Impedance spectrum prediction model training method and device, impedance spectrum prediction method and device and medium

Abstract

The invention relates to the technical field of batteries, in particular to an impedance spectrum prediction model training method and device, an impedance spectrum prediction method and device and a medium, and aims to solve the technical problem of how to predict electrochemical impedance spectrum characteristic values which are simpler, lower in cost and free of hardware transformation. In order to achieve the purpose, a preset impedance spectrum prediction model is trained according to a direct-current internal resistance characteristic value sample and a corresponding impedance spectrum characteristic value sample in the pulse process of the battery, and a trained impedance spectrum characteristic model is obtained. And applying the trained impedance spectrum prediction model to predict the impedance spectrum characteristic value of the to-be-predicted battery. According to the method, the mapping relation between the direct-current internal resistance characteristic value and the impedance spectrum characteristic value of the battery is obtained through a machine learning method, complex and expensive testing equipment does not need to be used, and the electrochemical impedance spectrum characteristic value can be predicted more simply at low cost and without hardware transformation.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to an impedance spectrum prediction model training, prediction method, device and medium. Background Art

[0002] Electrochemical impedance spectroscopy (EIS) uses impedances of different frequencies to reflect the characteristics of electrochemical processes. It can reflect the rich kinetic characteristics inside the battery and is often used in the fields of battery state estimation, fault diagnosis and failure analysis. Using equivalent circuit models or relaxation time analysis, the electrochemical impedance spectrum can be unloaded into several characteristic values ​​that can reflect the internal information of the battery according to different frequency responses, such as Ro (ohmic internal resistance), Rct (interface polarization impedance), and Rcp (diffusion impedance). These characteristic values ​​can accurately reflect the electrochemical reactions of the battery in different frequency domains, thereby accurately describing the changes in the mechanism inside the battery, and then predicting the aging of the battery and the potential risk of failure. However, the measurement process of EIS usually relies on expensive instruments that can excite high-frequency sine waves, and the measurement takes a long time. This makes it difficult to directly implement the EIS measurement process of the battery in a real vehicle or energy storage station.

[0003] At present, the most widely used impedance spectrum acquisition method in the laboratory is to sequentially inject sinusoidal current or voltage disturbance signals of different frequencies into the battery, and then process the disturbance and its response to obtain the impedance. This method has the advantage of high measurement accuracy. However, due to the sequential injection of disturbances, especially for very low-frequency impedance, it usually takes a long time to complete the measurement. For example, most commercial electrochemical workstations take 9 minutes to 30 minutes to complete impedance measurements when the frequency range is 1kHz to 0.01Hz and there are ten frequency points in each order of magnitude. Such long measurements are not conducive to on-board applications, maintenance assessments, or rapid screening after retirement of electric vehicles. In addition, the implementation of this method requires a high-precision and universal potential difference dynamic tester / constant current tester to generate sinusoidal disturbances, which is expensive and redundant for applications outside the laboratory. Therefore, for practical application scenarios such as electric vehicles, battery swap stations, and energy storage stations, how to conveniently and efficiently obtain broadband impedance spectra has become a key issue.

[0004] In some existing technologies, in order to obtain the electrochemical impedance spectrum of the battery at a lower cost, the hardware equipment on the application side is often modified to add some hardware equipment to assist in the measurement. However, this will not only increase the hardware cost, but may also affect the operating efficiency of the vehicle or energy storage station.

[0005] Accordingly, the art needs a new solution for predicting characteristic values ​​of electrochemical impedance spectroscopy of batteries to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects, the present application is proposed to solve or at least partially solve the technical problem of how to achieve the prediction of electrochemical impedance spectroscopy characteristic values ​​in a simpler, low-cost and hardware-free manner.

[0007] In a first aspect, a method for training an impedance spectrum prediction model is provided, the method comprising:

[0008] Obtaining a DC internal resistance characteristic value sample of the battery during a pulse process;

[0009] Obtaining an impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample;

[0010] The DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples are used as training data to train a preset impedance spectrum prediction model to obtain the trained impedance spectrum prediction model.

[0011] In a technical solution of the impedance spectrum prediction model training method, the step of obtaining a DC internal resistance characteristic value sample of a battery during a pulse process includes:

[0012] applying a pulse signal to the battery;

[0013] A plurality of DC internal resistance characteristic values ​​of the battery during the pulse signal application process are acquired as the DC internal resistance characteristic value samples.

