Lithium ion battery thermal runaway early warning method based on multi-parameter monitoring

Abstract

The present application relates to a kind of lithium ion battery thermal runaway early warning method based on multi-parameter monitoring, to early identification and prevent the risk of lithium ion battery thermal runaway.The present application constructs combination early warning mechanism, based on the dynamic monitoring of voltage, current, temperature and other basic parameters, combined with statistical analysis and electrochemical model, forms multi-level early warning characteristic parameter system.The present application can be based on the synergistic effect of multiple parameters, dynamically adjust early warning boundary, significantly improve the accuracy and adaptability of battery thermal runaway early warning, provide effective guarantee for battery safe operation, applicable to power battery and energy storage battery and other various application scenarios.

Description

TECHNICAL FIELD

[0001] The present application relates to a lithium ion battery thermal runaway early warning method based on multi-parameter monitoring, belonging to the technical field of lithium ion battery safety monitoring and early warning. BACKGROUND

[0002] Lithium ion batteries are widely used in portable electronic devices, electric vehicles and renewable energy storage due to their superior energy density, long cycle life and low self-discharge rate. However, lithium ion batteries are prone to thermal runaway under extreme conditions such as high temperature, overcharge and internal short circuit, leading to battery overheating, fire and even explosion. These safety hazards have attracted widespread attention to the safety of lithium ion batteries, and effective thermal runaway early warning technology is urgently needed to ensure the safe use of batteries.

[0003] Currently, the main methods for lithium ion battery thermal runaway early warning include the following:

[0004] Single parameter monitoring: Many existing technologies rely only on the monitoring of a single parameter, such as temperature, single cell voltage or current. These methods usually set a threshold value, and once the monitoring value exceeds the threshold value, an alarm is issued. However, single parameter monitoring is easily affected by noise and environmental factors, which may lead to false positives or false negatives.

[0005] Model-based prediction: Some researchers use electrochemical models of batteries to predict thermal runaway. These models are based on complex mathematical equations and can provide some accuracy, but their calculation process is complex and real-time performance is poor, making them difficult to apply to rapidly changing real-world situations.

[0006] Data-driven methods: In recent years, with the development of machine learning and big data technology, data-driven methods have gradually emerged. These methods can identify potential thermal runaway risks by learning from historical data. However, these methods still require a large amount of data support and may have the problem of model overfitting.

[0007] Although existing technologies have made some progress in lithium ion battery thermal runaway early warning, there are still some shortcomings, such as insufficient accuracy, poor real-time performance and insufficient comprehensive analysis of multiple factors. Therefore, there is an urgent need for an early warning method that can consider multiple parameters comprehensively.

[0008] According to the search, the Chinese patent with the patent number 2021111691677 discloses a cylindrical lithium ion battery thermal anomaly detection and positioning method, which comprises the following steps: establishing a normalized temperature spatial distribution model when a disturbance and a thermal anomaly exist; identifying the parameters of the normalized temperature spatial distribution model according to a preset optimization objective function; collecting battery surface temperature data, battery current and terminal voltage data, and solving the error system by the successive approximation method to obtain an output temperature estimation error; the distributed residual quantity of the output temperature estimation error can be obtained, and an evaluation value can be obtained by substituting the distributed residual quantity into a preset residual evaluation function; when the evaluation value is greater than a preset threshold value, the thermal anomaly existing at the spatial position is located. However, the early warning method of the patent is single, and the algorithm is complex, so that the situation of missed judgment is easily caused. SUMMARY

[0009] In order to solve the above problems, the present application discloses a lithium ion battery thermal runaway early warning method based on multi-parameter monitoring, and the specific technical scheme is as follows:

[0010] A lithium ion battery thermal runaway early warning method based on multi-parameter monitoring, comprising the following steps:

[0011] Step 1. Collecting the voltage V(t), current I(t) and temperature T(t) data of the battery monomer, and monitoring the abnormal value and / or abnormal growth rate to generate the first type of early warning characteristic parameter;

[0012] Step 2. Data statistical analysis is performed on the battery monomer voltage to detect the mean deviation and / or information entropy deviation to generate the second type of early warning characteristic parameter;

[0013] Step 3. Based on the battery electrochemical model, the particle swarm optimization method is used to identify the electrochemical parameters of the battery monomer, and the third type of early warning characteristic parameter is generated according to the abnormal electrochemical parameters obtained by identification;

[0014] Step 4. When any parameter in any early warning characteristic parameter of steps 1-3 is abnormal, the early warning is started immediately.

