A battery overcharge failure warning method and system based on electrochemical impedance

Abstract

The present application provides a battery overcharge failure warning method and system based on electrochemical impedance, which belongs to the field of battery management. The method includes: measuring the impedance of the battery at different temperatures and charge states, and obtaining the characteristic frequency of the battery overcharge failure warning; collecting characteristic impedance data to calculate the impedance curvature sequence; calculating the neighborhood radius and the minimum number of points based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, and constructing a battery overcharge failure warning model based on a clustering algorithm; measuring the characteristic impedance data of the battery at the characteristic frequency, calculating the impedance curvature value, and inputting the impedance curvature value into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and sending a warning signal when the impedance curvature value is an abnormal category and a positive value. The present application can reflect the safety status of the battery from the abnormal changes inside the battery, has a fast response speed, and solves the problem that the traditional threshold method is difficult to determine the alarm threshold.

Description

Technical Field

[0001] The present application belongs to the field of battery management, and more specifically, to a battery overcharge failure warning method and system based on electrochemical impedance. Background Art

[0002] New energy storage has been vigorously developed in recent years. Lithium-ion batteries have dominated the field of energy storage power stations and electric vehicles due to their high energy density, long life, no memory effect and environmental protection. However, the safety of lithium-ion batteries remains a pain point in the industry. Inconsistency in the performance of the battery cells themselves, inconsistency in the operating conditions and failure of the battery management system may all cause battery overcharging. Severe overcharging may cause thermal runaway or failure of the battery, which is extremely harmful. Therefore, timely and reliable early warning of battery overcharge failure is of great significance, which can improve the energy supply stability and service life of the energy storage system.

[0003] At present, most battery management systems use terminal voltage or state of charge to prevent and control overcharge failure. However, the terminal voltage of the battery is greatly affected by the charge and discharge current, and the state of charge is difficult to calculate accurately. Therefore, the actual application effect of these two methods is general. Some advanced battery management systems will warn of battery overcharge abuse based on temperature, sound waves or gas signals. The temperature-based overcharge warning method is usually generally fast; the sound wave-based overcharge warning method requires additional sound wave transceiver equipment, which is costly; the gas-based overcharge warning method usually takes effect only after the battery ruptures, and the speed is poor. Secondly, these warning methods are usually judged by threshold methods. The determination of the threshold value itself is relatively difficult. The algorithm is prone to miss judgment when the threshold value is large, and the algorithm is prone to misjudgment when the threshold value is small. Finally, most of these warning methods are designed for thermal runaway, and their effectiveness for overcharge failure remains to be verified. Summary of the invention

[0004] In view of the defects of the prior art, the purpose of this application is to provide a battery overcharge failure warning method and system based on electrochemical impedance, aiming to solve the problems of difficult to determine thresholds and slow warning response speed in the existing battery overcharge failure warning methods. If the warning threshold of the traditional method is too high, it is easy for the algorithm to miss the judgment, and if the warning threshold is too low, it is easy for the algorithm to make a wrong judgment.

[0005] To achieve the above objectives, in a first aspect, the present application provides a battery overcharge failure warning method based on electrochemical impedance, specifically:

[0006] Measure the characteristic impedance data of the battery at the characteristic frequency and calculate the impedance curvature value;

[0007] The impedance curvature value is input into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and a warning signal is issued when the impedance curvature value is an abnormal category and a positive value;

[0008] The method for constructing the battery overcharge failure warning model based on the clustering algorithm specifically includes the following steps:

[0009] S1: Measure the impedance of the battery at different temperatures and charge states to obtain the characteristic frequency of the battery overcharge failure warning;

[0010] S2: Measure characteristic impedance data at characteristic frequencies during battery charging and calculate impedance curvature sequence;

[0011] S3: Based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, the neighborhood radius and the minimum number of points are calculated, and a battery overcharge failure warning model based on the clustering algorithm is constructed.

