Power battery thermal runaway early warning method and device, storage medium and electronic equipment

Abstract

The invention relates to the technical field of power battery safety monitoring, in particular to a power battery thermal runaway early warning method and device, a storage medium and electronic equipment. The method comprises the following steps: training based on historical sampling data of single batteries of power batteries of a normal vehicle and a problem vehicle to obtain a time-varying linear relation between an alarm result and a feature vector; sampled data of a target power battery are obtained, a feature vector representing the thermal runaway risk of the power battery at the current moment is determined according to the sampled data, a corresponding alarm result predicted value is obtained through calculation according to a linear relation, and if the alarm result predicted value exceeds a first preset threshold value, it is judged that the thermal runaway risk exists, and if not, it is judged that the thermal runaway risk exists. Otherwise, the thermal runaway risk does not exist. The thermal runaway risk of the power battery can be accurately predicted in real time, and the false alarm condition is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power battery safety monitoring, and particularly relates to a method, device, storage medium and electronic device for predicting thermal runaway of a power battery. Background Art

[0002] At present, with the booming development of electric vehicles and energy storage systems, as a core component among them, the safety and reliability of power batteries are of crucial importance. Developing a precise and effective algorithm for warning of power battery thermal runaway has become the key to ensuring the stable operation of the battery system.

[0003] By virtue of the real-time monitoring of key parameters such as battery temperature, voltage, and current, the thermal runaway warning algorithm can sensitively capture abnormal conditions during battery operation and promptly send out alarm signals. This alarm will quickly prompt users or the system to take effective measures such as power-off, cooling, and emergency stop, thereby winning precious time for preventing accidents and reducing accident risks.

[0004] In the existing related technologies, it is proposed that under the charging state, when the voltage change rate of at least one battery cell within a unit time is negative and the difference in the voltage change rate within a unit time from other battery cells is greater than a first threshold, it can be determined that the battery is in a thermal runaway state. However, such technologies have obvious defects. Judging thermal runaway only based on the difference in the voltage change rate of one or several battery cells within a unit time being greater than a specific threshold, this single judgment method leads to a high misreport rate and poor accuracy. Due to frequent misreports, subsequent human efforts have to be invested in eliminating misreports, which not only consumes time and energy but also seriously affects the timeliness of thermal runaway handling, resulting in the inability of the battery system to respond quickly in the face of potential dangers. Summary of the Invention

[0005] The object of the present invention is to provide a method, device, storage medium and electronic device for predicting thermal runaway of a power battery, which can predict the thermal runaway risk of a power battery in real time and accurately and reduce the occurrence of misreports.

[0006] To achieve the above object, the technical scheme adopted by the present invention is as follows:

[0007] In a first aspect, the present invention discloses a method for predicting thermal runaway of a power battery, including:

[0008] Training a linear relationship varying with time between the alarm result and the feature vector based on the historical sampling data of battery cells of the power batteries of normal vehicles and problem vehicles;

[0009] Obtain the sampling data of the target power battery, determine the characteristic vector representing the thermal runaway risk of the power battery at the current moment based on the sampling data, calculate the corresponding predicted value of the alarm result according to the linear relationship formula, and if the predicted value of the alarm result exceeds the first preset threshold, it is determined that the first battery cell has a thermal runaway risk, otherwise it is determined that the first battery cell does not have a thermal runaway risk.

[0010] Furthermore, the linear relationship formula between the alarm result and the characteristic vector changing with time trained based on the historical sampling data of the battery cells of normal vehicles and problem vehicles specifically includes:

[0011] Obtain the historical sampling data of the battery cells of normal vehicles and problem vehicles, where the historical sampling data of the battery cells includes the temperature data of the first battery cell with a fixed time length, and the voltage data of the second battery cell adjacent to the first battery cell;

[0012] Based on the historical sampling data, determine the characteristic vector representing the thermal runaway risk of the power battery at each moment within a fixed time length, and the alarm result corresponding to this characteristic vector;

[0013] Fit to obtain the linear relationship formula between the alarm result and the characteristic vector changing with time.

[0014] Furthermore, based on the least squares method, fit to obtain the linear relationship formula between the alarm result and the characteristic vector changing with time.

[0015] Furthermore, it also includes: after fitting to obtain the linear relationship formula between the alarm result and the characteristic vector changing with time, regularly update the historical sampling data and refit.