[0015] Further, the voltage V(t), current I(t) and temperature T(t) data of the battery monomer collected in step 1 are preprocessed before the abnormal value and / or abnormal growth rate operation, and the specific process of the preprocessing is as follows:

[0016] 1.1: Data cleaning is performed to remove noise and abnormal value processing on the collected data, so as to ensure the quality and accuracy of the data and remove obviously incorrect data beyond the reasonable range;

[0017] 1.2: data segmentation, continuous time series data is segmented, the segmentation method is based on time window or event trigger, ensuring that each segment of data has similar characteristics.

[0018] Further, the abnormal values of the voltage V(t), current I(t) and temperature T(t) of the battery cell in step 1 are determined by comparing the measured values with the preset threshold value, and the measured values of the voltage V(t), current I(t) and temperature T(t) exceeding the preset threshold value are determined as abnormal values.

[0019] Further, the calculation formula of the growth rate in step 1 is as follows:

[0020]

[0021] Where y represents the type of collected data, including voltage, current or temperature, σ y is the real-time growth rate of data y, y(i) is the real-time value of data type y at the i-th moment, and y(i+1) is the real-time value of data type y at the i+1-th moment;

[0022] The preset voltage, current and temperature growth rate threshold, when any of the real-time growth rates of voltage, current or temperature exceeds the corresponding preset growth rate threshold of voltage, current or temperature, is determined as an abnormal growth rate.

[0023] Further, the calculation formula of the mean deviation in step 2 is as follows:

[0024]

[0025] Where V i is the voltage of the battery cell at the i-th moment, is the mean value of all voltages, and N is the number of data points;

[0026] The calculation formula of information entropy is as follows:

[0027]

[0028] Where P(V j ) is the probability distribution of the voltage value of the j-th cell, k is the number of voltage value categories,

[0029] Based on statistical analysis of historical data, the determination criteria of mean deviation and information entropy deviation are determined, and when any measured value of mean deviation or information entropy deviation is greater than the determination criteria of corresponding mean deviation or information entropy deviation, the second type of early warning feature parameter is generated.

[0030] Further, in step 3, the lithium ion concentration C Li+ and the equivalent conductivity σ sepAs an electrochemical parameter to characterize the internal state of the battery, lithium ion concentration C Li+ and separator equivalent conductivity σ sep The change directly reflects the potential failure characteristics of short circuit, separator failure in the battery,

[0031] The specific process of electrochemical parameter identification is: combining electrochemical model and particle swarm optimization algorithm, real-time monitoring, collecting voltage, current external signals in the process of battery operation, and dynamically inverting the change of C Li+ and σ sep Make the simulation voltage fit the collected voltage, that is, the loss function of parameter identification, calculate the difference between the simulation voltage and the measured voltage, and the difference between the simulation voltage and the measured voltage is minimum, then the value of the objective function reaches the minimum, so as to obtain the electrochemical parameter value in the electrochemical model;

[0032] The particle swarm optimization algorithm takes the root mean square error minimization of the simulation voltage and the measured voltage of the lithium ion battery as the objective function, iteratively optimizes the model parameters, and obtains high-precision real-time C Li+ and σ sep electrochemical parameters,

[0033]

[0034] Where V sim (i) is the simulation voltage at the i-th time point, V meas (i) is the measured voltage at the i-th time point, and N is the number of data points.

[0035] Further, the process of obtaining the simulation voltage is: collecting the voltage data of the battery monomer in the process of battery operation, and obtaining the simulation voltage of the battery through the electrochemical model simulation.

[0036] Further, in the combination early warning mechanism of the third type of electrochemical parameters in step 3, based on the monitored lithium ion concentration C Li+ and separator equivalent conductivity σ sep Electrochemical parameters, set the combination early warning boundary, when the lithium ion concentration C Li+ and separator equivalent conductivity σ sep The combination of electrochemical parameters meets the specific conditions, triggering the thermal runaway early warning, and the specific conditions are:

[0037]

[0038] Wherein is the separator equivalent conductivity threshold dynamically adjusted according to the lithium ion concentration, and the formula is as follows:

[0039]

[0040] wherein σ0 is the base separator equivalent conductivity threshold, k is an adjustment coefficient, used to represent the influence of lithium ion concentration on the conductivity of the separator, C ref is the reference lithium ion concentration.