[0014] In a technical solution of the impedance spectrum prediction model training method, the obtaining of multiple DC internal resistance characteristic values ​​of the battery during the pulse signal application process includes obtaining the DC internal resistance characteristic values ​​according to the following formula;

[0015]

[0016] Among them, D i V is the DC internal resistance characteristic value at the ith second during the pulse signal application process; o is the voltage value at the last moment of the pulse signal application process, I o is the current value at the last moment of the pulse signal application process, V i is the voltage value at the ith second during the pulse signal application process, I i is the current value at the ith second during the application of the pulse signal.

[0017] In a technical solution of the impedance spectrum prediction model training method, the step of obtaining the impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample includes:

[0018] Obtaining an electrochemical impedance spectrum of the battery;

[0019] The electrochemical impedance spectrum is analyzed to obtain a characteristic value sample of the impedance spectrum.

[0020] In a technical solution of the impedance spectrum prediction model training method, the step of obtaining the electrochemical impedance spectrum of the battery includes:

[0021] The battery is tested using a potential difference dynamic tester or a constant current tester to obtain the electrochemical impedance spectrum.

[0022] In a technical solution of the impedance spectrum prediction model training method, the step of analyzing the electrochemical impedance spectrum to obtain the characteristic value sample of the impedance spectrum includes:

[0023] The electrochemical impedance spectrum is analyzed by an equivalent circuit model or a relaxation time distribution algorithm to obtain a sample of characteristic values ​​of the impedance spectrum.

[0024] In a technical solution of the impedance spectrum prediction model training method, obtaining the impedance spectrum characteristic value sample includes:

[0025] At least one of the ohmic internal resistance, interface polarization impedance, and diffusion impedance of the battery is obtained as the impedance characteristic value sample.

[0026] In a technical solution of the above-mentioned impedance spectrum prediction model training method, the impedance spectrum prediction model is obtained by a machine learning method using a random forest model algorithm or a naive Bayes algorithm.

[0027] In a technical solution of the above-mentioned impedance spectrum prediction model training method, the pulse process is a process of removing current or applying current during the battery charging process.

[0028] In a second aspect, a method for predicting characteristic values ​​of an impedance spectrum is provided, the method comprising:

[0029] Obtaining a DC internal resistance characteristic value of the battery to be predicted during a pulse process;

[0030] Based on the trained impedance spectrum prediction model, obtaining the impedance spectrum characteristic value of the battery to be predicted according to the DC internal resistance characteristic value;

[0031] Wherein, the trained impedance spectrum prediction model is obtained by training according to any one of the impedance spectrum prediction model training methods described above.

[0032] In a third aspect, a smart device is provided, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any one of the technical solutions of the above-mentioned impedance spectrum prediction model training method or the method described in the above-mentioned impedance spectrum characteristic value prediction method technical solution is implemented.

[0033] In a fourth aspect, a computer-readable storage medium is provided, which stores a plurality of program codes, wherein the program codes are suitable for being loaded and run by a processor to execute the method described in any one of the technical solutions of the above-mentioned impedance spectrum prediction model training method or the method described in the above-mentioned impedance spectrum characteristic value prediction method technical solution.

[0034] Solution 1. A method for training an impedance spectrum prediction model, characterized in that the method comprises:

[0035] Obtaining a DC internal resistance characteristic value sample of the battery during a pulse process;

[0036] Obtaining an impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample;

[0037] The DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples are used as training data to train a preset impedance spectrum prediction model to obtain the trained impedance spectrum prediction model.

[0038] Solution 2. The impedance spectrum prediction model training method according to Solution 1 is characterized in that:

[0039] The step of obtaining a DC internal resistance characteristic value sample of the battery during a pulse process includes:

[0040] applying a pulse signal to the battery;

[0041] A plurality of DC internal resistance characteristic values ​​of the battery during the pulse signal application process are acquired as the DC internal resistance characteristic value samples.