[0012] Further preferably, the method for obtaining the characteristic frequency of the battery overcharge failure warning in S1 specifically includes the following steps:

[0013] S1.1: Calculate the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under the preset state of charge;

[0014] S1.2: Calculate the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency at a preset temperature;

[0015] S1.3: Calculate the comprehensive evaluation value at each frequency based on the comprehensive evaluation expression constructed based on the Pearson correlation coefficient obtained in S1.1 and the Pearson correlation coefficient obtained in S1.2;

[0016] S1.4: If the frequency corresponding to the maximum comprehensive evaluation value is higher than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value is used as the characteristic frequency; otherwise, the lower limit of the characteristic frequency is used as the characteristic frequency;

[0017] Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ; Among them, R T is the Pearson correlation coefficient between the real part of the battery impedance and temperature at each frequency under the preset state of charge; R C is the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; R is the comprehensive evaluation value.

[0018] Further preferably, the calculation formula of impedance curvature is:

[0019]

[0020] Among them, Kz is the impedance curvature; z″ is the second-order differential of the characteristic impedance curve at a certain point; z′ is the first-order differential of the characteristic impedance curve at a certain point; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm.

[0021] Further preferably, step S3 is specifically:

[0022] Calculate the minimum number of points based on the charging current, the rated capacity of the battery, and the sampling rate;

[0023] Calculate the neighborhood radius of the clustering algorithm based on the minimum number of points and impedance curvature data;

[0024] A battery overcharge failure warning model based on clustering algorithm is constructed according to the neighborhood radius and the minimum number of points; wherein the input variable of the battery overcharge failure warning model is impedance curvature, and the output variable is normal category or abnormal category.

[0025] Further preferably, the minimum number of points is:

[0026]

[0027] Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f z is the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes;

[0028] The neighborhood radius is:

[0029] R D = n·m(OD(X))

[0030] Where n is a coefficient, m(OD(X)) means calculating the average Euclidean distance of X sample points; X is 2MP; and the sample points are impedance curvature sequences.

[0031] In a second aspect, the present application provides a battery overcharge failure warning system based on electrochemical impedance, including: a characteristic frequency acquisition module, an impedance curvature calculation module, a model construction module and an abnormality diagnosis module;

[0032] The characteristic frequency acquisition module is used to measure the impedance of the battery at different temperatures and charge states to obtain the characteristic frequency of the battery overcharge failure warning;

[0033] The impedance curvature calculation module is used to measure characteristic impedance data at a characteristic frequency during battery charging and calculate an impedance curvature sequence;

[0034] The model building module is used to calculate the neighborhood radius and the minimum number of points based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, and to build a battery overcharge failure warning model based on a clustering algorithm;

[0035] The abnormal diagnosis module is used to measure the characteristic impedance data of the battery at the characteristic frequency and calculate the impedance curvature value; the impedance curvature value is input into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and a warning signal is issued when the impedance curvature value is of an abnormal category and is a positive value.

[0036] Further preferably, the characteristic frequency acquisition module includes a temperature box, a battery charge and discharge tester, an electrochemical workstation and a data processing unit;

[0037] The temperature box is used to keep the battery in a preset temperature environment; the battery charge and discharge tester is used to charge and discharge the battery and change the battery's state of charge; the electrochemical workstation is used to measure the real part of the battery's impedance at different SOCs and temperatures at various frequencies;

[0038] The data processing unit is used to calculate the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under a preset state of charge; calculate the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under a preset temperature; calculate the comprehensive evaluation value at each frequency according to the comprehensive evaluation expression constructed based on the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under a preset state of charge and the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under a preset temperature; if the frequency corresponding to the maximum comprehensive evaluation value is higher than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value is used as the characteristic frequency, otherwise the lower limit of the characteristic frequency is used as the characteristic frequency;

[0039] Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ; R T is the Pearson correlation coefficient between the real part of the battery impedance and temperature at each frequency under the preset state of charge; R C is the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; R is the comprehensive evaluation value.

[0040] Further preferably, the calculation formula of impedance curvature is:

[0041]

[0042] Among them, K z is the impedance curvature; z″ is the second-order differential of the characteristic impedance curve at a certain point; z′ is the first-order differential of the characteristic impedance curve at a certain point; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm.

[0043] Further preferably, the model building module includes a point number calculation unit, a radius calculation unit and a model building unit;

[0044] The point number calculation unit is used to calculate the minimum number of points according to the charging current, the rated capacity of the battery and the sampling rate;

[0045] The radius calculation unit is used to calculate the neighborhood radius of the clustering algorithm based on the minimum number of points and impedance curvature data;

[0046] The model building unit is used to build a battery overcharge failure warning model based on a clustering algorithm according to the neighborhood radius and the minimum number of points; wherein the input variable of the battery overcharge failure warning model is impedance curvature, and the output variable is a normal category or an abnormal category.