[0016] Furthermore, the characteristic vector is (X1, X2), where X1 is the temperature characteristic vector of the first battery cell, and X2 is the voltage characteristic vector of the second battery cell; the alarm result Y is a scalar value.

[0017] Furthermore, the temperature characteristic vector X1 of the first battery cell is [ΔT, dT, Var(T)];

[0018] Where ΔT = T i -med(T), T i is the temperature of the first battery cell at the current moment, med(T) is the median temperature of all battery cells in the power battery at the current moment; dT is the temperature change rate of the first battery cell compared to the previous moment; Var(T) is the temperature variance of the first battery cell within a fixed time length;

[0019] The voltage characteristic vector X2 of the second battery cell is [ΔV, dV, Var(V)],

[0020] is the voltage of the j-th second battery cell at the current moment, V mid is the median voltage of all second battery cells at the current moment;

[0021] r(V) i is the voltage change rate of the second battery cells at the current moment compared to the previous moment, r[med(V)] i is the change rate of the median voltage of all second battery cells at the current moment compared to the previous moment;

[0022] Var(V) is the voltage variance of the second battery cells within a fixed time length.

[0023] Furthermore, it further includes: after calculating the predicted value of the alarm result, based on map the predicted value of the alarm result to the interval (0, 1), and compare the processing result Z with the second preset threshold. If the processing result Z exceeds the second preset threshold, it is determined that the first battery cell has a thermal runaway risk; otherwise, it is determined that the first battery cell does not have a thermal runaway risk.

[0024] In a second aspect, the present invention discloses a power battery thermal runaway warning device, which includes:

[0025] a training module for training a linear relationship that changes over time between the alarm result and the feature vector based on the historical sampling data of the battery cells of normal vehicles and problem vehicles;

[0026] a prediction module for obtaining the sampling data of the target power battery, determining the feature vector representing the thermal runaway risk of the power battery at the current moment according to the sampling data, calculating the corresponding predicted value of the alarm result according to the linear relationship. If the predicted value of the alarm result exceeds the first preset threshold, it is determined that the first battery cell has a thermal runaway risk; otherwise, it is determined that the first battery cell does not have a thermal runaway risk.

[0027] In a third aspect, the present invention discloses a storage medium on which a computer program is stored. When the computer program is run by the computer, it executes the above-mentioned power battery thermal runaway warning method.

[0028] In a fourth aspect, the present invention discloses an electronic device, which includes a memory and a processor, and the memory is connected to the processor; the memory is used for storing programs; the processor is used for calling the programs stored in the memory to execute the above-mentioned power battery thermal runaway warning method.

[0029] The present invention has the following unexpected beneficial effects:

[0030] 1. The battery thermal runaway warning method of the present invention comprehensively considers the historical sampling data of battery cells in normal vehicles and problem vehicles, which can more comprehensively reflect the battery state. By determining the feature vectors at multiple moments and comprehensively considering various factors, the judgment of the thermal runaway risk is more accurate, reducing false alarms. Furthermore, it greatly reduces the manual intervention cost, improves the efficiency and timeliness of thermal runaway handling, and ensures the safe and stable operation of the battery system.

[0031] 2. By obtaining the sampling data of the target power battery and calculating the predicted value of the alarm result according to the fitted linear relationship, the present invention can monitor the thermal runaway risk of the battery of the target vehicle in real time. As time goes by and the battery state changes, the data is continuously updated to dynamically evaluate the thermal runaway risk, timely discover potential problems, and win time for accident prevention. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0033] Figure 1 FIG. shows a schematic flow chart of the battery thermal runaway warning method provided by the embodiment of the present invention.

[0034] Figure 2 FIG. shows a schematic flow chart of the training process of the linear relationship provided by the embodiment of the present invention.

[0035] Figure 3 FIG. shows a schematic diagram of the positional relationship between the first battery cell and the second battery cell according to the embodiment of the present invention.

[0036] Figure 4 FIG. shows a schematic flow chart of a preferred embodiment of the battery thermal runaway warning method provided by the embodiment of the present invention.

