Method and device for detecting lithium precipitation condition of lithium battery, and nonvolatile storage medium

Abstract

The invention discloses a method and device for detecting the lithium precipitation condition of a lithium battery and a nonvolatile storage medium. The method comprises the following steps: acquiring various operation data of a target battery, the various operation data including a voltage value, a capacity value and a temperature difference value; the multiple kinds of operation data are imported into a target machine learning model, the target machine learning model outputs a detection result used for representing whether the target battery separates out lithium or not, and the target machine learning model outputs the detection result through the multiple kinds of operation data of the multiple sample batteries. And training the initial machine learning model according to the actual lithium precipitation conditions of the plurality of sample batteries. The technical problem of low detection efficiency of the lithium precipitation condition of the lithium battery is solved.

Description

Technical Field

[0001] The present invention relates to the field of battery health monitoring and management, and in particular to a method, device and non-volatile storage medium for detecting lithium deposition in a lithium battery. Background Art

[0002] Lithium batteries have high energy density, are environmentally friendly, safe and efficient, and have a long service life. They have been widely used in the field of energy storage. However, lithium-ion battery energy storage also presents certain safety risks, and lithium plating is one of the factors that affect the safety of lithium batteries. Lithium plating will reduce battery capacity and shorten battery life. In severe cases, it will cause a short circuit in the battery, posing the risk of fire and explosion. In order to reduce the danger caused by lithium plating in lithium batteries, lithium batteries need to be tested regularly. Currently, there are two types of lithium plating detection methods for lithium batteries: destructive and non-destructive. The destructive method is to disassemble and observe the lithium battery, which is more complicated and costly. The non-destructive method is based on the electrochemical signals during the battery operation process. It is necessary to manually analyze the detection of each battery separately to determine whether there is lithium plating, which is less efficient.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a method, device, and non-volatile storage medium for detecting lithium deposition in a lithium battery, so as to at least solve the technical problem of low efficiency in detecting lithium deposition in a lithium battery.

[0005] According to one aspect of an embodiment of the present invention, a method for detecting lithium deposition in a lithium battery is provided, comprising: obtaining a variety of operating data of a target battery, wherein the various operating data include voltage value, capacity value, and temperature; importing the various operating data into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium deposition, wherein the target machine learning model is obtained by training an initial machine learning model through the various operating data of a plurality of sample batteries and the actual lithium deposition conditions of the plurality of sample batteries.

[0006] Optionally, the target machine learning model is trained in the following manner, including: obtaining multiple operating data of each of the multiple sample batteries, and actual lithium plating conditions of the multiple sample batteries; performing feature extraction on the multiple operating data of each of the multiple sample batteries, and determining multiple characteristic parameters of the multiple sample batteries; generating data points corresponding to each of the multiple sample batteries based on the multiple characteristic parameters of the multiple sample batteries, and obtaining multiple data points; importing the multiple data points into the initial machine learning model for classification, and obtaining classification results of the multiple data points; training the initial machine learning model based on the actual lithium plating conditions of the multiple sample batteries, and the classification results of the multiple data points, and obtaining the target machine learning model.

[0007] Optionally, feature extraction is performed on a plurality of operating data of each of the plurality of sample batteries to determine a plurality of characteristic parameters of each of the plurality of sample batteries, including: performing a plurality of charge and discharge cycles on the plurality of sample batteries to obtain sample batteries after a plurality of charge and discharge cycles; determining the voltage, capacity, and temperature of each of the sample batteries after a plurality of charge and discharge cycles, and generating a curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries; and obtaining a plurality of characteristic parameters of each of the plurality of sample batteries based on the curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries.

[0008] Optionally, multiple sample batteries are subjected to multiple charge and discharge cycles to obtain sample batteries after multiple charge and discharge cycles, including: grouping the multiple sample batteries to obtain multiple sample battery groups; setting different temperatures and charge rates for the multiple sample battery groups; based on the respective temperatures and charge rates of the multiple sample battery groups, respectively subjecting the sample batteries in the multiple sample battery groups to charge and discharge cycles until a predetermined termination condition is reached, to obtain sample batteries after multiple charge and discharge cycles.