[0047] Further preferably, the minimum number of points in the battery overcharge failure warning model construction module is:

[0048]

[0049] Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f z is the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes;

[0050] The neighborhood radius in the battery overcharge failure warning model construction module is:

[0051] R D = n·m(OD(X))

[0052] Where n is a coefficient, m(OD(X)) means calculating the average Euclidean distance of X sample points; X is 2MP; and the sample points are impedance curvature sequences.

[0053] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the prior art:

[0054] Compared with the existing security control method using raw impedance data, this application uses impedance curvature for security warning. A warning signal is issued only when the impedance curvature value is an abnormal category and the impedance curvature value is positive. It can filter out interference signals in the impedance data and reduce the misjudgment rate and missed judgment rate of the clustering algorithm.

[0055] Compared with the existing battery overcharge safety control methods based on voltage, temperature or gas, the present application provides a battery overcharge failure warning method based on electrochemical impedance, wherein the safety warning based on electrochemical impedance can reflect the safety status of the battery from the abnormal changes inside the battery, with higher reliability and better speed.

[0056] Compared with the existing threshold-based battery safety prevention and control methods, this application adopts an adaptive clustering algorithm to achieve safety warning. The algorithm can adjust the model parameters according to the battery's own data characteristics (Q is the rated capacity of the battery) and the current size (charging current size x). It has strong adaptability and solves the problem that the traditional threshold method is difficult to determine the alarm threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a schematic diagram of a battery overcharge failure warning method based on electrochemical impedance provided by the present application;

[0058] Figure 2 It is the experimental system architecture diagram provided by this application;

[0059] Figure 3 It is the clustering algorithm warning flow chart in step S4 provided in this application;

[0060] Figure 4 is the recorded battery terminal voltage curve provided by the present application;

[0061] Figure 5 is a recorded battery characteristic impedance curve provided in an embodiment of the present application;

[0062] Figure 6 is an impedance curvature diagram provided in an embodiment of the present application;

[0063] Figure 7 It is a clustering result diagram of the early warning algorithm provided in the embodiment of the present application;

[0064] Figure 8 is a warning signal diagram of the warning algorithm provided in the embodiment of the present application;

[0065] Fig. 9 It is a schematic diagram of a battery overcharge failure warning system based on electrochemical impedance provided in an embodiment of the present application. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0067] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0068] Example 1

[0069] like Figure 1 As shown, taking 18650 lithium iron phosphate battery as an example, the present application provides a battery overcharge failure warning method based on electrochemical impedance, comprising the following steps:

[0070] S100: Characteristic impedance screening: measure the impedance of the battery at different temperatures and states of charge (SOC), and select characteristic frequencies suitable for battery overcharge failure warning;

[0071] S100 specifically includes the following steps:

[0072] Measure the electrochemical impedance spectroscopy data of the battery at different temperatures and different states of charge (SOC);

[0073] Calculate the Pearson correlation coefficient R between the real part of the battery impedance and temperature at each frequency when the SOC is 60% T ; It should be pointed out that the SOC of 60% here is the SOC value of the lithium battery when it is in normal operation. In practical applications, the SOC is not limited to 60%;

[0074] Calculate the Pearson correlation coefficient R between the real part of the battery impedance and the state of charge (SOC) at each frequency when the temperature is 25°C C ; It should be pointed out that the value 25℃ here is the normal working temperature of lithium battery; in actual application, the temperature is not limited to 25℃;

[0075] The comprehensive evaluation value R at each frequency is calculated according to the Pearson correlation coefficient and the comprehensive evaluation expression. If the frequency corresponding to the maximum comprehensive evaluation value is greater than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value R is selected as the characteristic frequency; wherein the comprehensive evaluation expression is: R = 0.8R C -0.2R T ;

[0076] In this embodiment, the lower limit of the characteristic frequency is 5 Hz, that is, when the calculated characteristic frequency is lower than 5 Hz, the characteristic frequency takes the lower limit value of 5 Hz;

[0077] The real part of the impedance at the characteristic frequency has a strong correlation with the state of charge, which can characterize the battery charge and thus provide an early warning of battery overcharge failure.