[0041] Further, in order to adapt to high load or low temperature operating environment, the combined early warning boundary is dynamically adjusted in combination with real-time data and historical data, when the battery is in a high load state for a long time, then the threshold is reduced so as to issue an early warning signal in advance, once the combined early warning condition is triggered, the system will issue an alarm to the user, suggesting to stop using the battery immediately;The relevant data triggering the early warning are automatically recorded, including abnormal parameters and working condition information at the time of triggering.

[0042] The working principle of the present application is:

[0043] The present application generates multiple early warning characteristic parameters by real-time acquisition of battery monomer voltage, current and temperature data, and combination with electrochemical parameter analysis, so as to realize effective early warning of thermal runaway. This method considers multiple key parameters at the same time, provides more comprehensive monitoring information, reduces the risk of false positives and false negatives;Through statistical analysis and parameter identification, the thermal runaway risk of the battery can be more accurately judged, and the accuracy and reliability of the early warning can be improved;The method of the present application can be applied to different types and scales of lithium ion batteries, and has wide application prospect.

[0044] The beneficial effects of the present application are:

[0045] The present application not only constructs multiple early warning characteristic parameters, but also can more accurately judge the thermal runaway risk of the battery through statistical analysis and parameter identification.

[0046] The implementation of the present application will significantly improve the safety of lithium ion batteries, reduce the risk of thermal runaway events, prevent the occurrence of thermal runaway events, and provide practical technical support for the safe use of batteries, which has important practical significance and application value. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 is the flow chart of the lithium ion battery thermal runaway early warning method based on multi-parameter monitoring of the present application. DETAILED DESCRIPTION

[0048] The present application will be further illustrated below in combination with the drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present application and not to limit the scope of the present application.

[0049] In combination with the drawings Figure 1 It can be seen that the specific execution process of the present application is:

[0050] 1. Data collection: Collect the voltage, current, and temperature data of the battery cells through the sensors installed in the battery pack and store them in the database for subsequent analysis.

[0051] 2. Data preprocessing:

[0052] First, perform data cleaning to remove noise and outliers from the collected data, ensuring data quality and accuracy. For example, remove obvious erroneous data (such as negative values or data outside the reasonable range). Then segment the continuous time series data for subsequent analysis. The segmentation method can be based on time windows or event triggers, ensuring that each segment of data has similar characteristics.

[0053] 3. Outlier detection: First type of early warning characteristic parameter

[0054] Monitor the values of voltage, current, and temperature. The outliers of the battery cell voltage V(t), current I(t), and temperature T(t) are determined by comparing the measured values with the preset threshold.

[0055] Calculate the abnormal growth rate. When the real-time data value or the growth rate of the data value exceeds the set threshold, trigger the early warning. The formula for calculating the growth rate is as follows:

[0056]

[0057] where y represents various types of collected data, including voltage, current, or temperature, σ y is the growth rate of data y, y(i) is the real-time value of data type y at the i-th moment, and y(i+1) is the real-time value of data type y at the i+1-th moment.

[0058] 4. Statistical analysis: Second type of early warning characteristic parameter

[0059] Analyze historical data to calculate mean deviation and information entropy, generate the second type of early warning characteristic parameter, and assess the battery state in real time:

[0060] The formula for calculating the mean deviation is as follows:

[0061]

[0062] where V i is the voltage of the battery cell at the i-th moment, is the mean value of all voltages, and N is the number of data points.

[0063] The formula for calculating the information entropy is as follows:

[0064]

[0065] where P(V j) is the probability distribution of the j-th type of monomer voltage value, and k is the number of voltage value types. When the mean deviation is too large or the information entropy deviation is too large, the second type of early warning feature parameter is generated;

[0066] The determination criteria for the mean deviation being too large and the information entropy deviation being too large are based on statistical analysis of historical data. Due to the relevance of specific battery pack structures, cell types, etc., there may be different threshold determination criteria, therefore, the determination criteria need to be formulated according to the historical early normal data of the actual battery pack.

