Method and device for estimating state of health of power battery, electronic equipment and storage medium

Abstract

The application relates to a SOH estimation method and device of a power battery, an electronic device and a storage medium. The method comprises the following steps: obtaining a first SOH according to data from a starting moment to a current moment of a charging and discharging process of the power battery; obtaining a second SOH according to data of a complete charging and discharging process of the power battery; and obtaining a SOH of the power battery at the current moment according to the first SOH and the second SOH.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of batteries, in particular to a SOH estimation method and device of a power battery, an electronic device and a storage medium. BACKGROUND

[0002] In the process of calculating SOH, when using a static OCV-SOC characteristic curve to calculate SOH, there is a problem of large calculation error; when using a dynamic OCV-SOC characteristic curve to calculate SOH, a large amount of experimental resources and time cost are wasted. In addition, in the prior art, it is impossible to monitor the dynamic SOH of the power battery in real time during the charging and discharging process of the power battery.

[0003] In the process of calculating SOH, when using a static OCV-SOC characteristic curve to calculate SOH, there is a problem of large calculation error; when using a dynamic OCV-SOC characteristic curve to calculate SOH, a large amount of experimental resources and time cost are wasted. In addition, in the prior art, it is impossible to monitor the dynamic SOH of the power battery in real time during the charging and discharging process of the power battery. SUMMARY

[0004] In view of the above problems, the present application provides a SOH estimation method and device of a power battery, an electronic device and a storage medium, which can solve the problem of monitoring the dynamic SOH of the power battery in real time during the charging and discharging process of the power battery.

[0005] In the first aspect, the present application provides a SOH estimation method of a power battery, comprising: obtaining a first SOH according to data from a starting time to a current time of a charging and discharging process of the power battery; obtaining a second SOH according to data of a complete charging and discharging process of the power battery; and obtaining a SOH of the power battery at the current time according to the first SOH and the second SOH.

[0006] In the technical scheme of the present application, the dynamic SOH predicted in the charging and discharging process of the power battery is combined with the static SOH estimated in the complete charging and discharging process of the power battery, which not only realizes the effect of monitoring the dynamic SOH of the power battery in real time during the charging and discharging process of the power battery, but also solves the problem of large deviation caused by using a static OCV-SOC characteristic curve to estimate a dynamic OCV-SOC. This is helpful for real-time understanding of the health status of the power battery, so that the power battery can be disposed in time in case of safety problems.

[0007] In the second aspect, the present application also provides a SOH estimation device of a power battery, characterized in that,

[0008] comprising a first calculation module, a second calculation module and a fusion module. The first calculation module is configured to obtain a first SOH according to data from a starting time to a current time of a charging and discharging process of the power battery; the second calculation module is configured to obtain a second SOH according to data of a complete charging and discharging process of the power battery; and the fusion module is configured to obtain a SOH of the power battery at the current time according to the first SOH and the second SOH.

[0009] In a third aspect, the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor implements the steps of any of the above methods when executing the program.

[0010] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the computer program, when executed by a processor, implements the steps of any of the above methods.

[0011] In a fifth aspect, the present application provides a computer program product, characterized in that, comprising computer readable code, or a non-volatile computer readable storage medium carrying computer readable code, when the computer readable code runs in an electronic device, the processor in the electronic device executes the steps of any of the above methods.

[0012] The above description is only a summary of the technical solutions of the present application. In order to enable one of ordinary skill in the art to better understand the technical means of the present application and implement it according to the content of the description, and in order to enable the above and other purposes, characteristics and advantages of the present application to be more apparent and easy to understand, the following specific embodiments of the present application are described in detail. BRIEF DESCRIPTION OF DRAWINGS

[0013] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments, and are not meant to limit the present application. Moreover, the same reference numerals in all the drawings represent the same or similar elements. In the drawings:

[0014] Figure 1 Flowchart of the SOH estimation method of the power battery in an embodiment of the present application;

[0015] Figure 2 Flowchart of the first sample acquisition of the cell charge and discharge data in an embodiment of the present application;

[0016] Figure 3 Schematic diagram of the change of the voltage of the power battery with time in the charging process of the power battery in an embodiment of the present application;

[0017] Figure 4 Schematic diagram of the change of the current of the power battery with time in the charging process of the power battery in an embodiment of the present application;

[0018] Figure 5 Flowchart of the preset generation network training in an embodiment of the present application;

[0019] Figure 6 Schematic diagram of the overall network framework of the generative adversarial network in an embodiment of the present application;

[0020] Figure 7 A schematic diagram of a generative adversarial network training process in an embodiment of the present application is shown in FIG. 1.