[0078] In some embodiments, Figure 2 As shown, S100 specifically includes the following steps:

[0079] S11: Use a battery charge and discharge tester to discharge the battery to 2.5V at a constant current of 0.5C, and then charge the battery to a state of charge (SOC) of 20% at a constant current of 0.1C;

[0080] S12: Set the temperature of the incubator to 5°C;

[0081] S13: leaving the battery in the incubator for 2 hours to allow the internal and external temperatures of the battery to reach equilibrium and equal the set temperature of the incubator;

[0082] S14: using an electrochemical workstation to measure the electrochemical impedance spectroscopy (EIS) of the battery; wherein the measurement frequency range is 0.1 Hz to 10 kHz, and the excitation current amplitude is 300 mA;

[0083] S15: Record the current battery state of charge (SOC) and electrochemical impedance spectroscopy (EIS) data;

[0084] S16: setting the temperature of the incubator to 25°C and 55°C in sequence, and repeating steps S13 to S15 at each set temperature;

[0085] S17: setting the state of charge (SOC) of the battery to 40%, 60% and 80% in sequence, and repeating steps S12 to S16 at each state of charge (SOC);

[0086] S18: Calculate the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency when the SOC is 60%, and then calculate the Pearson correlation coefficient between the real part of the battery impedance and the state of charge (SOC) at 25° C.;

[0087] S19: Calculate the comprehensive evaluation value at each frequency according to the comprehensive evaluation expression and the Pearson correlation coefficient, and select the frequency corresponding to the maximum comprehensive evaluation value as the characteristic frequency; in the embodiment of the present application, the characteristic frequency selected according to the experimental data is 10 Hz, and the characteristic impedance is the real part of the impedance of 10 Hz;

[0088] Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ;

[0089] S200: Acquire characteristic impedance data and calculate impedance curvature: collect characteristic impedance data, and calculate the corresponding impedance curvature sequence according to the impedance curvature calculation formula;

[0090] In some embodiments, S200 specifically includes the following steps:

[0091] S21: Use a charge and discharge tester to perform constant current charging on the battery, with a charging current multiple of x, to simulate the constant current charging process of the battery in an actual energy storage system;

[0092] S22: Use an electrochemical workstation to measure the characteristic impedance of the battery, adopt a single-frequency measurement mode, and measure and record the real part of the impedance at 10 Hz;

[0093] S23: Calculate the impedance curvature of the characteristic impedance. The calculation formula of the impedance curvature is:

[0094]

[0095] Among them, K z is the impedance curvature, z″ is the second-order differential of the characteristic impedance curve at a certain point, and z′ is the first-order differential of the characteristic impedance curve at a certain point; the impedance curvature can characterize the curvature of the characteristic impedance curve; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm;

[0096] S300: Building a clustering warning model: Calculating the parameters of the clustering model according to the battery charging current, the rated capacity of the battery, and the impedance curvature sequence, including the neighborhood radius and the minimum number of points, and then building a battery overcharge failure warning model based on the clustering algorithm;

[0097] In some embodiments, S300 specifically includes the following steps:

[0098] A battery overcharge failure warning model based on clustering algorithm is built, and the density-based spatial clustering algorithm of noise applications (DBSCAN) is used as the warning model; the DBSCAN model divides data into different clusters according to the density of samples, and has obvious advantages in data clustering of lower dimensions; the DBSCAN model is divided into two steps: the first step is to scan all sample points, find the core points among them, and form temporary clusters with points with direct density; the second step is to merge temporary clusters; scan the points in the temporary clusters to determine whether these points are core points; if there are core points, merge the temporary clusters corresponding to the core points with the current temporary clusters; repeat the merging operation until the final cluster is formed;

[0099] S31: Calculate the neighborhood radius R of two key parameters in the battery overcharge failure warning model based on clustering algorithm D and the minimum number of points MP; in order to enhance the adaptability of the algorithm, the relevant parameters are selected based on the total sample size and the characteristics of the data itself;

[0100] More specifically, the formula for calculating the minimum number of points MP based on parameters such as the charging current, the rated capacity of the battery, and the sampling rate is as follows:

[0101]

[0102] Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f zis the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes; here F is 50, that is, the minimum number of points MP is 1 / 50 of the theoretical total number of sample points;

[0103] More specifically, the neighborhood radius R of the clustering algorithm is calculated based on the minimum number of points and impedance curvature data. D The calculation formula is as follows:

[0104] R D = n·m(OD(X))

[0105] Where n is a coefficient, and m(OD(X)) represents the calculation of the average Euclidean distance of X sample points; here n is 3, and X is 2MP, that is, the neighborhood radius R D is three times the average Euclidean distance; the sample points are the impedance curvature sequence calculated in S200;

[0106] S32: According to the neighborhood radius R D The battery overcharge failure warning model based on the clustering algorithm can be constructed by using the minimum number of points MP; wherein the input variable of the battery overcharge failure warning model based on the clustering algorithm is the impedance curvature, and the output variable is the normal category or the abnormal category;

[0107] S400: Online measurement of impedance data and failure warning: Online measurement of characteristic impedance data of the battery, calculation of the corresponding impedance curvature value, and abnormal diagnosis of the impedance curvature value according to the battery overcharge failure warning model based on the clustering algorithm. If the impedance curvature value is of an abnormal category and is a positive value, the algorithm issues a warning signal.

[0108] More specifically, Figure 3 As shown, online measurement of impedance data for failure warning specifically includes the following steps:

[0109] S41: Continuously record characteristic impedance data during battery charging;

[0110] S42: Calculate the impedance curvature according to the impedance curvature calculation formula in S23;

[0111] S43: clustering the impedance curvature according to the DBSCAN model constructed in S300;

[0112] S44: Determine whether the current sample point (impedance ratio) is of the same type; if the current sample point is of the same type as the previous sample point, continue to perform clustering operations on subsequent sample points; if the current sample point is of an abnormal type, it is necessary to further determine whether the current sample point is a positive value; if it is a positive value, the algorithm issues a warning signal and ends the diagnosis; if it is a negative value, continue to perform clustering operations on subsequent sample points.

[0113] In this embodiment, during the process of constant current charging of the battery until failure, the recorded battery terminal voltage curve is as follows: Figure 4 shown by Figure 4 It can be seen that within the normal charging range of the battery, the terminal voltage rises slowly and there is a large plateau period; after the battery is overcharged, the terminal voltage rises rapidly and suddenly jumps to 28V in 4261s; 28V is the external power supply voltage, which indicates that the battery has been short-circuited and the battery has been overcharged and failed; this is because the gas production inside the battery increases after overcharging, the gas pressure gradually increases, and finally the current cut-off protection device CID (current interrupt device) inside the battery is activated to cut off the charging circuit; the action of the protection device CID can prevent the battery from exploding, but the irreversibility of its action also means battery failure, which in turn has a serious impact on the energy storage system.

[0114] The characteristic impedance curve recorded in this embodiment is as follows Figure 5 As shown by Figure 5 It can be seen that within the normal charging range of the battery, the characteristic impedance gradually decreases with the increase of the state of charge (SOC); this is because the characteristic impedance (real part at 10Hz) is negatively correlated with both SOC and temperature. When the battery is charged, both SOC and temperature increase, causing the characteristic impedance to decrease; after the battery is overcharged, the characteristic impedance begins to increase in the opposite direction. This is because the internal structure of the overcharged battery is destroyed, and a large number of abnormal electrochemical reactions begin to occur inside, thereby increasing the characteristic impedance; therefore, before the battery fails due to overcharge, the characteristic impedance shows a minimum value feature at 3902s, and this application performs an overcharge failure warning based on this minimum value feature.

[0115] The impedance curvature curve calculated in this embodiment is as follows: Figure 6 As shown in the figure, during the normal charging of the battery, the characteristic impedance curve has a small curvature and the impedance curvature fluctuates around 0; before the battery fails due to overcharging, the characteristic impedance reaches a minimum value and the impedance curvature increases suddenly; therefore, the clustering algorithm can be used to diagnose the abnormal impedance curvature point; specifically, the impedance curvature in the normal stage is near 0 and these sample points will be clustered into the normal class by DBSCAN, and the impedance curvature corresponding to the characteristic impedance minimum point is obviously far away from other points, and it will be clustered into the abnormal class; in addition to the category anomaly, the area near the minimum point is a local concave function, the corresponding second-order differential is greater than 0, and the impedance curvature is also greater than 0, which can also be used as another feature;

