peak temperature change and voltage drop information of the square curve of the internal strain derivative in the mechanical abuse, electrical abuse, and thermal abuse experiments.

7. A method for warning of battery thermal runaway by multi-sensor information fusion according to claim 1, characterized in that, The hierarchical early warning signals in S4 include the abnormal characteristic peaks of the square curve of the internal strain derivative caused by gas production or material decomposition, the sudden increase in temperature exceeding the preset threshold, and the voltage drop in the internal short circuit or thermal runaway stage.

Images