[0009] Optionally, the initial machine learning model is trained according to the actual lithium plating conditions of multiple sample batteries and the classification results of multiple data points to obtain a target machine learning model, including: determining multiple neighboring data points of each of the multiple data points, wherein the distance between the multiple data points and the corresponding multiple neighboring data points is less than a predetermined distance; determining the skewness of each of the multiple data points according to the distances between the multiple data points and the corresponding multiple neighboring data points, wherein the skewness of each of the multiple data points is used to characterize the distribution of the distances between the corresponding data points and the corresponding multiple neighboring data points; and classifying the multiple data points according to the skewness of each of the multiple data points to obtain classification results of the multiple data points.

[0010] Optionally, the multiple data points are classified according to their respective skewnesses to obtain classification results for the multiple data points, including: if the skewness is greater than a predetermined threshold, the corresponding data point is classified into a first category; if the skewness is less than or equal to the predetermined threshold, the corresponding data point is classified into a second category.

[0011] According to another aspect of an embodiment of the present invention, there is provided a device for detecting lithium deposition in a lithium battery, comprising: a first acquisition module, for acquiring a variety of operating data of a target battery, wherein the various operating data include voltage value, capacity value, and temperature; a first import module, for importing the various operating data into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium deposition, wherein the target machine learning model is obtained by training an initial machine learning model through the various operating data of a plurality of sample batteries and the actual lithium deposition conditions of the plurality of sample batteries.

[0012] According to another aspect of an embodiment of the present invention, a non-volatile storage medium is provided, wherein the non-volatile storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing any one of the methods for detecting lithium deposition in a lithium battery.

[0013] According to another aspect of an embodiment of the present invention, there is provided an electronic device, comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the methods for detecting lithium deposition in a lithium battery.

[0014] According to another aspect of the embodiments of the present invention, a computer program product is provided, including a computer program, which implements any one of the above methods for detecting lithium deposition in a lithium battery when executed by a processor.

[0015] In an embodiment of the present invention, multiple operating data of a target battery are obtained, wherein the multiple operating data include voltage value, capacity value, and temperature; the multiple operating data are imported into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium deposition. The target machine learning model is obtained by training an initial machine learning model with multiple operating data of multiple sample batteries and actual lithium deposition conditions of multiple sample batteries, thereby solving the technical problem of low efficiency in detecting lithium deposition conditions of lithium batteries, and achieving the purpose of directly judging whether a lithium battery has lithium deposition conditions by using a machine learning model, thereby achieving the technical effect of improving the low efficiency in detecting lithium deposition conditions of lithium batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 is a flow chart of a method for detecting lithium deposition in a lithium battery provided according to an embodiment of the present invention;

[0018] Figure 2 is a schematic diagram of a device for detecting lithium deposition in a lithium battery according to an embodiment of the present invention;

[0019] Figure 3 is a schematic diagram of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] According to an embodiment of the present invention, a method embodiment for detecting lithium deposition in a lithium battery is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0023] Figure 1 is a flow chart of a method for detecting lithium deposition in a lithium battery according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0024] Step S102, obtaining various operating data of the target battery, wherein the various operating data include voltage value, capacity value, and temperature.

[0025] In this step, the lithium electron lithium precipitation phenomenon refers to the phenomenon that the lithium in the lithium battery will precipitate on the electrode surface to form lithium dendrites during the charging or discharging process. From the perspective of external conditions, there are many factors that induce lithium precipitation. In addition to the defects of the battery itself, the probability of lithium precipitation under working conditions such as low temperature, fast charging, and overcharging is high. And from the fundamental understanding of electrochemical reaction, lithium precipitation occurs because the negative electrode potential is lower than the equilibrium potential of the lithium precipitation reaction under certain conditions. Therefore, the internal resistance R of the battery is large at low temperature, the charging current I is large during fast charging, and the internal resistance R at the high charge state (State of Charge, SOC) during overcharging also increases significantly, all of which will lead to a larger negative electrode reaction overpotential (η=IR), thereby quickly reducing the negative electrode potential so that the negative electrode potential is lower than the equilibrium potential of the lithium precipitation reaction, causing the lithium precipitation phenomenon. The lithium precipitation phenomenon of lithium batteries will lead to a decrease in the safety of the battery, which may cause problems such as short circuit and overheating, and even cause battery explosion. Therefore, the lithium precipitation phenomenon of lithium batteries is one of the issues that need to be focused on and prevented during the use of lithium batteries. In order to reduce the probability of danger, lithium batteries need to be inspected regularly. When inspecting lithium batteries, it is necessary to collect data generated by the lithium batteries during the charging and discharging process, including the voltage value, capacity and surface temperature of the lithium batteries. Based on these data, it can be determined whether lithium plating occurs.