[0037] Figure 5 FIG. shows a schematic structural diagram of the battery thermal runaway warning device provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] The following will describe the embodiments of the present invention with reference to the drawings and preferred embodiments. Those skilled in the art can easily understand the other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for explaining the present invention, rather than for limiting the protection scope of the present invention.

[0039] It should be noted that the illustrations provided in the following embodiments only schematically illustrate the basic concept of the present invention. The components shown in the drawings only show the components related to the present invention, rather than being drawn according to the number, shape and size of the components in actual implementation. The types, quantities and proportions of the components in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0040] In one embodiment, referring to Figure 1 as shown, the present invention discloses a method for warning of thermal runaway of a power battery, including the following steps:

[0041] S1. Based on the historical sampling data of battery cells of normal vehicles and problem vehicles, a linear relationship between the alarm result and the feature vector changing with time is trained;

[0042] S2. Obtain the sampling data of the target power battery, determine the feature vector representing the thermal runaway risk of the power battery at the current moment according to the sampling data, calculate the corresponding predicted alarm result value according to the linear relationship, and if the predicted alarm result value exceeds the first preset threshold, it is determined that the first battery cell has a thermal runaway risk, otherwise it is determined that the first battery cell does not have a thermal runaway risk.

[0043] By obtaining the sampling data of the target power battery, calculating the predicted alarm result value and comparing it with the preset threshold to determine the risk, the real-time monitoring and warning of the thermal runaway risk of the battery of the target vehicle are realized. Once the predicted value exceeds the threshold, an alarm can be issued in time to prompt the user or the system to take measures, effectively preventing the serious consequences caused by thermal runaway, ensuring the safety of personnel and property, and improving the safety and reliability of the battery system.

[0044] The method for warning of thermal runaway of the power battery according to the present invention comprehensively considers the historical sampling data of battery cells of normal vehicles and problem vehicles, covering temperature data and adjacent cell voltage data, and can more comprehensively reflect the battery state. By determining the feature vectors at multiple moments and comprehensively considering various factors, the judgment of the thermal runaway risk is more accurate, and the occurrence of false alarms is reduced. Furthermore, the cost of manual intervention is greatly reduced, the efficiency and timeliness of thermal runaway treatment are improved, and the safe and stable operation of the battery system is ensured.

[0045] The present invention can perform real-time monitoring of the thermal runaway risk of the battery of the target vehicle by obtaining the sampling data of the target power battery and calculating the predicted alarm result value according to the fitted linear relationship. As time goes by and the battery state changes, the data is continuously updated, the thermal runaway risk is dynamically evaluated, potential problems are discovered in time, and time is won for preventing accidents.

[0046] As a preferred embodiment of the present invention, referring to Figure 2As shown in the figure, the linear relationship between the alarm result and the feature vector that changes over time, which is trained based on the historical sampling data of battery cells of normal vehicles and problem vehicles, specifically includes the following steps:

[0047] S11. Obtain the historical sampling data of battery cells of normal vehicles and problem vehicles. The historical sampling data of battery cells includes the temperature data of the first battery cell with a fixed time length, and the voltage data of the second battery cell adjacent to the first battery cell.

[0048] The historical sampling data covers the battery state information under different working conditions, provides rich samples for subsequent fitting analysis, helps to explore the characteristic change rules of the battery before and after thermal runaway, and improves the early warning accuracy. If abnormal change patterns of temperature and voltage can be found from a large amount of data, it provides a strong basis for judging the thermal runaway risk.

[0049] S12. Based on the historical sampling data, determine the feature vector representing the thermal runaway risk of the power battery at each moment within a fixed time length, and the alarm result corresponding to the feature vector.

[0050] Comprehensively consider various features of temperature and voltage. For example, the temperature feature vector includes the difference between the current temperature and the temperature median, the temperature change rate, the temperature variance, etc., which comprehensively reflects the temperature change trend and fluctuation; the voltage feature vector is the same. In this way, the battery state can be more accurately reflected, providing a reliable data basis for subsequent fitting and prediction, and improving the accuracy of the early warning model for judging the thermal runaway risk.

[0051] S13. Fit the linear relationship between the alarm result and the feature vector that changes over time.

[0052] Build a quantitative relationship between the battery state characteristics and the thermal runaway risk. With the help of this linear relationship, the predicted value of the alarm result can be calculated based on the real-time collected battery data, realizing the quantitative assessment of the thermal runaway risk, and making the early warning more scientific and accurate.

