CN-122017599-A - Lithium battery thermal runaway analysis method and system

CN122017599ACN 122017599 ACN122017599 ACN 122017599ACN-122017599-A

Abstract

The invention belongs to the technical field of lithium batteries and discloses a thermal runaway analysis method and a system of the lithium battery, wherein the method comprises the steps of obtaining a multisource observation sequence to form an alignment observation set, extracting response time sequence characteristics from a detector, grouping operation time periods to form a working condition group, generating a response time sequence chain and actual peak distribution from the working condition group to form a working condition grouping response map, constructing a topology change indication quantity, comparing the topology change indication quantity with a topology change threshold set to determine a super threshold detector pair to generate a space alias mark, searching a hypothetical path subset by using the working condition group identification, screening to obtain a candidate release path set, calculating sequence consistency rate and peak consistency rate, screening an effective path set, obtaining a dynamic space mapping relation, outputting a risk source positioning result, executing differentiation treatment based on the risk source positioning result, and determining a risk source by extracting the response time sequence characteristics, screening the effective path and generating the dynamic space mapping relation, thereby reducing the risk of error treatment.

Inventors

REN YUANYUAN
PENG WEI
ZHANG SHAOJIE
LI XIN
GAO XINGYU
Sheng Rongkai
MENG FANXU

Assignees

安徽理工大学

Dates

Publication Date: 20260512
Application Date: 20260303

Claims (10)