[0042] Solution 3. The impedance spectrum prediction model training method according to Solution 2 is characterized in that:

[0043] The obtaining of a plurality of DC internal resistance characteristic values ​​of the battery during the pulse signal application process comprises obtaining the DC internal resistance characteristic values ​​according to the following formula;

[0044]

[0045] Among them, D i V is the DC internal resistance characteristic value at the ith second during the pulse signal application process;o is the voltage value at the last moment of the pulse signal application process, I o is the current value at the last moment of the pulse signal application process, V i is the voltage value at the ith second during the pulse signal application process, I i is the current value at the ith second during the application of the pulse signal.

[0046] Solution 4. The impedance spectrum prediction model training method according to Solution 1 is characterized in that:

[0047] The step of obtaining an impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample comprises:

[0048] Obtaining an electrochemical impedance spectrum of the battery;

[0049] The electrochemical impedance spectrum is analyzed to obtain a characteristic value sample of the impedance spectrum.

[0050] Solution 5. The impedance spectrum prediction model training method according to Solution 4 is characterized in that:

[0051] The step of obtaining the electrochemical impedance spectrum of the battery comprises:

[0052] The battery is tested using a potential difference dynamic tester or a constant current tester to obtain the electrochemical impedance spectrum.

[0053] Solution 6. The impedance spectrum prediction model training method according to Solution 4 is characterized in that:

[0054] The step of analyzing the electrochemical impedance spectrum to obtain a sample of characteristic values ​​of the impedance spectrum includes:

[0055] The electrochemical impedance spectrum is analyzed by an equivalent circuit model or a relaxation time distribution algorithm to obtain a sample of characteristic values ​​of the impedance spectrum.

[0056] Scheme 7. The impedance spectrum prediction model training method according to any one of schemes 1 to 6, characterized in that:

[0057] Acquiring the impedance spectrum characteristic value sample, including:

[0058] At least one of the ohmic internal resistance, interface polarization impedance, and diffusion impedance of the battery is obtained as the impedance characteristic value sample.

[0059] Scheme 8. The impedance spectrum prediction model training method according to any one of schemes 1 to 6, characterized in that:

[0060] The impedance spectrum prediction model is obtained by a machine learning method using a random forest algorithm or a naive Bayes algorithm.

[0061] Scheme 9. The impedance spectrum prediction model training method according to any one of schemes 1 to 6, characterized in that:

[0062] The pulse process is a process of removing current or applying current during the battery charging process.

[0063] Solution 10. A method for predicting characteristic values ​​of an impedance spectrum, characterized in that the method comprises:

[0064] Obtaining a DC internal resistance characteristic value of the battery to be predicted during a pulse process;

[0065] Based on the trained impedance spectrum prediction model, obtaining the impedance spectrum characteristic value of the battery to be predicted according to the DC internal resistance characteristic value;

[0066] Wherein, the trained impedance spectrum prediction model is obtained by training according to the impedance spectrum prediction model training method described in any one of Schemes 1 to 9.

[0067] Solution 11. A smart device, comprising:

[0068] at least one processor;

[0069] and, a memory communicatively coupled to the at least one processor;

[0070] Wherein, a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the impedance spectrum prediction model training method described in any one of Schemes 1 to 9 or the impedance spectrum characteristic value prediction method described in Scheme 10 is implemented.

[0071] Scheme 12. A computer-readable storage medium storing a plurality of program codes, characterized in that the program codes are suitable for being loaded and run by a processor to execute the impedance spectrum prediction model training method described in any one of Schemes 1 to 9 or the impedance spectrum characteristic value prediction method described in Scheme 10.

[0072] The above one or more technical solutions of the present application have at least one or more of the following beneficial effects:

[0073] In the technical solution of the impedance spectrum characteristic value prediction method provided by the present application, the present application trains the preset impedance spectrum prediction model according to the DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples during the pulse process of the battery to obtain the trained impedance spectrum characteristic model. The trained impedance spectrum prediction model is used to predict the impedance spectrum characteristic value of the battery to be predicted. Through the above configuration, the present application can achieve the mapping relationship between the DC internal resistance characteristic value and the impedance spectrum characteristic value of the battery through machine learning methods, eliminating the complex and expensive testing equipment costs in the process of predicting the impedance spectrum characteristic value of the battery to be predicted, and at the same time, it can efficiently and accurately establish the relationship between the easily obtained DC internal resistance characteristic value and the impedance spectrum characteristic value, and can achieve a simpler, low-cost and hardware-free electrochemical impedance spectrum characteristic value prediction, which provides information at the electrochemical mechanism level for battery fault diagnosis and failure analysis, and greatly improves the safety of battery application. In the context of vigorously developing electric vehicles and carbon neutrality, it has great application prospects and value. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The disclosure of the present application will become easier to understand with reference to the accompanying drawings. It is easy for those skilled in the art to understand that these drawings are only for illustrative purposes and are not intended to limit the scope of protection of the present application. Among them:

[0075] Figure 1 It is a schematic flow chart of the main steps of the impedance spectrum prediction model training method according to an embodiment of the present application;

[0076] Figure 2 It is a schematic flow chart of the main steps of a method for predicting characteristic values ​​of impedance spectrum according to an embodiment of the present application;

[0077] Figure 3 It is a flowchart of the main steps of a method for predicting characteristic values ​​of an impedance spectrum according to an implementation of an embodiment of the present application;

[0078] Figure 4 is a schematic diagram of the process of voltage changing with time during a pulse process according to an implementation of an embodiment of the present application;

[0079] Figure 5 It is a schematic diagram of the impedance spectrum characteristic value result obtained by DRT analysis according to one implementation of an embodiment of the present application;

[0080] Figure 6 It is a schematic diagram of the main structure of a smart device according to an embodiment of the present application.

[0081] Reference numerals:

[0082] 11: memory; 12: processor. DETAILED DESCRIPTION

[0083] Some embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application and are not intended to limit the protection scope of the present application.

[0084] In the description of the present application, "module" and "processor" may include hardware, software or a combination of the two. A module may include hardware circuits, various suitable sensors, communication ports, memory, and may also include software parts, such as program code, or a combination of software and hardware. The processor may be a central processing unit, a microprocessor, an image processor, a digital signal processor or any other suitable processor. The processor has data and / or signal processing functions. The processor may be implemented in software, hardware or a combination of the two. Computer-readable storage media include any suitable medium that can store program code, such as a disk, a hard disk, an optical disk, a flash memory, a read-only memory, a random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as only A, only B or A and B. The term "at least one A or B" or "at least one of A and B" has a similar meaning to "A and / or B" and may include only A, only B or A and B. The singular terms "one" and "the" may also include plural forms.

[0085] The relevant user personal information that may be involved in the various embodiments of this application is strictly in accordance with the requirements of laws and regulations, following the principles of legality, legitimacy and necessity, based on the reasonable purposes of business scenarios, to process the personal information that users actively provide during the use of products / services or generated due to the use of products / services, as well as the personal information obtained with the user's authorization.

[0086] The user personal information processed by this application will vary depending on the specific product / service scenario, and shall be based on the specific scenario in which the user uses the product / service, and may involve the user's account information, device information, driving information, vehicle information or other related information. This application will treat the user's personal information and its processing with a high degree of diligence.

[0087] This application attaches great importance to the security of user personal information and has taken reasonable and feasible security protection measures that meet industry standards to protect user information and prevent personal information from being accessed, disclosed, used, modified, damaged or lost without authorization.

[0088] Here we first explain some terms involved in this application.

[0089] Electrochemical Impedance Spectroscopy (EIS), when the electrode system is disturbed by an AC signal of a sinusoidal voltage (current), a corresponding current (voltage) response signal will be generated, from which the impedance or admittance of the electrode can be obtained. The impedance spectrum generated by a series of sinusoidal wave signals is called electrochemical impedance spectroscopy.

[0090] The characteristic values ​​of the impedance spectrum are the characteristic values ​​that reflect the electrochemical reactions of the battery in different frequency domains, thereby accurately describing the mechanism changes inside the battery. They may include Ro (ohmic internal resistance), Rct (interface polarization impedance), Rcp (diffusion impedance), etc.

[0091] DC internal resistance is an important performance indicator of lithium batteries, representing the dynamic internal resistance value of the battery under different charging and discharging conditions.

[0092] The Distribution of Relaxation Times (DRT) is an electrochemical impedance spectroscopy (EIS) analysis technique that does not rely on prior knowledge of the research object and is used to separate and analyze highly overlapping physical and chemical processes in EIS. Relaxation time refers to the average time interval for a system to reach another equilibrium state from one equilibrium state, and the time required for a system to tend to a certain steady state from a transient state.