[0053] It should be noted that as shown in Figure 3 the figure, the judgment method of the second battery cell 20 adjacent to the first battery cell 10 includes: by checking the battery layout drawing or directly observing the battery pack, clarify the position of the first battery cell 10 in the battery pack box 30, and observe the battery cells around the first battery cell 10, and use the battery cell that is directly adjacent to the first battery cell 10 and has no other battery cells in between as the second battery cell 20.

[0054] Furthermore, the historical sampling data of battery cells of normal vehicles and problem vehicles obtained are respectively divided into a training set and a test set according to a preset ratio.

[0055] Specifically, the historical sampling data includes the historical sampling data of battery cells of power batteries of a normal vehicles and the historical sampling data of battery cells of power batteries of b problem vehicles, and the preset ratio is c. Then the training set includes a×c and b×c, and the test set is the remaining historical sampling data other than the training set.

[0056] Use the training set data to fit and train the linear relationship between the alarm result and the feature vector, so that the fitting relationship can learn the rules and features in the data. Then use the test set data to test the trained fitting relationship and observe the performance of the fitting relationship on the data that has not participated in the training. By comparing the predicted results of the test set with the actual situation, the performance indicators such as the accuracy and generalization ability of the model can be accurately evaluated. If the fitting relationship performs well on the test set, it indicates that the model can better adapt to different data samples and has reliability in predicting the risk of thermal runaway of the power battery; on the contrary, if the error of the fitting relationship on the test set is large, the fitting parameters of the fitting relationship need to be adjusted and optimized.

[0057] As a preferred embodiment of the present invention, a linear relationship between the alarm result and the feature vector changing with time is obtained by least squares fitting.

[0058] The core principle of the least squares method is to find the best function matching of the data by minimizing the sum of the squares of the errors, which can effectively handle the noise and errors existing in the historical sampling data, so that the fitted linear relationship fits the actual data distribution to the greatest extent. Based on a large amount of historical thermal runaway data and normal data, the least squares method can accurately determine the coefficients of each item in the linear relationship, so that the predicted value of the alarm result can more accurately reflect the true thermal runaway risk status of the power battery. Compared with other fitting methods, the least squares method can more reasonably allocate the influence weights of each feature vector on the alarm result, avoid prediction deviations caused by data fluctuations, and thus significantly improve the accuracy of thermal runaway early warning.

[0059] Moreover, the calculation process of the least squares method is relatively simple and efficient. When dealing with a large amount of historical sampling data, the fitting result can be quickly obtained. When new target power battery sampling data is obtained, the predicted value of the alarm result can be quickly calculated according to the fitted linear relationship, realizing real-time monitoring and rapid judgment of the thermal runaway risk. During the driving of an electric vehicle, the battery management system can quickly process the collected data and send out early warning signals in time to gain valuable time for ensuring the safety of the vehicle and personnel.

[0060] Meanwhile, as the battery usage time increases, the operating conditions change, and new data accumulates, it is necessary to continuously update the linear relationship between the alarm result and the feature vector to improve the early warning accuracy. The fitting method based on the least squares method is easy to operate. When new historical sampling data is added, only the least squares method needs to be recalculated to obtain new fitting parameters and achieve the dynamic update of the model. Regularly updating the historical sampling data and refitting can make the early warning model better adapt to the changes in the battery state and continuously maintain a high early warning performance.

[0061] Specifically, the steps for obtaining the linear relationship between the alarm result and the feature vector that changes with time based on the least squares method fitting are as follows:

[0062] S131, construct the initial relationship between the alarm result Y and the feature vector X: Y = ω·X + b.

[0063] S132, solve the objective function is the actual value of the alarm result at the i-th moment, and h is the number of time series data points within a fixed time length.

[0064] S1321, set the initial values ω0 and b0 of ω and b, and calculate the partial derivatives of the objective function with respect to ω and b respectively

[0065] S1322, set the value of the parameter η, and update the values of ω and b in the way of gradient descent, that is:

[0066]

[0067] S1323, substitute the initial values of ω and b and the corresponding updated values obtained by the gradient descent method into the objective function to calculate L, and take the ω and b corresponding to the smallest L as the solution of the objective function.