[0067] 5. Electrochemical parameter monitoring: the third type of early warning feature parameter

[0068] The equivalent conductivity σ of the separator and the negative lithium ion concentration C are obtained in real time by using an electrochemical model combined with a particle swarm optimization algorithm. sep Li+ First, the voltage data of the battery monomer is collected during the operation of the battery, and then the voltage of the battery is obtained by simulation through the electrochemical model. The simulated voltage is fitted to the collected voltage (this function is the loss function of parameter identification, which is the difference between the simulated voltage and the measured voltage, and the goal of parameter identification is to minimize the value of this objective function), so as to obtain the electrochemical parameter values in the electrochemical model, which is parameter identification. For the voltage data of the battery monomer, the values of the two electrochemical parameters are obtained through parameter identification, and then it is judged whether they exceed the safety threshold, so as to realize early warning. The optimization objective function is to minimize the root mean square error between the simulated voltage and the measured voltage:

[0069]

[0070] Where, V sim (i) is the simulated voltage at the i-th time point, V meas (i) is the measured voltage at the i-th time point, and N is the number of data points.

[0071] 6. Combined early warning mechanism:

[0072] In the combined early warning mechanism of the third type of electrochemical parameter, the system sets the combined early warning boundary based on the existing monitored electrochemical parameters, including lithium ion concentration and separator equivalent conductivity. When the combination of these parameters meets certain conditions, the system triggers a thermal runaway early warning. The specific conditions include:

[0073]

[0074] The calculation formula is as follows:

[0075]

[0076] Where, ​is the equivalent conductivity threshold of the separator dynamically adjusted according to the lithium ion concentration, σ0 is the equivalent conductivity threshold of the base separator, k is the adjustment coefficient, used to represent the influence of lithium ion concentration on the conductivity of the separator, C ref is the reference lithium ion concentration.

[0077] In order to adapt to different operating environments (such as high load or low temperature working conditions), the system will dynamically adjust the combined early warning boundary by combining real-time data and historical data. For example, when the battery is in a high load state for a long time, the system will reduce the threshold of the equivalent conductivity of the separator, so that the early warning signal is issued in advance. Thus, the early warning signal is issued in advance.

[0078] Once the combined early warning condition is triggered, the system will issue an alarm to the user, suggesting that the use of the battery be stopped immediately; automatically record the relevant data triggering the early warning, including abnormal parameters and working condition information at the time of triggering; use the recorded data for subsequent optimization and improvement of the early warning algorithm.

[0079] This design can more accurately judge the thermal runaway risk of the battery by comprehensively analyzing the synergistic relationship between multiple parameters, and has good adaptability and scalability.

[0080] Those skilled in the art can understand that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood by those skilled in the art to which the present application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have meanings consistent with those in the context of the prior art, and unless otherwise defined, should not be interpreted to have idealized or overly formal meanings.

[0081] The meaning of "and / or" described in the present application means that each single existence or both existences are included.

[0082] Based on the above ideal embodiments according to the present application, through the above description, relevant personnel can make various changes and modifications without deviating from the technical idea of the present application. The technical scope of the present application is not limited to the contents of the specification, and the technical scope must be determined according to the scope of the claims.