[0093] See attached Figure 1 , Figure 1 FIG. 1 is a flow chart of the main steps of the impedance spectrum prediction model training method according to an embodiment of the present application. Figure 1 As shown, the impedance spectrum prediction model training method in the embodiment of the present application mainly includes the following steps S101 to S103.

[0094] Step S101: obtaining a DC internal resistance characteristic value sample of the battery during a pulse process, wherein the pulse process refers to a process of applying a pulse signal.

[0095] In this embodiment, a certain pulse signal may be applied to the battery, and a DC internal resistance characteristic value sample of the battery during the process of applying the pulse signal may be obtained.

[0096] In one embodiment, during the process of collecting the DC internal resistance characteristic value samples of the battery, the DC internal resistance characteristic value samples may be collected multiple times at preset temperature intervals. For example, a certain pulse signal is applied to the battery at every 5°C interval, and the DC internal resistance characteristic value samples of the battery are obtained during each application of the pulse signal, so as to obtain more sufficient DC internal resistance characteristic value samples.

[0097] In one embodiment, the pulse process may be a pulse signal formed by removing or applying current during the battery charging process. The magnitude of the current may include but is not limited to 0C to 4C, and the pulse time of the pulse signal may include but is not limited to 10-200s.

[0098] In one implementation, a certain pulse signal may be applied to the battery, and a plurality of DC internal resistance characteristic values ​​of the battery during the process of applying the pulse signal are obtained as DC internal resistance characteristic value samples.

[0099] Specifically, the DC internal resistance characteristic value can be obtained according to the following formula (1):

[0100]

[0101] Among them, D i is the DC internal resistance characteristic value at the ith second during the pulse signal application process; V o is the voltage value at the last moment of the pulse signal application process, I o is the current value at the last moment of the pulse signal application process, V i is the voltage value at the i-th second during the pulse signal application process, I i is the current value at the i-th second during the pulse signal application process.

[0102] The number of DC internal resistance characteristic values ​​obtained during each pulse signal application process may be 3-20, and those skilled in the art may determine the number of DC internal resistance characteristic values ​​according to the needs of practical applications and the actual requirements of the impedance spectrum prediction model.

[0103] Step S102: obtaining a battery impedance spectrum characteristic value sample corresponding to the DC internal resistance characteristic value sample.

[0104] In this embodiment, the impedance spectrum characteristic value samples of the battery corresponding to the DC internal resistance characteristic value samples may be obtained.

[0105] In one implementation, the electrochemical impedance spectrum of the battery may be tested first, and the tested electrochemical impedance spectrum may be analyzed to obtain an impedance spectrum characteristic value sample.

[0106] In one embodiment, a potential difference dynamic tester or a constant current tester may be used to test the battery to obtain an electrochemical impedance spectrum.

[0107] In one embodiment, the impedance spectrum characteristic value may include the ohmic internal resistance R0, the interface polarization impedance Rct, and the diffusion impedance Rcp of the battery. The ohmic internal resistance, the interface polarization impedance, and the diffusion impedance of the battery may be obtained as impedance spectrum characteristic value samples.

[0108] In one embodiment, the electrochemical impedance spectrum can be analyzed by an equivalent circuit model or a relaxation time distribution algorithm to obtain a sample of characteristic values ​​of the impedance spectrum. Among them, the analysis of the electrochemical impedance spectrum based on the equivalent circuit model is well known to those skilled in the art and will not be repeated in this application.

[0109] The following is an exemplary description of the process of analyzing the electrochemical impedance spectrum using the relaxation time distribution algorithm:

[0110] The DRT analytical formula is as follows (2):

[0111]

[0112] Where Z(f) is the DRT analytical impedance data, R0 is the ohmic resistance, f is the frequency, τ is the relaxation time, C in is the capacitance, and γ(τ) is the DRT function.

[0113] The above formula (2) is a typical first-order Fredholm anti-integral equation, which can be calculated using the Tikhonov regularization method. Through the measured analytical impedance data Z(f), the DRT function about the characteristic time constant can be obtained. The polarization resistance can be obtained by the following formula (3):

[0114]

[0115] Among them, R p is the polarization resistance.

[0116] Step S103: using the DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples as training data, training a preset impedance spectrum prediction model to obtain a trained impedance spectrum prediction model.