[0068] As a preferred embodiment of the present invention, the method for early warning of thermal runaway of the power battery further includes: after obtaining the linear relationship between the alarm result and the feature vector that changes with time by fitting, regularly update the historical sampling data and refit.

[0069] The performance of the power battery will change with the usage time, the number of charge and discharge cycles, and the change of the working environment. As the battery gradually ages, the internal chemical reactions and physical properties will be different, and the manifestation characteristics of the thermal runaway risk may also change accordingly. Regularly updating the historical sampling data can incorporate this newly emerging battery state information into the model, and the refitted linear relationship can more accurately reflect the relationship between the current thermal runaway risk of the battery and the feature vector.

[0070] As data accumulates, the old data may contain some outliers caused by measurement errors or special circumstances, and these outliers may affect the accuracy of the linear relationship. Regularly updating the historical sampling data can remove some old outliers and at the same time add more representative new data, making the model more stable and reliable. The process of refitting enables the model to better adapt to data changes, reducing false alarms and missed alarms caused by data fluctuations or outliers, thereby improving the stability and reliability of the entire thermal runaway warning system.

[0071] Further, refer to Figure 4 As shown, using the sampling data of the power battery that newly experiences thermal runaway as input, update the historical sampling data and refit again.

[0072] The sampling data of the power battery that newly experiences thermal runaway contains the latest characteristics of the current battery thermal runaway. With the development of battery technology and the change of the use environment, the manifestation forms of thermal runaway may be different. These new data can show the latest characteristics related to thermal runaway, such as unique temperature and voltage change patterns caused by new types of faults, etc. Incorporating them into the historical sampling data and refitting, the warning model can accurately capture these new characteristics, so that subsequent risk predictions are more in line with the actual situation, greatly improving the accuracy of the warning.

[0073] Using fixed historical sampling data for model training in the long term is prone to overfitting, resulting in the model having a poor adaptability to new data. Introducing new thermal runaway sampling data can enrich the data diversity and provide a wider range of samples for model training. The refitting process is based on more comprehensive data, which helps to avoid the model relying too much on certain specific patterns in the old data, enhances the generalization ability of the model, and enables it to maintain good prediction performance when facing various actual situations.

[0074] Each update of new thermal runaway data and refitting is an optimization and upgrade of the warning system. With the continuous addition of new data, the model's judgment of thermal runaway risk will be more accurate, and the situations of false alarms and missed alarms will be further reduced. This is crucial for ensuring the safe operation of power batteries, can provide users with more reliable thermal runaway warning services, and reduce the risk of safety accidents caused by thermal runaway.

[0075] As a preferred implementation manner of the present invention, the feature vector is (X1, X2), where X1 is the temperature feature vector of the first battery cell, and X2 is the voltage feature vector of the second battery; the alarm result Y is a scalar value.

[0076] Temperature and voltage are important indicators reflecting the thermal runaway risk of power batteries. The combination of the two can more accurately capture the change characteristics before battery thermal runaway. The temperature change of the first battery cell may be caused by abnormal internal chemical reactions, while the voltage change of the adjacent second battery cell may reflect the change in electrical performance inside the battery pack. Considering both comprehensively is more accurate than relying solely on a single indicator to judge the thermal runaway risk.

[0077] Define the alarm result Y as a scalar value to make the evaluation result concise and intuitive. In practical applications, the system or operator can quickly judge whether there is a thermal runaway risk in the first battery cell based on this single scalar value, without having to process complex multi-dimensional evaluation results. When the alarm result Y exceeds the preset threshold, it directly indicates the existence of a risk. This simple and clear judgment method helps to make quick decisions and improve the timeliness of thermal runaway warning.

[0078] The clear feature vector and scalar value alarm result are convenient for integration with other systems. In electric vehicles or energy storage systems, this thermal runaway warning method can be seamlessly docked with battery management systems, vehicle control systems, etc. The battery management system can monitor the battery state in real time based on the feature vector X and the alarm result Y, and the vehicle control system can take corresponding measures according to the warning information, such as restricting power output, starting the cooling system, etc., enhancing the safety and reliability of the entire system and improving the system practicality.

[0079] As a preferred implementation manner of the present invention, the temperature feature vector X1 of the first battery cell is [ΔT, dT, Var(T)].