[0116] Therefore, the impedance curvature near the minimum point of the characteristic impedance has two characteristics: (1) the impedance curvature value near the required minimum point is significantly larger than the other curvature values, and will be diagnosed as an abnormal category by the DBSCAN model; (2) the impedance curvature value near the required minimum point is positive. Based on these two characteristics, the minimum point of the characteristic impedance can be detected, and then an early warning of battery overcharge failure can be issued;

[0117] The clustering results and warning signals of this embodiment are as follows: Figure 7 and Figure 8 As shown in the figure, in order to show the clustering results more intuitively, the clustering results are plotted in two-dimensional coordinates, and both the horizontal and vertical coordinates are impedance curvatures; Figure 6 and Figure 7 It can be seen that the clustering algorithm diagnosed two anomalies, the first anomaly was around 3600s, and the second anomaly was around 3900s; in the two-dimensional graph, these two anomaly points are far away from other points; the first anomaly is caused by the accelerated drop in impedance, and the second anomaly corresponds to the minimum point of the failure warning; since the first anomaly point has a negative impedance curvature value, the DBSCAN algorithm did not make a misjudgment; the second anomaly point has a positive impedance curvature value, which meets the judgment criteria of the positive impedance curvature value and the abnormal category, and the algorithm issues an early warning signal. After the early warning signal is issued, the clustering and diagnosis of new data can be stopped. The early warning signal changes from 0 to 1 at 3902s, and the battery overcharge failure time is 4261s. The algorithm issues an early warning signal 359s ahead of time; the experimental results verify the effectiveness and feasibility of the clustering algorithm.

[0118] Example 2

[0119] like Fig. 9 As shown, the present application provides a battery overcharge failure warning system based on electrochemical impedance, including: a characteristic frequency acquisition module 100, an impedance curvature calculation module 200, a model construction module 300 and an abnormality diagnosis module 400;

[0120] The characteristic frequency acquisition module 100 is used to measure the impedance of the battery at different temperatures and charge states, and obtain the characteristic frequency of the battery overcharge failure warning;

[0121] The impedance curvature calculation module 200 is used to measure characteristic impedance data at a characteristic frequency during battery charging and calculate an impedance curvature sequence;

[0122] The model building module 300 is used to calculate the neighborhood radius and the minimum number of points based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, and obtain a battery overcharge failure warning model based on a clustering algorithm;

[0123] The abnormal diagnosis module 400 is used to measure the characteristic impedance data of the battery at the characteristic frequency and calculate the impedance curvature value; the impedance curvature value is input into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and a warning signal is issued when the impedance curvature value is an abnormal category and is a positive value.

[0124] Further preferably, the characteristic frequency acquisition module 100 includes a temperature box, a battery charge and discharge tester, an electrochemical workstation and a data processing unit;

[0125] The temperature box is used to keep the battery in a preset temperature environment;

[0126] The battery charge and discharge tester is used to charge and discharge the battery and change the battery's state of charge;

[0127] The electrochemical workstation is used for calculating the real part of the impedance of the battery at various frequencies under different SOCs and temperatures; the data processing unit is used to calculate the Pearson correlation coefficient between the real part of the battery impedance at various frequencies and the temperature at a preset state of charge; calculate the Pearson correlation coefficient between the real part of the battery impedance at various frequencies and the state of charge at a preset temperature; calculate the comprehensive evaluation value at each frequency according to the comprehensive evaluation expression constructed based on the Pearson correlation coefficient between the real part of the battery impedance at various frequencies and the temperature at a preset state of charge and the Pearson correlation coefficient between the real part of the battery impedance at various frequencies and the state of charge at a preset temperature; if the frequency corresponding to the maximum comprehensive evaluation value is higher than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value is used as the characteristic frequency, otherwise the lower limit of the characteristic frequency is used as the characteristic frequency;

[0128] Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ; Among them, R T is the Pearson correlation coefficient between the real part of the battery impedance and temperature at each frequency under the preset state of charge; R C is the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; R is the comprehensive evaluation value.

[0129] Further preferably, the calculation formula of impedance curvature is:

[0130]

[0131] Among them, K z is the impedance curvature; z″ is the second-order differential of the characteristic impedance curve at a certain point; z′ is the first-order differential of the characteristic impedance curve at a certain point; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm.