Claims

1. A method for thermal runaway early warning of lithium ion battery based on multi-parameter monitoring, characterized in that, The method comprises the following steps: Step 1. Collecting voltage V(t), current I(t) and temperature T(t) data of the battery monomer, and monitoring the abnormal values and / or growth rates thereof to generate first type of early warning characteristic parameters; Pre-set voltage, current and temperature growth rate thresholds, when any of the real-time growth rates of voltage, current or temperature exceeds the pre-set growth rate threshold of the corresponding voltage, current or temperature, it is determined that the growth rate is abnormal; Step 2. Statistically analyzing the battery monomer voltage data to detect mean deviation and / or information entropy deviation to generate second type of early warning characteristic parameters; Based on statistical analysis of historical data, determining the determination standard values of mean deviation and information entropy deviation, when any of the measured values of mean deviation or information entropy deviation is greater than the determination standard value of the corresponding mean deviation or information entropy deviation, the second type of early warning characteristic parameters is generated; Step 3. Based on the battery electrochemical model, the particle swarm optimization method is used to identify the electrochemical parameters of the battery monomer, and the third type of early warning characteristic parameters is generated according to the abnormal electrochemical parameters obtained by identification; Lithium ion concentration C Li+ and separator equivalent conductivity σ sep The lithium ion concentration C Li+ and separator equivalent conductivity σ sep changes directly reflect the potential failure characteristics of short circuit, separator failure within the battery, The specific process of the electrochemical parameter identification is as follows: combining an electrochemical model and a particle swarm optimization algorithm, performing real-time monitoring, collecting voltage and current external signals in the battery operation process, dynamically inverting changes of C Li+ and sigma sep ; making the simulation voltage fit the collected voltage, that is, the loss function of the parameter identification, calculating the difference between the simulation voltage and the measured voltage, the difference between the simulation voltage and the measured voltage being minimum, the value of the target function being minimum, so that the electrochemical parameter value in the electrochemical model is obtained; The particle swarm optimization algorithm takes the minimization of the root mean square error between the simulated voltage and the measured voltage of the lithium-ion battery as the objective function, iteratively optimizes the model parameters, and thus obtains high-precision real-time C Li+ and σ sep electrochemical parameters, wherein V sim (i) is the simulated voltage at the i-th time point, V meas (i) is the measured voltage at the i-th time point, N is the number of data points; In the combination warning mechanism of the third type of electrochemical parameters, based on the monitored lithium ion concentration C Li+ and the equivalent conductivity of the separator σ se electrochemical parameters, set the combination warning boundary, when the lithium ion concentration C Li+ and the equivalent conductivity of the separator σ sep electrochemical parameter combination meets a certain condition, trigger the thermal runaway warning, the specific condition is: wherein is a threshold value of the equivalent conductivity of the separator dynamically adjusted according to the lithium ion concentration, and the formula is as follows: where σ0is the base separator equivalent conductivity threshold, k is an adjustment factor to represent the effect of lithium ion concentration on the separator conductivity, C ref is the reference lithium ion concentration; Step 4. When any type of early warning characteristic parameters in steps 1-3 is abnormal, the early warning is started.

2. The multi-parameter monitoring based thermal runaway early warning method for lithium ion battery according to claim 1, characterized in that, The voltage V(t), current I(t) and temperature T(t) data of the battery monomer collected in step 1 are pre-processed before abnormal value and / or growth rate operation, and the specific process of pre-processing is as follows: 1.1: Data cleaning is performed to remove noise and abnormal values of collected data, so as to ensure the quality and accuracy of data and remove obviously incorrect data beyond the reasonable range; 1.2: Data segmentation is performed to segment continuous time series data, and the segmentation method is based on time window or event triggering to ensure that each segment of data has similar characteristics.

3. The multi-parameter monitoring based thermal runaway early warning method for lithium-ion battery according to claim 2, characterized in that, The abnormal values of voltage V(t), current I(t) and temperature T(t) of the battery monomer in step 1 are determined by comparing the measured values with pre-set thresholds, and the measured values of voltage V(t), current I(t) and temperature T(t) exceeding the pre-set thresholds are determined as abnormal values.

4. The multi-parameter monitoring based thermal runaway early warning method for lithium ion battery according to claim 2, characterized in that, The calculation formula of the growth rate in step 1 is as follows: where y represents the type of collected data, including voltage, current, or temperature, σ y is the real-time growth rate of data y, y(i) is the real-time value of the data type y at the i th moment, and y(i+1) is the real-time value of the data type y at the i+1 th moment.

5. The multi-parameter monitoring based thermal runaway early warning method for lithium-ion battery according to claim 1, wherein, The calculation formula of the mean deviation in step 2 is as follows: where V i is the battery cell voltage at the ith time point, is the mean of all voltages, and N is the number of data points. The calculation formula of the information entropy is as follows: where P(V j ) is the probability distribution of the voltage value of the jth class, and k is the number of classes of voltage values.

6. The multi-parameter monitoring based thermal runaway early warning method for lithium ion battery according to claim 1, wherein, The acquisition process of the simulation voltage is as follows: the voltage data of the battery monomer is collected during the operation of the battery, and the simulation voltage of the battery is obtained by simulating the electrochemical model.

7. The multi-parameter monitoring based thermal runaway early warning method for lithium-ion battery according to claim 1, wherein, In order to adapt to the high load or low temperature operation environment, the combined early warning boundary is dynamically adjusted combined with real-time data and historical data, and when the battery is in a high load state for a long time, the Therefore, the early warning signal is sent in advance, once the combined early warning condition is triggered, the system will alarm the user to immediately stop the use of the battery; the related data triggering the early warning, including the abnormal parameters and the working condition information at the time of triggering, are automatically recorded.

Images