[0080] In the formula, ΔT = T i -med(T), where T i is the temperature of the first battery cell at the current moment, and med(T) is the median temperature of all battery cells in the power battery at the current moment; the introduction of ΔT can intuitively understand the relative high and low situation of the temperature of the first battery cell in the entire battery pack. If ΔT is large, it means that the temperature of this cell deviates from the overall average level and may be abnormal, providing an important clue for judging the thermal runaway risk. For example, when the temperatures of other battery cells are relatively stable and the ΔT of a certain cell continues to increase, the thermal runaway risk needs to be vigilant.

[0081] dT is the temperature change rate of the first battery cell at the current moment compared with the previous moment. Specifically, T i-1is the temperature of the first battery cell at the previous moment, and Δt is the time difference between the current moment and the previous moment. dT reflects the rate of change of the battery cell temperature. A rapid temperature rise is often an early sign of thermal runaway. By monitoring this indicator, an abnormal rapid change in temperature can be detected in a timely manner, and an early warning can be issued. In the initial stage of thermal runaway, the temperature may rise sharply, and at this time, the temperature change rate will increase significantly. This indicator can sensitively capture this change, which helps to take measures in the early stage of the development of thermal runaway.

[0082] Var(T) is the temperature variance of the first battery cell within a fixed time length, which is used to measure the temperature fluctuation. A stable temperature fluctuation range indicates that the battery is in a normal working state, while a large variance means that the temperature fluctuates violently, which may indicate internal faults or unstable factors, increasing the risk of thermal runaway.

[0083] In the present invention, the temperature characteristics [ΔT, dT, Var(T)] are comprehensively incorporated into the temperature feature vector X1, which comprehensively covers information such as the absolute value, relative position, change speed, and fluctuation stability of the temperature, and can more accurately describe the thermal state of the first battery cell. Compared with single temperature measurement, this multi-dimensional feature description can more accurately identify the thermal runaway risk, reduce false alarms and missed alarms, and provide more reliable guarantee for the safe operation of the power battery.

[0084] As a preferred embodiment of the present invention, the voltage feature vector X2 of the second battery cell is [ΔV, dV, Var(V)].

[0085] is the voltage of the j-th second battery cell at the current moment, V mid is the median voltage of all second battery cells at the current moment. The introduction of ΔV clearly shows the relative high and low of the voltage of each second battery cell in the battery pack. If the ΔV of a certain second battery cell deviates from the normal range, it means that the voltage of this second battery cell is abnormal, and this abnormality may be an early signal of the thermal runaway risk. Under normal circumstances, the voltages of the individual cells in the battery pack should be relatively balanced. If the ΔV of a certain second battery cell is too large or too small, it indicates that this cell may have internal short circuit, open circuit or other faults, increasing the possibility of thermal runaway.

[0086] r(V) i is the voltage change rate of the second battery cell at the current moment compared with the previous moment, r[med(V)] iis the change rate of the median voltage of all second battery cells at the current moment compared to the previous moment; the introduction of dV can intuitively reflect the change of voltage over time. A rapid increase or decrease in voltage may be an indication of abnormal chemical reactions inside the battery, which is closely related to the risk of thermal runaway. Monitoring the voltage change rate can promptly capture these abnormal change trends and early warning of the thermal runaway risk. During the battery charging process, if the voltage change rate of a certain second battery cell suddenly accelerates and exceeds the normal range, it may indicate that the second battery cell is about to experience thermal runaway.

[0087] Wherein, V i is the voltage of the second battery cell at the current moment, V i-1 is the voltage of the second battery cell at the previous moment, and Δt is the time difference between the current moment and the previous moment.

[0088] med(V) i is the median voltage of all second battery cells adjacent to the first battery cell at the current moment, med(V) i-1 is the median voltage of all second battery cells adjacent to the first battery cell at the previous moment, and Δt is the time difference between the current moment and the previous moment.

[0089] Var(V) is the voltage variance of the second battery cell within a fixed time length. A stable voltage fluctuation range is an important indicator of the normal operation of the battery, while an increase in the voltage variance indicates an intensification of voltage fluctuations, and there may be unstable factors inside the battery, which will increase the risk of thermal runaway. During the battery cycling process, as the battery ages, the voltage variance may gradually increase. By monitoring the voltage variance, the aging trend of the battery and the potential thermal runaway risk can be detected in a timely manner.