[0026] Step S104, importing multiple operating data into the target machine learning model, and the target machine learning model outputs a detection result used to characterize whether the target battery has lithium plating, wherein the target machine learning model is obtained by training the initial machine learning model through the multiple operating data of multiple sample batteries and the actual lithium plating conditions of the multiple sample batteries.

[0027] In this step, an initial machine learning model is constructed by collecting various operating data of multiple sample batteries, such as voltage, temperature, capacity and other information, as well as the actual lithium deposition of these batteries. These data are imported into the initial machine learning model for training so that the initial machine model can learn how to predict whether the battery will deposit lithium based on the operating data of the battery. For example, suppose there are 100 sample batteries, each battery has 3 different operating data, and it is known whether each battery is lithium deposited. These data can be input into the machine learning model to let the model learn the operating data of the battery and divide the operating data into two categories, one representing the existence of lithium deposition, and the other representing the absence of lithium deposition, and the relationship between the lithium deposition situation. Through training, the model can find a pattern or law, so that new batteries can be tested for lithium deposition. When a new battery needs to be tested for lithium deposition, its operating data can be input into the trained machine learning model, and the model will automatically output a result to characterize whether the battery will deposit lithium. This method can help quickly perform battery lithium deposition detection and improve the safety and reliability of the battery.

[0028] In an optional embodiment, the target machine learning model is trained in the following manner, including: obtaining multiple operating data of each of the multiple sample batteries, and actual lithium plating conditions of the multiple sample batteries; performing feature extraction on the multiple operating data of each of the multiple sample batteries, and determining multiple characteristic parameters of the multiple sample batteries; generating data points corresponding to each of the multiple sample batteries based on the multiple characteristic parameters of the multiple sample batteries, and obtaining multiple data points; importing the multiple data points into the initial machine learning model for classification, and obtaining classification results of the multiple data points; training the initial machine learning model based on the actual lithium plating conditions of the multiple sample batteries, and the classification results of the multiple data points, and obtaining the target machine learning model.

[0029] Optionally, this optional embodiment mainly illustrates the training method of the target machine learning model, and can collect various operating data of each sample battery after charging and discharging, such as voltage, capacity, temperature, etc., and extract multiple characteristic parameters of the operating data corresponding to each sample battery, and the number of characteristic parameters of each sample battery is the same. Take each sample battery as a data point, and the same number of characteristic parameters corresponding to each sample battery as the dimension of the data point. For example, 3 characteristic parameters are extracted from the operating data collected by sample battery 1, and the data point 1 generated according to sample battery 1 is a three-dimensional data point. All data points are imported into the initial machine learning model. The initial machine model can calculate the skewness of each data point and divide all data points into two categories according to the skewness of each data point, one category represents the sample battery with lithium plating phenomenon, and the other category represents the sample battery without lithium plating phenomenon. Then the sample batteries are disassembled one by one to determine whether the sample battery really has lithium plating phenomenon, and the disassembly results are compared with the classification results of the initial machine learning model to verify whether the classification results obtained by the initial machine learning model are accurate. If the classification results obtained by the initial machine learning model are accurate, it is determined that the initial machine learning model training is completed and the target machine learning model is obtained. If the classification result obtained by the initial machine learning model is inaccurate, the model can be trained further by adjusting the parameters in the model until the classification result is consistent with the result observed after disassembly. Then the training is stopped to obtain the target machine learning model.

[0030] In an optional embodiment, feature extraction is performed on a plurality of operating data of each of the plurality of sample batteries to determine a plurality of characteristic parameters of each of the plurality of sample batteries, including: performing a plurality of charge and discharge cycles on the plurality of sample batteries to obtain sample batteries after a plurality of charge and discharge cycles; determining a voltage, a capacity, and a temperature of each of the sample batteries after a plurality of charge and discharge cycles, and generating a curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries; and obtaining a plurality of characteristic parameters of each of the plurality of sample batteries based on the curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries.