[0021] Figure 8 A schematic diagram of a module of a SOH estimation device for a power battery in an embodiment of the present application is shown in FIG. 2. DETAILED DESCRIPTION

[0022] The embodiments of the technical solutions of the present application will be described in detail below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of this application; the description and the claims herein and the above description of drawings, the term "comprising" and any variations thereof, is intended to cover not exclusively include.

[0024] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0025] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor is it necessarily mutually exclusive or alternative to other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0026] The estimation of the state of health (SOH) of a battery currently generally uses a two-point calculation method, which is based on the initial voltage and the state of charge (SOC) corresponding to the cut-off voltage, and converts the corresponding SOH based on the corresponding SOC. The specific process is as follows:

[0027] The battery factory open circuit voltage (OCV) and SOC curve, i.e. the OCV-SOC curve, has the SOC as the horizontal axis and the OCV as the vertical axis. During the charging and discharging process of the battery, let the initial SOC be SOC0, and the corresponding open circuit voltage be V0. The terminal SOC is SOC1, and the corresponding open circuit voltage is V1. Let the initial current be I0, and the terminal current be I1.I1 The factory capacity of the battery is Q0, and the current capacity is Q now .

[0028] Taking the charging process as an example, the entire process lasts for At, and the SOC change of the battery is recorded as ASOC, then:

[0029] ASOC = |SOC1-SOC0|.

[0030] The ampere-hour integral of the entire charging and discharging process is recorded as Ah, then:

[0031] Ah = IAt.

[0032] In summary, the formula of SOH is:

[0033] SOH = Ah / ASOC*1 / Q0 = Q now / Q0.

[0034] The two-point calculation method relies heavily on the OCV-SOC characteristic data. The battery or battery pack decays during actual operation. When using the factory calibrated OCV-SOC characteristic curve of the power battery (i.e. static OCV-SOC characteristic curve) to calculate SOH, there is a large error. In addition, the OCV-SOC characteristic curve recorded under actual working conditions (i.e. dynamic OCV-SOC characteristic curve) is measured by the actual working state and condition of the battery or battery pack in its entire life cycle, which requires a large amount of experimental data as support. Since the measured SOC needs to be measured and calculated in the specified time of the battery, the calculation efficiency is very low, and the battery cannot be calculated during the charging and discharging process, so the use scenario is very limited.

[0035] In the process of calculating SOH, using static OCV-SOC characteristic curve to calculate SOH has a large calculation error; using dynamic OCV-SOC characteristic curve to calculate SOH wastes a lot of experimental resources and time cost, etc.

[0036] Based on the research of the inventor, a generative adversarial network algorithm model is created, and the running data of different types of batteries in different decay stages is used to train the generative adversarial network model to generate OCV-SOC characteristic data of each battery in its entire life cycle, form a dynamic OCV-SOC characteristic, and create a dynamic OCV-SOC algorithm model for fitting dynamic OCV-SOC characteristic data.

[0037] In addition, a recurrent neural network algorithm model is created to extract features from the power battery charging and discharging process data as a SOH adaptive calculation model. It is used to calculate the SOH value in the charging and discharging process.

[0038] The SOH is confirmed by fusing the SOH calculation, and the overall SOH calculation scheme is divided into two sections: in the first section, the SOH adaptive calculation model is used in the charging and discharging process; in the second section, after the charging and discharging process stops, the SOH is calculated according to the two-point method using the dynamic OCV-SOC model. By setting the difference threshold of the SOH calculation results of the above first section and second section, the SOH is weighted to obtain the final SOH value.