[0090] As a preferred embodiment of the present invention, the method for warning thermal runaway of the power battery further includes: after calculating the predicted value of the alarm result, based on map the predicted value of the alarm result to the interval (0, 1), and compare the processing result Z with the second preset threshold. If the processing result Z exceeds the second preset threshold, it is determined that the first battery cell has a risk of thermal runaway, otherwise it is determined that the first battery cell does not have a risk of thermal runaway.

[0091] Map the predicted value of the alarm result to the interval (0, 1), so that predicted values from different sources and of different magnitudes have a unified measurement standard. In practical applications, the predicted value of the alarm result may have different numerical ranges due to various factors (such as different battery types, differences in the accuracy of measurement devices, etc.). By mapping to the interval (0, 1), regardless of the original predicted value, the degree of thermal runaway risk can be represented on a standardized scale, facilitating subsequent comparison and judgment.

[0092] The numerical value in the interval (0, 1) can intuitively reflect the level of thermal runaway risk. The closer the numerical value is to 1, the higher the thermal runaway risk; the closer the numerical value is to 0, the lower the thermal runaway risk. This enables operators or systems to more intuitively understand and evaluate the thermal runaway risk status of the battery and make more reasonable decisions. For example, when the processing result Z is 0.9, it clearly indicates a very high thermal runaway risk and immediate measures need to be taken; while when Z is 0.1, it means the thermal runaway risk is very low and continuous observation can be carried out.

[0093] Compare the processing result Z with the second preset threshold to determine the risk, making the setting of the warning threshold more flexible. The size of the second preset threshold can be flexibly adjusted according to factors such as different application scenarios, the importance of the battery, and the risk tolerance.

[0094] This method combines the previous linear relationship calculation and mapping processing, further improving the accuracy and reliability of the warning. The linear relationship can preliminarily predict the thermal runaway risk based on the eigenvector, and the process of mapping to the interval (0, 1) and comparing with the threshold further refines and standardizes the prediction result, reducing misjudgments caused by the uncertainty and fluctuations of the original predicted value. By reasonably setting the second preset threshold, while ensuring timely detection of the thermal runaway risk, the occurrence of false alarms can be minimized as much as possible, improving the overall performance of the warning system.

[0095] In an embodiment, the present invention also discloses a thermal runaway warning device for a power battery. Refer to Figure 5 As shown, the warning device 50 includes a training module 51 and a prediction module 52.

[0096] The training module 51 is used to train a linear relationship that changes with time between the alarm result and the eigenvector based on the historical sampling data of the battery cells of the power batteries of normal vehicles and problem vehicles.

[0097] The prediction module 52 is configured to obtain the sampled data of the target power battery, determine the feature vector representing the thermal runaway risk of the power battery at the current moment based on the sampled data, calculate the corresponding predicted value of the alarm result according to the linear relationship formula, and if the predicted value of the alarm result exceeds the first preset threshold, it is determined that the first battery cell has a thermal runaway risk; otherwise, it is determined that the first battery cell does not have a thermal runaway risk.

[0098] In one embodiment, the present invention also discloses a storage medium, on which a computer program is stored. When the computer program is run by a computer, it executes the above-mentioned power battery thermal runaway warning method.

[0099] In one embodiment, the present invention also discloses an electronic device, including a memory and a processor, the memory is connected to the processor; the memory is used for storing programs; the processor is used for calling the program stored in the memory to execute the above-mentioned power battery thermal runaway warning method.

[0100] Wherein, the processor is connected to the memory through a bus, and the memory stores program codes. When the program codes are executed by the processor, various steps in the above-mentioned power battery thermal runaway warning method are enabled for the processor.

[0101] The processor may be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and can execute the power battery thermal runaway warning method, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the power battery thermal runaway warning method disclosed in combination with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0102] As a non-volatile storage medium, the memory can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The memory may include at least one type of storage medium, for example, it may include flash memory, hard disk, multimedia card, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), magnetic memory, magnetic disk, optical disc, and so on. The memory 12 is any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory in the embodiments of the present application may also be a circuit or any other system capable of implementing a storage function for storing program instructions and / or data.

[0103] The above embodiments are only preferred embodiments given to fully illustrate the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or transformations made by those skilled in the art on the basis of the present invention are within the protection scope of the present invention.