[0031] Optionally, a batch of lithium iron phosphate soft-pack batteries of the same model and capacity can be selected as sample batteries, and an aging experiment can be performed on the sample batteries to simulate the actual situation, in which lithium batteries may undergo lithium precipitation after being used for a period of time under different environments. The process of performing an aging test on the sample batteries is as follows: the sample batteries are divided into multiple groups, and each group of sample batteries is subjected to multiple charge and discharge cycles under different temperatures, different charge rates, and the same discharge rate to obtain aged sample batteries. The aged batteries are subjected to a final round of charge and discharge processes according to the original temperature conditions and charging conditions, as well as discharge conditions different from those in the aging stage, to obtain the sample batteries to be tested, and the voltage V, temperature T, and capacity Q of the sample batteries to be tested are extracted to generate two curves, one of which is the voltage-capacity curve, namely the VQ curve. Differentiate the VQ curve with respect to Q to obtain the dV / dQ curve, and draw the characteristic curve with dV / dQ as the ordinate and Q as the abscissa. Record the difference in capacity corresponding to the maximum and minimum values ​​of the curve, the difference between the maximum and minimum values ​​on the curve, the area of ​​the figure enclosed by the dV / dQ curve and the horizontal axis at the maximum point, and the area of ​​the figure enclosed by the dV / dQ curve and the horizontal axis at the minimum point, and get a total of 4 characteristic parameters. The second is the temperature difference-capacity curve of each sample battery, that is, the ΔT-Q curve. Among them, the temperature difference ΔT is the difference between the current surface temperature of the battery during the last round of discharge and the surface temperature when it just started to discharge. Record the area of ​​the figure enclosed by the ΔT-Q curve and the Q axis, as well as the maximum slope of the ΔT-Q curve, and get 2 characteristic parameters, a total of 6 characteristic parameters.

[0032] In an optional embodiment, multiple sample batteries are subjected to multiple rounds of charge and discharge cycles to obtain sample batteries after multiple charge and discharge cycles, including: grouping the multiple sample batteries to obtain multiple sample battery groups; setting different temperatures and charge rates for the multiple sample battery groups; based on the respective temperatures and charge rates of the multiple sample battery groups, respectively subjecting the sample batteries in the multiple sample battery groups to charge and discharge cycles until a predetermined termination condition is reached, to obtain sample batteries after multiple charge and discharge cycles.

[0033] Optionally, a batch of lithium iron phosphate soft-pack batteries of the same model and capacity can be selected as sample batteries, and multiple sample batteries can be grouped, each group containing multiple sample batteries, the purpose of which is to reduce the classification result error caused by other reasons, such as the sample battery has lithium precipitation before the experiment, set different temperatures and charge and discharge rates to simulate the actual charging and discharging of the lithium battery, such as 75 sample batteries can be selected, the 75 sample batteries can be divided into 25 groups, each group containing 3 sample batteries, the 25 groups of batteries are charged and discharged under five different temperatures (-10°C, 0°C, 15°C, 25°C, 35°C), five charging rates (0.2C, 0.5C, 0.8C, 1C, 1.5C) and the same 1C rate discharge conditions, until the capacity of the sample battery drops to 70% of the initial capacity or the number of charge and discharge cycles reaches 500 times, to obtain an aged sample battery, the aged sample battery is subjected to a round of charge and discharge under the same temperature, charging rate, and 0.05C rate discharge conditions to obtain a sample battery to be tested.

[0034] In an optional embodiment, an initial machine learning model is trained according to actual lithium plating conditions of multiple sample batteries and classification results of multiple data points to obtain a target machine learning model, including: determining multiple adjacent data points of each of the multiple data points, wherein the distance between the multiple data points and the corresponding multiple adjacent data points is less than a predetermined distance; determining the skewness of each of the multiple data points according to the distances between the multiple data points and the corresponding multiple adjacent data points, wherein the skewness of each of the multiple data points is used to characterize the distribution of the distances between the corresponding data points and the corresponding multiple adjacent data points; and classifying the multiple data points according to the skewness of each of the multiple data points to obtain classification results of the multiple data points.

[0035] Optionally, each sample battery is taken as a data point. Each sample battery has 6 characteristic parameters, that is, each data point is six-dimensional data. The distance Kdis(x) between the data point x and other data points is calculated. The calculation formula is:

[0036]

[0037] Among them, n is the dimension of the data object, x i is the current data object, x k is the kth data point.

[0038] Select data points whose distance to data point x is less than a predetermined distance as multiple neighbors of data point x, and determine the skewness KDS(x) between data point x and multiple neighbors. The calculation formula is:

[0039]

[0040] Among them, Kdis(i) represents the distance between the data point x and its k nearest neighbors, and knumber is the total number of neighbors of the data point x.

[0041] According to the above method, multiple neighbors of each data point are determined and the skewness of each data point is calculated, and all data points are divided into multiple categories according to the skewness of each data point.