1. A method for analyzing thermal runaway of a lithium battery, comprising: The method comprises the steps that a multi-source observation sequence of a battery pack in a current operation period is obtained through a detector to form an alignment observation set, wherein the multi-source observation sequence comprises a gas concentration sequence, a temperature sequence, a pressure sequence, a fan working condition sequence and a treatment action execution sequence; Extracting response timing characteristics for each detector based on the alignment observation set, the response timing characteristics including arrival time, rising edge duration, peak time, and peak amplitude; grouping the operation time periods by using a fan working condition sequence to form working condition groups, generating a response time sequence chain and actual peak distribution for each working condition group, and forming a working condition grouping response map; constructing a topology change indicating quantity based on the working condition grouping response map, comparing the topology change indicating quantity with a topology change threshold set to determine a super-threshold detector pair and generating a spatial alias mark; when the spatial aliasing mark exists, searching a hypothetical path subset from a pre-built hypothetical path library by taking the working condition group mark as an index, and screening according to the consistency of the actual response sequence of the super-threshold detector pair and the expected response sequence of each hypothetical path to obtain a candidate bleed path set; calculating the sequence consistency rate and the peak consistency rate of each candidate release path, and comparing and screening the sequence consistency rate and the peak consistency rate with corresponding thresholds respectively to obtain an effective path set; weighting, summarizing and normalizing based on the effective path set to obtain a dynamic space mapping relation, and outputting a risk source positioning result comprising a target module and a positioning credible mark according to the dynamic space mapping relation; Positioning beaconing based on risk source positioning results may be gated to perform differential treatments.
2. The method of claim 1, wherein the method of extracting the response timing characteristics for each detector comprises: Pre-configuring corresponding response thresholds for each type of detector; Traversing the aligned observation set along a time axis, and determining the moment as the arrival time when the observation value of a certain detector meets the response threshold value for the first time and reaches the persistence constraint; counting the time span of the continuous response state of the observed value from the arrival time to obtain the rising edge duration; and in a continuous time interval corresponding to the response state, positioning the time when the observed value reaches the maximum value, determining the peak value time, and recording the maximum value as the peak value amplitude.
3. The method for analyzing thermal runaway of a lithium battery according to claim 1, wherein the method for acquiring the operating condition grouping response map comprises the following steps: dividing the intervals with the same working condition identification and continuous time into the same working condition group based on the working condition sequence of the fan; summarizing the arrival time, rising edge duration time, peak time and peak amplitude of all detectors in each working condition group, and sequencing the detectors by taking the arrival time as a main sequence to obtain a response time sequence chain of the working condition group; Forming actual peak distribution by taking peak time and peak amplitude as indexes; And constructing a response time sequence chain and actual peak distribution corresponding to each working condition group as a response set, and establishing a corresponding relation between each working condition group and the response set to form a working condition grouping response map.
4. The method of claim 1, wherein the method of constructing the topology change indicator comprises: for each working condition group, reading a detector pair set matched with the working condition group identifier from a preset detector pair configuration table; For each detector pair in the same working condition group, calculating the difference value of arrival time of each detector pair as an arrival time difference, calculating the difference value of peak time of each detector pair as a peak time difference, and calculating the ratio of peak amplitude of each detector pair as a peak amplitude ratio; and splicing the arrival time differences, the peak value time differences and the peak value amplitude ratios of all the detector pairs in the working condition group in sequence according to a preset sequence to obtain the topology change indication quantity of the working condition group.
5. The method of claim 1, wherein the method of determining the pair of super-threshold detectors and generating the spatial aliasing signature comprises: in the calibration period, forming a normal fluctuation interval for three types of components of each detector pair respectively, wherein the normal fluctuation interval comprises an arrival time difference normal interval, a peak time difference normal interval and a peak amplitude ratio normal interval; when the operation period is compared, whether the arrival time difference of the detector pair falls into the arrival time difference normal interval, whether the peak time difference falls into the peak time difference normal interval and whether the peak amplitude ratio falls into the peak amplitude ratio normal interval are respectively judged; when any component exceeds the corresponding normal fluctuation interval, judging that the component meets the threshold exceeding condition, and marking the detector pair as a threshold exceeding detector pair under the working condition group; if any super-threshold detector pair exists in the working condition group, the working condition group is considered to be inconsistent, and the output space alias mark is the existence; if no super-threshold detector pair is present, the output spatial aliasing is marked as not present.
6. The method for analyzing thermal runaway of a lithium battery according to claim 1, wherein the method for acquiring the candidate bleed-off path set comprises: pre-establishing an assumed path library, wherein the assumed path library is an assumed path set pre-established when the battery pack is deployed; retrieving a hypothetical path subset matched with the working condition group from a hypothetical path library by taking the current working condition group identifier as an index; and for each hypothesized path, checking whether the response sequence of the super-threshold detector pairs is consistent with the expected response sequence of the hypothesized path one by one, and reserving the hypothesized path when the number of the consistent super-threshold detector pairs is not less than a preset matching number threshold value to obtain a candidate bleed path set.