Claims

1. A method for warning of thermal runaway of a power battery, characterized in that, Including: Training a linear relationship that changes over time between the alarm result and the feature vector based on the historical sampling data of the battery cells of normal vehicles and problem vehicles; Obtaining the sampling data of the target power battery, determining the feature vector representing the thermal runaway risk of the power battery at the current moment based on the sampling data, calculating the corresponding predicted alarm result according to the linear relationship, and if the predicted alarm result exceeds the first preset threshold, determining that the first battery cell has a thermal runaway risk, otherwise determining that the first battery cell does not have a thermal runaway risk.

2. The method for warning of thermal runaway of a power battery according to claim 1, wherein Specifically, training a linear relationship that changes over time between the alarm result and the feature vector based on the historical sampling data of the battery cells of normal vehicles and problem vehicles includes: Obtaining the historical sampling data of the battery cells of normal vehicles and problem vehicles, where the historical sampling data of the battery cells includes the temperature data of the first battery cell with a fixed time length and the voltage data of the second battery cell adjacent to the first battery cell; Based on the historical sampling data, determining the feature vector representing the thermal runaway risk of the power battery at each moment within a fixed time length and the alarm result corresponding to the feature vector; Fitting to obtain a linear relationship that changes over time between the alarm result and the feature vector.

3. The method for warning of thermal runaway of a power battery according to claim 2, characterized in that: Fitting to obtain a linear relationship that changes over time between the alarm result and the feature vector based on the least squares method.

4. The method for warning of thermal runaway of a power battery according to claim 2, wherein Also including: After fitting to obtain a linear relationship that changes over time between the alarm result and the feature vector, regularly update the historical sampling data and refit.

5. The method for warning of thermal runaway of a power battery according to claim 1, wherein: The feature vector is (X1, X2), where X1 is the temperature feature vector of the first battery cell and X2 is the voltage feature vector of the second battery; The alarm result Y is a scalar value.

6. The method for warning of thermal runaway of a power battery according to claim 5, wherein: The temperature feature vector X1 of the first battery cell is [ΔT, dT, Var(T)]; where ΔT = T i -med(T), T i is the temperature of the first battery cell at the current moment, med(T) is the median temperature of all battery cells in the power battery at the current moment; dT is the temperature change rate of the first battery cell at the current moment compared to the previous moment; Var(T) is the temperature variance of the first battery cell within a fixed time length; The voltage feature vector X2 of the second battery cell is [ΔV, dV, Var(V)], ΔV = V i j -V mid ,V i j is the voltage of the j-th second battery cell at the current moment, V mid is the median voltage of all second battery cells at the current moment; r(V) i is the voltage change rate of the second battery cell at the current moment compared to the previous moment, r[med(V)] i is the voltage median change rate of all second battery cells at the current moment compared to the previous moment; Var(V) is the voltage variance of the second battery cell within a fixed time length.

7. The method for warning of thermal runaway of a power battery according to claim 1, wherein, Also including: After obtaining the predicted value of the alarm result, based on map the predicted value of the alarm result to the interval (0, 1), and compare the processing result Z with the second preset threshold. If the processing result Z exceeds the second preset threshold, it is determined that the first battery cell has a risk of thermal runaway; otherwise, it is determined that the first battery cell does not have a risk of thermal runaway.

8. A thermal runaway warning device for a power battery, characterized in that, Including: A training module for training a linear relationship that changes over time between the alarm result and the feature vector based on the historical sampling data of the battery cells of normal vehicles and problem vehicles; A prediction module for obtaining the sampling data of the target power battery, determining the feature vector representing the thermal runaway risk of the power battery at the current moment based on the sampling data, calculating the corresponding predicted alarm result according to the linear relationship, and if the predicted alarm result exceeds the first preset threshold, determining that the first battery cell has a thermal runaway risk, otherwise determining that the first battery cell does not have a thermal runaway risk.

9. A storage medium, characterized in that: A computer program is stored thereon, and when the computer program is run by a computer, it executes the power battery thermal runaway warning method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: Including a memory and a processor, the memory is connected to the processor; The memory is used for storing programs; The processor is used for calling the program stored in the memory to execute the power battery thermal runaway warning method according to any one of claims 1 to 7.

Images