[0132] Further preferably, the model building module 300 includes a point number calculation unit, a radius calculation unit and a model building unit;

[0133] The point number calculation unit is used to calculate the minimum number of points according to the charging current, the rated capacity of the battery and the sampling rate;

[0134] The radius calculation unit is used to calculate the neighborhood radius of the clustering algorithm based on the minimum number of points and impedance curvature data;

[0135] The model building unit is used to obtain a battery overcharge failure warning model based on a clustering algorithm according to a neighborhood radius and a minimum number of points; wherein the input variable of the battery overcharge failure warning model is impedance curvature, and the output variable is a normal category or an abnormal category.

[0136] Further preferably, the minimum number of points in the battery overcharge failure warning model construction module is:

[0137]

[0138] Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f z is the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes;

[0139] The neighborhood radius in the battery overcharge failure warning model construction module is:

[0140] R D = n·m(OD(X))

[0141] Where n is a coefficient, m(OD(X)) means calculating the average Euclidean distance of X sample points; X is 2MP; and the sample points are impedance curvature sequences.

[0142] In general, compared with the prior art, this application has the following advantages:

[0143] Compared with the existing battery overcharge safety prevention and control methods based on voltage, temperature or gas, the present application performs safety warning based on electrochemical impedance, which can reflect the safety status of the battery from abnormal changes inside the battery, with higher reliability and better speed.

[0144] Compared with the existing safety control method using raw impedance data, the present application uses impedance curvature for safety warning, which can filter out interference signals in the impedance data and reduce the misjudgment rate of the algorithm.

[0145] Compared with the existing threshold-based battery safety prevention and control methods, this application adopts an adaptive clustering algorithm to achieve safety warning. The algorithm can adjust the model parameters according to the battery's own data characteristics and current size, has strong adaptability, and solves the problem that the traditional threshold method is difficult to determine the alarm threshold.

[0146] It should be understood that the above-mentioned system is used to execute the methods in the above-mentioned embodiments. The implementation principles and technical effects of the corresponding program modules in the system are similar to those described in the above-mentioned methods. The working process of the system can refer to the corresponding process in the above-mentioned method and will not be repeated here.

[0147] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A battery overcharge failure warning method based on electrochemical impedance, characterized in that: Specifically: Measure the characteristic impedance data of the battery at the characteristic frequency and calculate the impedance curvature value; The impedance curvature value is input into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and a warning signal is issued when the impedance curvature value is an abnormal category and a positive value; The method for constructing the battery overcharge failure warning model based on the clustering algorithm specifically includes the following steps: S1: Measure the impedance of the battery at different temperatures and charge states to obtain the characteristic frequency of the battery overcharge failure warning; S2: Measure characteristic impedance data at characteristic frequencies during battery charging and calculate impedance curvature sequence; S3: Based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, the neighborhood radius and the minimum number of points are calculated, and a battery overcharge failure warning model based on the clustering algorithm is constructed.

2. The battery overcharge failure warning method according to claim 1, characterized in that: The method for obtaining the characteristic frequency of the battery overcharge failure warning in S1 specifically includes the following steps: S1.1: Calculate the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under the preset state of charge; S1.2: Calculate the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency at a preset temperature; S1.3: Calculate the comprehensive evaluation value at each frequency based on the comprehensive evaluation expression constructed based on the Pearson correlation coefficient obtained in S1.1 and the Pearson correlation coefficient obtained in S1.2; S1.4: If the frequency corresponding to the maximum comprehensive evaluation value is higher than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value is used as the characteristic frequency; otherwise, the lower limit of the characteristic frequency is used as the characteristic frequency; Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ; R T is the Pearson correlation coefficient between the real part of the battery impedance and temperature at each frequency under the preset state of charge; R C is the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; R is the comprehensive evaluation value.

3. The battery overcharge failure warning method according to claim 1, characterized in that: The calculation formula of impedance curvature is: Among them, K z is the impedance curvature; z″ is the second-order differential of the characteristic impedance curve at a certain point; z′ is the first-order differential of the characteristic impedance curve at a certain point; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm.

4. The battery overcharge failure warning method according to any one of claims 1 to 3, characterized in that: Step S3 is specifically as follows: Calculate the minimum number of points based on the charging current, the rated capacity of the battery, and the sampling rate; Calculate the neighborhood radius of the clustering algorithm based on the minimum number of points and impedance curvature data; A battery overcharge failure warning model based on clustering algorithm is constructed according to the neighborhood radius and the minimum number of points; wherein the input variable of the battery overcharge failure warning model is impedance curvature, and the output variable is normal category or abnormal category.