[0039] The SOH estimation method of the power battery disclosed in the embodiments of the present application can be used to estimate the SOH of the battery of various electronic devices. The electronic device can be, but is not limited to, a vehicle, a mobile phone, a tablet, a notebook computer, an electric toy, an electric tool, an electric bicycle, an electric motorcycle, an electric vehicle, a ship, a spacecraft, etc.

[0040] For the convenience of description, the embodiments of the present application are described by taking the application in a vehicle as an example.

[0041] In one embodiment, Figure 1 The flowchart of the SOH estimation method of the power battery in an embodiment of the present application is shown. It should be understood that, Figure 1 The execution subject of the SOH estimation method of the power battery can be an electronic device, or a chip or processor in the electronic device. Exemplarily, the execution subject of the SOH estimation method of the power battery is an electronic device. Figure 1 The execution subject of the SOH estimation method of the power battery is an electronic device, which is described in detail as follows. As shown in Figure 1 The method comprises the following steps:

[0042] Step 110: obtaining a first SOH according to the data from the starting time to the current time of the charging and discharging process of the power battery;

[0043] Step 120: obtaining a second SOH according to the data of the complete charging and discharging process of the power battery;

[0044] Step 130: obtaining the SOH of the power battery at the current time according to the first SOH and the second SOH.

[0045] The first SOH is calculated using the SOH adaptive calculation model in the charging and discharging process. The complete charging and discharging process of the power battery can be that after the charging and discharging of the power battery stops, the SOC is calculated using the dynamic OCV-SOC model, and then the second SOH is calculated according to the two-point method.

[0046] The SOH of the power battery at the current time in the charging and discharging process is obtained by fusing the calculation results of the dynamic OCV-SOC feature and the static OCV-SOC feature.

[0047] In some embodiments, step 110 obtains the first SOH according to the data from the starting time to the current time of the charging and discharging process of the power battery, comprising:

[0048] Step 210: Obtain the first data from the start time of power battery charging to the current time;

[0049] Step 220: Use the preset first model to predict the first data to obtain the first SOH.

[0050] The first preset model is a deep learning model. Using a deep learning model to predict the data of the dynamic charging and discharging process of the power battery can ensure the accuracy of the prediction results.

[0051] In one embodiment, the first model is trained as follows:

[0052] Step 310: Obtain the first sample of cell charge and discharge data, wherein the label of the first sample is the SOH value of different charge and discharge data;

[0053] Step 320: Train the preset recurrent neural network model based on the first sample to obtain the first model.

[0054] The preset recurrent neural networks include, but are not limited to, bidirectional encoder representations from transformers (BERT), recurrent neural networks (RNN), long short-term memory networks (LSTM), gated recurrent units (GRU), and convolutional neural networks (CNN). By using variable-length charge-discharge cycle data for training, the first trained model can adaptively process the data of the power battery in the charge-discharge state, and has good robustness.

[0055] In one embodiment, such as Figure 2 As shown, Figure 2 This is a schematic diagram illustrating the first sample process for obtaining battery cell charge / discharge data in one embodiment of this application. It should be understood that... Figure 2 The executing entity can be an electronic device, or a chip or processor within that device. For example, using... Figure 2 The following is a detailed explanation using electronic devices as an example of the executing entity.

[0056] Obtain the first sample of battery cell charge / discharge data, including:

[0057] Step 411: Obtain the voltage-time curve and current-time curve for cell charging and discharging;

[0058] Step 420: Within a preset time range, intercept at least one voltage segment and at least one current segment from the voltage-time curve and the current-time curve;

[0059] Step 430: Obtain the SOH value corresponding to each voltage segment and each current segment, and confirm the SOH value as the tag information;

[0060] Step 440: Obtain a first sample based on at least one voltage segment, at least one current segment, and the SOH value.

[0061] Exemplarily, as Figure 3 and Figure 4 shown, Figure 3 is a schematic diagram of the change of the voltage of the power battery over time during the charging process in an embodiment of the present application, Figure 4 is a schematic diagram of the change of the current of the power battery over time during the charging process in an embodiment of the present application, t1 is the start time of charging, and t2 is the end time of charging.