[0042] In an optional embodiment, multiple data points are classified according to their respective skewnesses to obtain classification results for the multiple data points, including: if the skewness is greater than a predetermined threshold, the corresponding data point is classified into a first category; if the skewness is less than or equal to the predetermined threshold, the corresponding data point is classified into a second category.

[0043] Optionally, the data points can be divided into two categories based on the comparison result of the skewness of the data points with a predetermined threshold. For example, if the predetermined threshold is 0.5, the data points with a skewness greater than 0.5 are classified into the first category, and the result corresponding to the first category may be that there is no lithium deposition in the sample battery. The data points with a skewness less than or equal to 0.5 are classified into the second category, and the result corresponding to the second category may be that there is lithium deposition in the sample battery.

[0044] Optionally, when determining the predetermined threshold and the number of neighbors, a 5-fold cross validation method can be used to select them so as to obtain a more accurate classification result, that is, when training the initial machine learning model, the number of neighbors and the predetermined threshold are first given, and 75 data are randomly divided into 5 groups, each with 15 data, and the 5 groups of data are cross-validated 5-fold. The specific process of cross validation can be: randomly divide 75 data points into 5 groups, each with 15 data, and for each group of data, one of them is used as test data and the remaining 4 groups are used as training data. In each training, the initial machine learning model is trained using a given number of different neighbors and a predetermined threshold, and predictions are made on the test data. The accuracy of each test is calculated and recorded. Repeat the above steps until each group of data is used as test data, and the average accuracy of the 5 tests is calculated as an evaluation indicator of the number of neighbors and the predetermined threshold. Repeat the attempts of different combinations of the number of neighbors and the predetermined threshold, and select the number of neighbors and the predetermined threshold with higher accuracy as the final model parameters.

[0045] By the above method, a variety of operating data of the target battery is obtained, wherein the various operating data include voltage value, capacity value, and temperature; the various operating data are imported into the target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium deposition, wherein the target machine learning model is obtained by training an initial machine learning model with the various operating data of multiple sample batteries and the actual lithium deposition conditions of the multiple sample batteries, thereby solving the technical problem of low efficiency in detecting lithium deposition conditions of lithium batteries, and achieving the purpose of directly judging whether the lithium battery has lithium deposition conditions by using the machine learning model, thereby achieving the technical effect of improving the low efficiency in detecting lithium deposition conditions of lithium batteries.

[0046] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation mode:

[0047] Step S1, extracting features from a plurality of operating data of a plurality of sample batteries, and determining a plurality of characteristic parameters of a plurality of sample batteries, including: performing a plurality of charge and discharge cycles on a plurality of sample batteries, and obtaining sample batteries after a plurality of charge and discharge cycles; determining a voltage, a capacity, and a temperature of each sample battery after a plurality of charge and discharge cycles, and generating a curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries; and obtaining a plurality of characteristic parameters of each of the plurality of sample batteries according to the curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries.

[0048] Optionally, a batch of lithium iron phosphate soft-pack batteries of the same model and capacity can be selected as sample batteries. An aging experiment is performed on the sample batteries to simulate the actual situation. After being used for a period of time in different environments, lithium batteries may have lithium precipitation. The process of aging test on the sample batteries is as follows: the sample batteries are divided into multiple groups, and each group of sample batteries is subjected to multiple charge and discharge cycles under different temperatures, different charging rates, and the same discharge rate to obtain aged sample batteries. The aged batteries are subjected to a final round of charge and discharge process according to the original temperature conditions and charging conditions, as well as discharge conditions different from the aging stage, to obtain the sample batteries to be tested, and the voltage V, temperature T, and capacity Q of the sample batteries to be tested are extracted to generate two curves, one of which is the voltage-capacity curve, namely the VQ curve. Differentiate the VQ curve with respect to Q to obtain the dV / dQ curve, and draw the characteristic curve with dV / dQ as the ordinate and Q as the abscissa. Record the difference in capacity corresponding to the maximum and minimum values ​​of the curve, the difference between the maximum and minimum values ​​on the curve, the area of ​​the figure enclosed by the dV / dQ curve and the horizontal axis at the maximum point, and the area of ​​the figure enclosed by the dV / dQ curve and the horizontal axis at the minimum point, and get a total of 4 characteristic parameters. The second is the temperature difference-capacity curve of each sample battery, that is, the ΔT-Q curve. Among them, the temperature difference ΔT is the difference between the current surface temperature of the battery during the last round of discharge and the surface temperature when it just started to discharge. Record the area of ​​the figure enclosed by the ΔT-Q curve and the Q axis, as well as the maximum slope of the ΔT-Q curve, and get 2 characteristic parameters, a total of 6 characteristic parameters.