[0117] In this embodiment, the DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples may be used as training data to train a preset impedance spectrum prediction model to obtain a trained impedance spectrum prediction model.

[0118] In one embodiment, the impedance spectrum prediction model can be obtained by a machine learning method using a random forest algorithm.

[0119] In another embodiment, the impedance spectrum prediction model can be obtained by a machine learning method using a naive Bayes algorithm.

[0120] In some other embodiments, the impedance spectrum prediction model can be obtained by using other classification algorithms through machine learning methods.

[0121] In one implementation, a database may be established based on the DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples, and the impedance spectrum prediction model may be trained based on the database to obtain the trained impedance spectrum prediction model.

[0122] Based on the method described in steps S101 to S103 above, the embodiment of the present application trains the preset impedance spectrum prediction model according to the DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples during the pulse process of the battery to obtain the trained impedance spectrum characteristic model. The trained impedance spectrum prediction model is used to predict the impedance spectrum characteristic value of the battery to be predicted. Through the above configuration, the embodiment of the present application can achieve the mapping relationship between the DC internal resistance characteristic value and the impedance spectrum characteristic value of the battery through machine learning methods, eliminating the complex and expensive test equipment costs in the process of predicting the impedance spectrum characteristic value of the battery to be predicted, and at the same time can efficiently and accurately establish the relationship between the easily obtained DC internal resistance characteristic value and the impedance spectrum characteristic value, and can achieve a simpler, low-cost and hardware-free prediction of the electrochemical impedance spectrum characteristic value, which provides information at the electrochemical mechanism level for battery fault diagnosis and failure analysis, and greatly improves the safety of battery application.

[0123] Furthermore, the present application also provides a method for predicting characteristic values ​​of an impedance spectrum.

[0124] See attached Figure 2 , Figure 2 FIG. 1 is a flow chart of the main steps of a method for predicting characteristic values ​​of an impedance spectrum according to an embodiment of the present application. Figure 2 As shown, the impedance spectrum characteristic value prediction method in the embodiment of the present application mainly includes the following steps S201 to S202:

[0125] Step S201: obtaining a DC internal resistance characteristic value of a battery to be predicted during a pulse process.

[0126] In this embodiment, the DC internal resistance characteristic value of the battery to be predicted during the pulse process may be obtained.

[0127] In one implementation, the DC internal resistance characteristic value may be obtained according to formula (1) in the aforementioned embodiment.

[0128] Step S202: Based on the trained impedance spectrum prediction model, the impedance spectrum characteristic value of the battery to be predicted is obtained according to the DC internal resistance characteristic value; wherein the trained impedance spectrum prediction model is obtained by training according to the impedance spectrum prediction model training method in the above impedance spectrum prediction model training method embodiment.

[0129] In this embodiment, the trained impedance spectrum prediction model obtained by training in the above impedance spectrum prediction model training method embodiment can be applied, and based on the trained impedance spectrum prediction model, the impedance spectrum characteristic value of the battery to be predicted is predicted according to the DC internal resistance characteristic value.

[0130] In one embodiment, the characteristic values ​​of the impedance spectrum may include the ohmic internal resistance R0, the interface polarization impedance Rct, and the diffusion impedance Rcp of the battery.

[0131] In one embodiment, please refer to the attached Figure 3 , Figure 3 FIG. 1 is a flowchart of the main steps of a method for predicting characteristic values ​​of an impedance spectrum according to an embodiment of the present application. Figure 3 As shown, DCR feature extraction can be performed based on the DCR test data during the battery pulse process to obtain multiple DC internal resistance characteristic value samples such as D1, D2, D5, D10, D40, and D80; and according to the impedance spectrum test of the battery, the impedance spectrum characteristic value samples of the battery are extracted, including R0, Rct1, Rct2, and Rcp. According to the DC internal resistance characteristic value samples and the impedance spectrum characteristic value samples, the machine learning model is trained to obtain the trained impedance spectrum prediction model. After testing the actual vehicle pulse DCR data (that is, the DC internal resistance data during the pulse process of the battery to be predicted), DCR feature extraction is performed to obtain multiple DC internal resistance characteristic values ​​such as D1, D2, D5, D10, D40, and D80. The DC internal resistance characteristic values ​​are input into the trained impedance spectrum prediction model to obtain the impedance spectrum characteristic values ​​R0, Rct1, Rct2, and Rcp of the battery to be predicted. Based on these impedance spectrum characteristic values, the impedance characteristic abnormality diagnosis of the battery to be predicted is performed.