5. The battery overcharge failure warning method according to claim 4, characterized in that: The minimum number of points is: Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f z is the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes; The neighborhood radius is: R D =n·m(OD(X)) Where n is a coefficient, m(OD(X)) means calculating the average Euclidean distance of X sample points; X is 2MP; and the sample points are impedance curvature sequences.

6. A battery overcharge failure warning system based on electrochemical impedance, characterized in that: include: Characteristic frequency acquisition module, impedance curvature calculation module, model building module and abnormality diagnosis module; The characteristic frequency acquisition module is used to measure the impedance of the battery at different temperatures and charge states to obtain the characteristic frequency of the battery overcharge failure warning; The impedance curvature calculation module is used to measure characteristic impedance data at a characteristic frequency during battery charging and calculate an impedance curvature sequence; The model building module is used to calculate the neighborhood radius and the minimum number of points based on the battery charging current, the rated capacity of the battery and the impedance curvature sequence, and to build a battery overcharge failure warning model based on a clustering algorithm; The abnormal diagnosis module is used to measure the characteristic impedance data of the battery at the characteristic frequency and calculate the impedance curvature value; the impedance curvature value is input into the battery overcharge failure warning model based on the clustering algorithm for abnormal diagnosis, and a warning signal is issued when the impedance curvature value is of an abnormal category and is a positive value.

7. The battery overcharge failure warning system according to claim 6, characterized in that: The characteristic frequency acquisition module includes a temperature box, a battery charge and discharge tester, an electrochemical workstation and a data processing unit; The temperature box is used to keep the battery in a preset temperature environment; the battery charge and discharge tester is used to charge and discharge the battery and change the battery's state of charge; the electrochemical workstation is used to measure the real part of the battery's impedance at different SOCs and temperatures at various frequencies; The data processing unit is used to calculate the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under a preset state of charge; Calculate the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under a preset temperature; Calculate the comprehensive evaluation value at each frequency according to the comprehensive evaluation expression constructed by the Pearson correlation coefficient between the real part of the battery impedance and the temperature at each frequency under the preset state of charge and the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; If the frequency corresponding to the maximum comprehensive evaluation value is higher than the lower limit of the characteristic frequency, the frequency corresponding to the maximum comprehensive evaluation value is taken as the characteristic frequency, otherwise the lower limit of the characteristic frequency is taken as the characteristic frequency; Among them, the comprehensive evaluation expression is: R = 0.8R C -0.2R T ; R T is the Pearson correlation coefficient between the real part of the battery impedance and temperature at each frequency under the preset state of charge; R C is the Pearson correlation coefficient between the real part of the battery impedance and the state of charge at each frequency under the preset temperature; R is the comprehensive evaluation value.

8. The battery overcharge failure warning system according to claim 6, characterized in that: The calculation formula of impedance curvature is: Among them, K z is the impedance curvature; z″ is the second-order differential of the characteristic impedance curve at a certain point; z′ is the first-order differential of the characteristic impedance curve at a certain point; when calculating the impedance curvature, the unit of time is second and the unit of impedance is milliohm.

9. The battery overcharge failure warning system according to any one of claims 6 to 8, characterized in that: The model building module includes a point number calculation unit, a radius calculation unit and a model building unit; The point number calculation unit is used to calculate the minimum number of points according to the charging current, the rated capacity of the battery and the sampling rate; The radius calculation unit is used to calculate the neighborhood radius of the clustering algorithm based on the minimum number of points and impedance curvature data; The model building unit is used to build a battery overcharge failure warning model based on a clustering algorithm according to the neighborhood radius and the minimum number of points; wherein the input variable of the battery overcharge failure warning model is impedance curvature, and the output variable is a normal category or an abnormal category.

10. The battery overcharge failure warning system according to claim 9, characterized in that: The minimum number of points in the battery overcharge failure warning model construction module is: Where, f represents rounding down; Q is the rated capacity of the battery in ampere-hours; f z is the sampling rate of characteristic impedance; F is the segmentation factor; x is the charging current in amperes; The neighborhood radius in the battery overcharge failure warning model construction module is: R D =n·m(OD(X)) Where n is a coefficient, m(OD(X)) means calculating the average Euclidean distance of X sample points; X is 2MP; and the sample points are impedance curvature sequences.

Images