[0049] Step S11, performing multiple rounds of charge and discharge cycles on multiple sample batteries to obtain sample batteries after multiple charge and discharge cycles, including: grouping the multiple sample batteries to obtain multiple sample battery groups; setting different temperatures and charge rates for the multiple sample battery groups; based on the respective temperatures and charge rates of the multiple sample battery groups, performing charge and discharge cycles on the sample batteries in the multiple sample battery groups respectively until a predetermined termination condition is reached to obtain sample batteries after multiple charge and discharge cycles.

[0050] Optionally, a batch of lithium iron phosphate soft-pack batteries of the same model and capacity can be selected as sample batteries, and multiple sample batteries can be grouped, each group containing multiple sample batteries, the purpose of which is to reduce the classification result error caused by other reasons, such as the sample battery has lithium precipitation before the experiment, set different temperatures and charge and discharge rates to simulate the actual charging and discharging of the lithium battery, such as 75 sample batteries can be selected, the 75 sample batteries can be divided into 25 groups, each group containing 3 sample batteries, the 25 groups of batteries are charged and discharged under five different temperatures (-10°C, 0°C, 15°C, 25°C, 35°C), five charging rates (0.2C, 0.5C, 0.8C, 1C, 1.5C) and the same 1C rate discharge conditions, until the capacity of the sample battery drops to 70% of the initial capacity or the number of charge and discharge cycles reaches 500 times, to obtain an aged sample battery, the aged sample battery is subjected to a round of charge and discharge under the same temperature, charging rate, and 0.05C rate discharge conditions to obtain a sample battery to be tested.

[0051] Step S2, generating data points corresponding to the plurality of sample batteries respectively according to the plurality of characteristic parameters of the plurality of sample batteries respectively, to obtain a plurality of data points.

[0052] Optionally, each sample battery is taken as a data point, and multiple characteristic parameters corresponding to each sample battery are taken as dimensions of the data point. For example, three characteristic parameters are extracted from the operating data collected by sample battery 1, and data point 1 generated based on sample battery 1 is a three-dimensional data point.

[0053] Step S3, importing multiple data points into the initial machine learning model for classification, and obtaining classification results of the multiple data points.

[0054] Optionally, all data points are imported into an initial machine learning model, which can calculate the skewness of each data point and classify all data points into two categories according to the skewness of each data point, one category representing that lithium plating occurs in the sample battery, and the other category representing that lithium plating does not occur in the sample battery.

[0055] Step S4, training the initial machine learning model according to the actual lithium plating conditions of multiple sample batteries and the classification results of multiple data points to obtain a target machine learning model, including: determining multiple adjacent data points of each of the multiple data points, wherein the distance between the multiple data points and the corresponding multiple adjacent data points is less than a predetermined distance; determining the skewness of each of the multiple data points according to the distances between the multiple data points and the corresponding multiple adjacent data points, wherein the skewness of each of the multiple data points is used to characterize the distribution of the distances between the corresponding data points and the corresponding multiple adjacent data points; classifying the multiple data points according to the skewness of each of the multiple data points to obtain the classification results of the multiple data points.

[0056] Optionally, each sample battery is taken as a data point. Each sample battery has 6 characteristic parameters, that is, each data point is six-dimensional data. The distance Kdis(x) between the data point x and other data points is calculated. The calculation formula is:

[0057]

[0058] Among them, n is the dimension of the data object, x i is the current data object, x k is the kth data point.

[0059] Select data points whose distance to data point x is less than a predetermined distance as multiple neighbors of data point x, and determine the skewness KDS(x) between data point x and multiple neighbors. The calculation formula is:

[0060]

[0061] Among them, Kdis(i) represents the distance between the data point x and its k nearest neighbors, and knumber is the total number of neighbors of the data point x.

[0062] According to the above method, multiple neighbors of each data point are determined and the skewness of each data point is calculated, and all data points are divided into multiple categories according to the skewness of each data point.

[0063] Step S41, classifying the multiple data points according to their respective skewnesses to obtain classification results of the multiple data points, including: if the skewness is greater than a predetermined threshold, classifying the corresponding data point into a first category; if the skewness is less than or equal to the predetermined threshold, classifying the corresponding data point into a second category.