7. The method for analyzing thermal runaway of a lithium battery according to claim 1, wherein the method for acquiring the effective path set is as follows: The expected response sequence of the candidate paths is compared with the response time sequence chain actually extracted in the alignment observation set one by one, and the comparison unit is used for comparing the sequence relation of the detector pairs one by one; When the expected sequence is consistent with the actual sequence, the expected sequence is marked as inconsistent, and the ratio of the number of the statistically consistent detectors to the total number of the statistically participated detectors is used for obtaining the sequence consistency rate; the expected peak amplitude of each detector of the candidate path and the peak amplitude of each detector in the actual response are sequenced from big to small according to the peak amplitude, and the detector in the front r% is taken to form an expected peak priority detector set and an actual peak priority detector set; aligning and comparing the expected peak value priority detector set with the actual peak value priority detector set, counting the number of detectors in the intersection of the expected peak value priority detector set and the actual peak value priority detector set, and taking the ratio of the intersection number to the total number of the actual peak value priority detector set as a peak value consistency rate; When the sequence consistency rate is larger than a preset sequence consistency threshold value and the peak consistency rate is larger than a preset peak consistency threshold value, the candidate release paths are judged to be effective paths, and when any one item does not meet the requirement, the candidate release paths are rejected, and the effective paths are unified to be an effective path set.
8. The method for analyzing thermal runaway of a lithium battery according to claim 1, wherein the method for obtaining the dynamic space mapping relation based on the weighted summary and normalization of the effective path set comprises the following steps: Maintaining a correlation table of the detector and the module position index in advance, wherein the correlation table comprises a detector identifier, an adjacent module position index, a partition identifier, a hypothetical path identifier and a pointing weight; for each effective path in the effective path set, reading the expected response sequence of the effective path binding, and determining a detector set related to the path; reading the corresponding adjacent module position index and the pointing weight of each detector related to the effective path under the hypothetical path identifier according to the association table; the same module position index is subjected to accumulation of the pointing weights from different detectors to obtain the in-path pointing strength of the effective path to each module position index; configuring path weights for each effective path, wherein the path weights are obtained by carrying out equal weight synthesis on the sequence consistency rate and the peak value consistency rate of the effective path; aiming at each module position index, summarizing the intra-path directional strength of all effective paths to the module position index, and carrying out weighted accumulation according to the corresponding path weight to obtain a fusion directional value of the module position index; And carrying out normalization processing on the fusion pointing values of all the module position indexes to obtain normalized pointing values, thereby obtaining dynamic space mapping relation between the module position indexes and the normalized pointing values.
9. The method for analyzing thermal runaway of a lithium battery according to claim 1, wherein the method for acquiring the risk source positioning result comprises the following steps: Reading the normalized pointing value corresponding to each module position index, determining the module position index with the largest normalized pointing value as a target module, determining the module position index with the second largest normalized pointing value, and recording the normalized pointing value as the next highest normalized pointing value; setting a credible threshold C and a discrimination threshold D, comparing a normalized pointing value corresponding to the target module with the credible threshold C, continuing to judge when the normalized pointing value is not less than C, and outputting a positioning credible mark as failed when the normalized pointing value is less than C; On the premise of judging through a credible threshold, calculating a difference value between a normalized pointing value corresponding to the target module and a next highest normalized pointing value, and comparing the difference value with a discrimination threshold D; when the difference value is not smaller than D, the output positioning credible mark is passed, and when the difference value is smaller than D, the output positioning credible mark is not passed; And when the two positioning credible marks of the target module are passed, outputting the positioning credible marks as a risk source positioning result.
10. A lithium battery thermal runaway analysis system for implementing a lithium battery thermal runaway analysis method according to any one of claims 1 to 9, comprising: the data acquisition module acquires a multi-source observation sequence of the battery pack in the current operation period through the detector to form an alignment observation set, wherein the multi-source observation sequence comprises a gas concentration sequence, a temperature sequence, a pressure sequence, a fan working condition sequence and a treatment action execution sequence; The response extraction module is used for extracting response time sequence characteristics for each detector based on the alignment observation set, wherein the response time sequence characteristics comprise arrival time, rising edge duration time, peak time and peak amplitude; The map generation module is used for grouping the operation time periods by utilizing the fan working condition sequence to form working condition groups, and generating a response time sequence chain and actual peak distribution for each working condition group to form a working condition grouping response map; The alias marking module is used for constructing a topology change indicating quantity based on the working condition grouping response map, comparing the topology change indicating quantity with a topology change threshold set to determine a super-threshold detector pair and generating a spatial alias mark; The bleed-off path screening module is used for searching a hypothetical path subset from a pre-built hypothetical path library by taking the working condition group identifier as an index when the spatial alias mark exists, and screening to obtain a candidate bleed-off path set according to the consistency of the actual response sequence of the super-threshold detector pair and the expected response sequence of each hypothetical path; The effective path screening module is used for calculating the sequence consistency rate and the peak consistency rate of each candidate release path, and comparing and screening the sequence consistency rate and the peak consistency rate with corresponding thresholds respectively to obtain an effective path set; The risk source positioning module is used for weighting, summarizing and normalizing the effective path set to obtain a dynamic space mapping relation and outputting a risk source positioning result comprising a target module and a positioning credible mark according to the dynamic space mapping relation; And the risk treatment module is used for gating the positioning beaconing mark based on the risk source positioning result to execute differentiated treatment.