[0132] Based on the method described in steps S201 to S202 above, the embodiment of the present application applies the trained impedance spectrum prediction model to predict the impedance spectrum characteristic value of the battery to be predicted. Through the above configuration, the embodiment of the present application can achieve the mapping relationship between the DC internal resistance characteristic value and the impedance spectrum characteristic value of the battery through machine learning, eliminating the complex and expensive testing equipment costs in the process of predicting the impedance spectrum characteristic value of the battery to be predicted. At the same time, it can efficiently and accurately establish the relationship between the easily obtained DC internal resistance characteristic value and the impedance spectrum characteristic value, and can achieve a simpler, low-cost and hardware-free electrochemical impedance spectrum characteristic value prediction, which provides information at the electrochemical mechanism level for battery fault diagnosis and failure analysis, and greatly improves the safety of battery application. In the context of vigorously developing electric vehicles and carbon neutrality, it has great application prospects and value.

[0133] In one example, a 100Ah ternary lithium battery can be used. During the charging process, a pulse signal is applied to the lithium battery to remove the current, the voltage after the current is removed is recorded, and the DC internal resistance characteristic value is calculated. The schematic diagram of the process of voltage changing with time during the pulse process is as follows: Figure 4 As shown, Figure 4 The horizontal axis is time and the vertical axis is voltage. The calculated DC internal resistance characteristic values ​​are shown in Table 1:

[0134] Table 1 DC internal resistance characteristic values D1 D2 D5 D10 D20 D40 D80 0.00057 0.000643 0.000708 0.000748 0.000817 0.000864 0.000995

[0135] The polarization resistance calculated by the above formula (2) and formula (3) can be referred to in the attached Figure 5 , Figure 5 It is a schematic diagram of the impedance spectrum characteristic value results obtained by DRT analysis according to one implementation of an example of the present application. Figure 5 The distribution results of the impedance spectrum characteristic values ​​are shown in FIG. 2 , wherein the impedance spectrum characteristic values ​​include R0, Rct1, Rct2, Rct3, and Rcp. The characteristic values ​​of each impedance spectrum are shown in Table 2:

[0136] Table 2 Impedance spectrum characteristic values R0 Rct1 Rct2 Rct3 Rcp 0.000045 0.000044 0.000010 0.000033 0.000288

[0137] A correspondence is established between the multiple DC internal resistance characteristic values ​​of the lithium battery and the corresponding impedance spectrum characteristic values, and a database is established. According to the database, a preset impedance spectrum prediction model is trained using a machine learning method to obtain a trained impedance spectrum prediction model. The trained impedance spectrum prediction model can be used to predict the impedance spectrum characteristic values ​​of the same lithium battery.

[0138] It should be pointed out that although the various steps in the above embodiments are described in a specific order, those skilled in the art can understand that in order to achieve the effect of the present application, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders. These adjusted schemes are equivalent to the technical schemes described in this application, and therefore will also fall within the scope of protection of this application.

[0139] It is understood by those skilled in the art that all or part of the processes in the method for implementing the above-mentioned embodiment of the present application can also be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device, medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium, etc. that can carry the computer program code.

[0140] Another aspect of the present application also provides a computer-readable storage medium.

[0141] In an embodiment of a computer-readable storage medium according to the present application, the computer-readable storage medium may be configured to store a program for executing the above method embodiment, which may be loaded and run by a processor to implement the method in the above method embodiment. For ease of explanation, only the parts related to the embodiment of the present application are shown. For specific technical details not disclosed, please refer to the method section of the embodiment of the present application. The computer-readable storage medium may be a storage device formed by various electronic devices. Optionally, the computer-readable storage medium in the embodiment of the present application is a non-temporary computer-readable storage medium.

[0142] Another aspect of the present application also provides a smart device.

[0143] In an embodiment of an intelligent device according to the present application, the intelligent device may include at least one processor; and a memory connected to the at least one processor; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any of the above embodiments is implemented. The intelligent device described in the present application may include a driving device, a smart car, a robot, an electronic device, and the like. Figure 6 , Figure 6 FIG. 4 exemplarily shows that the memory 11 and the processor 12 are communicatively connected via a bus.