[0064] Optionally, the data points can be divided into two categories based on the comparison result of the skewness of the data points with a predetermined threshold. For example, if the predetermined threshold is 0.5, the data points with a skewness greater than 0.5 are classified into the first category, and the result corresponding to the first category may be that there is no lithium deposition in the sample battery. The data points with a skewness less than or equal to 0.5 are classified into the second category, and the result corresponding to the second category may be that there is lithium deposition in the sample battery.

[0065] Optionally, when determining the predetermined threshold and the number of neighbors, a 5-fold cross validation method can be used to select them so as to obtain a more accurate classification result, that is, when training the initial machine learning model, the number of neighbors and the predetermined threshold are first given, and 75 data are randomly divided into 5 groups, each with 15 data, and the 5 groups of data are cross-validated 5-fold. The specific process of cross validation can be: randomly divide 75 data points into 5 groups, each with 15 data, and for each group of data, one of them is used as test data and the remaining 4 groups are used as training data. In each training, the initial machine learning model is trained using a given number of different neighbors and a predetermined threshold, and predictions are made on the test data. The accuracy of each test is calculated and recorded. Repeat the above steps until each group of data is used as test data, and the average accuracy of the 5 tests is calculated as an evaluation indicator of the number of neighbors and the predetermined threshold. Repeat the attempts of different combinations of the number of neighbors and the predetermined threshold, and select the number of neighbors and the predetermined threshold with higher accuracy as the final model parameters.

[0066] The above optional implementation manner achieves at least the following effects: the technical problem of low efficiency in detecting lithium deposition in lithium batteries is solved, and the purpose of directly judging whether lithium deposition occurs in lithium batteries by using a machine learning model can be achieved, thereby achieving the technical effect of improving the low efficiency in detecting lithium deposition in lithium batteries.

[0067] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0068] In this embodiment, a device for detecting lithium deposition in a lithium battery is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the terms "module" and "device" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0069] According to an embodiment of the present invention, there is also provided an embodiment of a device for implementing a method for detecting lithium deposition in a lithium battery. Figure 2 is a schematic diagram of a device for detecting lithium deposition in a lithium battery according to an embodiment of the present invention. Figure 2 As shown, the above-mentioned device for detecting lithium deposition of a lithium battery includes a first acquisition module 21 and a first introduction module 22. The device is described below.

[0070] The first acquisition module 21 is used to acquire various operating data of the target battery, wherein the various operating data include voltage value, capacity value, and temperature.

[0071] The first import module 22 is connected to the first acquisition module 21, and is used to import a variety of operating data into a target machine learning model, and the target machine learning model outputs a detection result used to characterize whether the target battery has lithium plating. The target machine learning model is obtained by training an initial machine learning model through a variety of operating data of multiple sample batteries and the actual lithium plating conditions of the multiple sample batteries.

[0072] In a device for detecting lithium deposition of a lithium battery provided by an embodiment of the present invention, a first acquisition module is set to obtain multiple operating data of a target battery, wherein the multiple operating data include voltage value, capacity value, and temperature; a first import module is used to import the multiple operating data into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium deposition, wherein the target machine learning model is obtained by training an initial machine learning model with multiple operating data of multiple sample batteries and actual lithium deposition conditions of the multiple sample batteries, thereby solving the technical problem of low efficiency in detecting lithium deposition of lithium batteries, and achieving the purpose of directly judging whether a lithium battery has lithium deposition using a machine learning model, thereby achieving the technical effect of improving the low efficiency in detecting lithium deposition of lithium batteries.

[0073] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0074] It should be noted that the first acquisition module 21 and the first import module 22 correspond to steps S102 to S104 in the embodiment, and the examples and application scenarios implemented by the modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiments. It should be noted that the modules as part of the device can be run in a computer terminal.

[0075] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in the embodiment, which will not be repeated here.

[0076] The above-mentioned device for detecting lithium deposition of lithium batteries may also include a processor and a memory, including the first acquisition module 21, the first import module 2, etc., which are stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0077] The processor includes a kernel, which retrieves the corresponding program unit from the memory. There can be one or more kernels. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0078] An embodiment of the present invention provides a non-volatile storage medium on which a program is stored. When the program is executed by a processor, a method for detecting lithium deposition of a lithium battery is implemented.

[0079] like Figure 3 As shown, an embodiment of the present invention provides an electronic device, the electronic device 10 includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented: the memory is used to store a computer program, wherein, when the computer program is executed by the processor, the processor implements the above-mentioned method for detecting lithium deposition in a lithium battery. The device in this article may be a server, a PC, etc.