Description

Lithium battery thermal runaway analysis method and system Technical Field The invention relates to the technical field of lithium batteries, in particular to a thermal runaway analysis method and a thermal runaway analysis system for a lithium battery. Background With the wide deployment of lithium ion batteries in electrochemical energy storage cabinets, energy storage containers, electric automobile power battery packs and other scenes, the monitoring and disposal of surrounding thermal runaway precursors are commonly realized by adopting various sensors in a coordinated manner, wherein a gas detector, a temperature detector and a pressure detector are arranged in a cabinet body or a battery pack, and an air duct, a discharge channel, a flow deflector and a compartment structure are matched, so that decomposed gas and heat change generated under abnormal working conditions can be timely perceived, and further safety actions such as isolation, power reduction, inerting, water-based cooling inhibition and the like are triggered, so that the thermal runaway expansion risk is reduced. In the prior art, more schemes focus on design optimization of an exhaust channel and a diversion structure, or analysis of gas diffusion and concentration distribution based on fixed geometry and fixed ventilation conditions, so as to improve the discharge controllability in the structural design level. However, the spatial mapping relationship and the bleed path of the multiple default sensors in the above scheme remain stable during the operation period, and the risk source positioning often depends on a preset fixed correspondence rule or fixed propagation sequence assumption of the "sampling point-spatial position". In actual operation, the energy storage cabinet and the battery pack may experience working conditions such as transportation, installation and maintenance, vibration impact or slight collision, and the like, and the boundaries of the guide vane, the air duct member, the discharge channel and the compartment may be deformed, displaced or partially shielded, so that the precedence relationship and the peak distribution position of the discharge gas reaching each detector are changed. The key problem caused by the method is that the response time sequence and the peak distribution of the sensor deviate from the preset mapping, if the positioning rule under the fixed topology assumption is still used, the risk source positioning distortion is easy to occur, namely, the module which is really discharged is misjudged as other modules, and further the action object of the subsequent treatment action is deviated. The positioning distortion can directly amplify the treatment risk, namely under the condition of mispositioning, actions such as isolation, inerting or water-based cooling inhibition can be applied to a non-risk module and cannot timely cover a real discharge module, so that heat continues to accumulate and spread to an adjacent module, meanwhile, treatment resources are unnecessarily consumed, insufficient treatment capacity or unstable action effect can occur when a higher risk stage is subsequently entered, even unnecessary derating and shutdown of a system are caused, and additional safety and operation loss are brought. In view of the above, the present invention provides a method and a system for analyzing thermal runaway of a lithium battery to solve the above problems. Disclosure of Invention In order to overcome the defects in the prior art, the invention provides the following technical scheme that the method for analyzing the thermal runaway of the lithium battery comprises the following steps: The method comprises the steps that a multi-source observation sequence of a battery pack in a current operation period is obtained through a detector to form an alignment observation set, wherein the multi-source observation sequence comprises a gas concentration sequence, a temperature sequence, a pressure sequence, a fan working condition sequence and a treatment action execution sequence; Extracting response timing characteristics for each detector based on the alignment observation set, the response timing characteristics including arrival time, rising edge duration, peak time, and peak amplitude; grouping the operation time periods by using a fan working condition sequence to form working condition groups, generating a response time sequence chain and actual peak distribution for each working condition group, and forming a working condition grouping response map; constructing a topology change indicating quantity based on the working condition grouping response map, comparing the topology change indicating quantity with a topology change threshold set to determine a super-threshold detector pair and generating a spatial alias mark; when the spatial aliasing mark exists, searching a hypothetical path subset from a pre-built hypothetical path library by taking the working condition group mark as an inde