[0144] In some embodiments of the present application, the smart device may further include at least one sensor, which is used to sense information. The sensor is connected to any type of processor mentioned in the present application. Optionally, the smart device may also include an autonomous driving system, which is used to guide the smart device to drive itself or assist driving. The processor communicates with the sensor and / or the autonomous driving system to complete the method described in any of the above embodiments. The electronic device described in the present application may be, but is not limited to, a mobile phone, a tablet computer, a desktop, a laptop, a handheld computer, a notebook computer, a vehicle-mounted device, an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (PDA), an augmented reality (AR)\virtual reality (VR) device, etc., and the embodiments of the present application are not limited to this.

[0145] So far, the technical solution of the present application has been described in conjunction with an embodiment shown in the accompanying drawings, but it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments. Without departing from the principles of the present application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present application.

Claims

1. A method for training an impedance spectrum prediction model, characterized in that: The method comprises: Obtaining a DC internal resistance characteristic value sample of the battery during a pulse process; Obtaining an impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample; The DC internal resistance characteristic value samples and the corresponding impedance spectrum characteristic value samples are used as training data to train a preset impedance spectrum prediction model to obtain the trained impedance spectrum prediction model.

2. The impedance spectrum prediction model training method according to claim 1, characterized in that: The step of obtaining a DC internal resistance characteristic value sample of the battery during a pulse process includes: applying a pulse signal to the battery; A plurality of DC internal resistance characteristic values ​​of the battery during the pulse signal application process are acquired as the DC internal resistance characteristic value samples.

3. The impedance spectrum prediction model training method according to claim 2, characterized in that: The obtaining of a plurality of DC internal resistance characteristic values ​​of the battery during the pulse signal application process comprises obtaining the DC internal resistance characteristic values ​​according to the following formula; Among them, D i V is the DC internal resistance characteristic value at the ith second during the pulse signal application process; o is the voltage value at the last moment of the pulse signal application process, I o is the current value at the last moment of the pulse signal application process, V i is the voltage value at the ith second during the pulse signal application process, I i is the current value at the ith second during the application of the pulse signal.

4. The impedance spectrum prediction model training method according to claim 1, characterized in that: The step of obtaining an impedance spectrum characteristic value sample of the battery corresponding to the DC internal resistance characteristic value sample comprises: Obtaining an electrochemical impedance spectrum of the battery; The electrochemical impedance spectrum is analyzed to obtain a characteristic value sample of the impedance spectrum.

5. The impedance spectrum prediction model training method according to claim 4, characterized in that: The step of obtaining the electrochemical impedance spectrum of the battery comprises: The battery is tested using a potential difference dynamic tester or a constant current tester to obtain the electrochemical impedance spectrum.

6. The impedance spectrum prediction model training method according to claim 4, characterized in that: The step of analyzing the electrochemical impedance spectrum to obtain a sample of characteristic values ​​of the impedance spectrum includes: The electrochemical impedance spectrum is analyzed by an equivalent circuit model or a relaxation time distribution algorithm to obtain a sample of characteristic values ​​of the impedance spectrum.

7. The impedance spectrum prediction model training method according to any one of claims 1 to 6, characterized in that: Acquiring the impedance spectrum characteristic value sample, comprising: At least one of the ohmic internal resistance, interface polarization impedance, and diffusion impedance of the battery is obtained as the impedance characteristic value sample.

8. The impedance spectrum prediction model training method according to any one of claims 1 to 6, characterized in that: The impedance spectrum prediction model is obtained by a machine learning method using a random forest algorithm or a naive Bayes algorithm.

9. The impedance spectrum prediction model training method according to any one of claims 1 to 6, characterized in that: The pulse process is a process of removing current or applying current during the battery charging process.

10. A method for predicting characteristic values ​​of an impedance spectrum, characterized in that: The method comprises: Obtaining a DC internal resistance characteristic value of the battery to be predicted during a pulse process; Based on the trained impedance spectrum prediction model, obtaining the impedance spectrum characteristic value of the battery to be predicted according to the DC internal resistance characteristic value; Wherein, the trained impedance spectrum prediction model is obtained by training according to the impedance spectrum prediction model training method according to any one of claims 1 to 9.

Images