[0080] The present invention also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that initializes the following method steps: computer instructions are executed by a processor to perform the above-mentioned method for detecting lithium deposition in a lithium battery.

[0081] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0082] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0086] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0087] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0088] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0089] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0090] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for detecting lithium deposition in a lithium battery, characterized in that: include: Acquire various operating data of the target battery, wherein the various operating data include voltage value, capacity value, and temperature difference value; The multiple operating data are imported into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium plating, wherein the target machine learning model is obtained by training an initial machine learning model through the multiple operating data of each of the multiple sample batteries and the actual lithium plating conditions of the multiple sample batteries.

2. The method according to claim 1, characterized in that The target machine learning model is trained in the following manner, including: Acquire various operation data of each of the plurality of sample batteries and actual lithium deposition conditions of the plurality of sample batteries; Extracting features from the various operating data of the multiple sample batteries to determine the multiple characteristic parameters of the multiple sample batteries; Generating data points corresponding to each of the plurality of sample batteries according to the plurality of characteristic parameters of each of the plurality of sample batteries to obtain a plurality of data points; Importing the multiple data points into the initial machine learning model for classification to obtain classification results of the multiple data points; The initial machine learning model is trained according to the actual lithium plating conditions of the multiple sample batteries and the classification results of the multiple data points to obtain the target machine learning model.

3. The method according to claim 2, characterized in that The extracting features of the various operating data of the plurality of sample batteries to determine the plurality of characteristic parameters of the plurality of sample batteries includes: Performing multiple charge-discharge cycles on the multiple sample batteries to obtain sample batteries after multiple charge-discharge cycles; Determine the voltage, capacity, and temperature of each of the sample batteries after the plurality of charge and discharge cycles, and generate a curve for characterizing the charge and discharge characteristics of each of the plurality of sample batteries; A plurality of characteristic parameters of each of the plurality of sample batteries are obtained according to the curves used to characterize the charge and discharge characteristics of each of the plurality of sample batteries.

4. The method according to claim 3, characterized in that The step of performing multiple charge-discharge cycles on the multiple sample batteries to obtain sample batteries after multiple charge-discharge cycles includes: Grouping the plurality of sample batteries to obtain a plurality of sample battery groups; Setting different temperatures and charging rates for the multiple sample battery packs respectively; Based on the respective temperatures and charging rates of the plurality of sample battery groups, the sample batteries in the plurality of sample battery groups are respectively subjected to charge and discharge cycles until a predetermined termination condition is reached, thereby obtaining the sample batteries after the plurality of charge and discharge cycles.

5. The method according to claim 2, characterized in that: The initial machine learning model is trained according to the actual lithium deposition conditions of the multiple sample batteries and the classification results of the multiple data points to obtain the target machine learning model, including: determining a plurality of neighboring data points of each of the plurality of data points, wherein a distance between the plurality of data points and the corresponding plurality of neighboring data points is less than a predetermined distance; Determine the skewness of each of the multiple data points according to the distances between each of the multiple data points and the corresponding multiple adjacent data points, wherein the skewness of each of the multiple data points is used to characterize the distribution of the distances between each of the corresponding data points and the corresponding multiple adjacent data points; The multiple data points are classified according to the skewness of each of the multiple data points to obtain classification results of the multiple data points.

6. The method according to claim 5, characterized in that According to the skewness of each of the plurality of data points, Classifying the multiple data points to obtain classification results of the multiple data points includes: If the skewness is greater than a predetermined threshold, classifying the corresponding data point into a first category; If the skewness is less than or equal to a predetermined threshold, the corresponding data point is classified into a second category.

7. A device for detecting lithium deposition in a lithium battery, characterized in that: include: A first acquisition module, used to acquire various operating data of the target battery, wherein the various operating data include voltage value, capacity value, and temperature; The first import module is used to import the multiple operating data into a target machine learning model, and the target machine learning model outputs a detection result for characterizing whether the target battery has lithium plating, wherein the target machine learning model is obtained by training an initial machine learning model through the multiple operating data of each of the multiple sample batteries and the actual lithium plating conditions of the multiple sample batteries.

8. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method for detecting lithium deposition in a lithium battery as described in any one of claims 1 to 6.

9. An electronic device, characterized in that: include: One or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method for detecting lithium deposition in a lithium battery as described in any one of claims 1 to 6.

10. A computer program product comprising computer instructions, characterized in that The computer instructions are executed by the processor to implement the method for detecting lithium deposition in a lithium battery as described in any one of claims 1 to 6.

Images