[0062] The above first sample can be obtained by intercepting segments on Figure 3 and Figure 4 When intercepting, the intercept start point is set as S1, and the intercept end point is set as S2, keeping t1 = <S1 < S2 < t2, that is, the S1 point can be at the t1 point or between t1 and t2, and the S2 point can be between t1 and t2. By randomly intercepting a sufficient number of voltage and current data of different lengths and at different temperatures between S1 and S2, and using the SOH value of the corresponding voltage and current as its tag to form the first sample. This can ensure that the intercepted sample can reflect the dynamic process of the battery, making the first model trained have good adaptability.

[0063] In one embodiment, Step 220: Obtain a second SOH according to the data of the complete charge and discharge process of the power battery, including:

[0064] Step​​​​​​​​​

[0068] In one embodiment, step 530: obtaining all second SOHs according to the SOC, the starting voltage and the cut-off voltage of the power battery charging and discharging, comprises:

[0069] According to the SOC, the starting voltage and the cut-off voltage of the power battery charging and discharging, the charge and discharge ampere-hour integral of the power battery is calculated to obtain the second SOH.

[0070] It should be noted that the method for calculating the second SOH is not limited thereto, and methods such as joint estimation, collaborative estimation, fusion, empirical / fitting, optimization algorithm, sample entropy, etc. can also be used.

[0071] In one embodiment, the second model is obtained by training in the following manner:

[0072] Step 610: obtaining a second sample of complete charging and discharging data of the battery cell, wherein the second sample comprises complete charging and discharging data of different models of battery cells under at least one temperature and at least one current rate combination;

[0073] Step 620: training a preset integrated algorithm model according to the second sample to obtain a second model.

[0074] The integrated algorithm model includes but is not limited to gradient boosting algorithm (Categorical Boosting, CatBoost), extreme gradient boosting algorithm (eXtreme Gradient Boosting, XGBoost), distributed gradient boosting framework based on decision tree algorithm (Light Gradient Boosting Machine, lightGBM), random forest (Random Forest, RF) model, etc. Through any of the above integrated algorithm models, the multiple OCV-SOC feature data obtained in step 610 are modeled, and the OCV-SOC feature model is trained and saved. That is, the second model.

[0075] In one embodiment, in order to solve the problem that the measured SOC in the actual process needs to be measured and calculated within a specified time of the battery cell, which is very low in measurement efficiency, a small amount of real data is used to generate sample data based on a generative adversarial network, which can ensure the training of the model. For example, Figure 5 As shown in Figure 5 is a preset generative network training process diagram in an embodiment of the present application. It should be understood that Figure 5 The execution subject of Figure 5 is an electronic device, or a chip or processor in the electronic device. Exemplarily, the execution subject of is an electronic device, which is described in detail.

[0076] The preset generation network is trained by the following method:

[0077] Step 710: inputting a random noise sample into the preset generation network to obtain a fake sample;

[0078] Step 720: inputting the real sample and the fake sample into the preset discriminator network to obtain a prediction result;

[0079] Step 730: updating the preset generation network and the preset discriminator network based on the prediction result until the discriminator network cannot distinguish the authenticity of the fake sample generated by the preset generation network, and obtaining the preset generation network.

[0080] As shown in Figure 6 , Figure 6 is a schematic diagram of a whole network framework of a generative adversarial network in an embodiment of the present application. The generative adversarial network (GAN) mainly includes two parts: a preset generation network (Generator Network, GN) denoted as a G model, and a discriminator network (Discriminator Network, DN) denoted as a D model. The G model and the D model can be, but are not limited to, a deep neural network (Deep Neural Networks, DNN), an RNN, a Bert, a CNN, a fully connected (Fully Connected, FC) model, and the like.

[0081] The GAN mainly has two inputs: real data (Real Sample, RS) denoted as RS data, and random noise (Random Noise, RN) denoted as RN data. The RS data is mainly real data, that is, a real sample. The running data of a small amount of different models of battery cells at different attenuation stages and different temperatures is obtained, wherein the running data includes the temperature, voltage, and current of the battery cell, and the static OCV-SOC curve data converted therefrom. The RN data is mainly randomly generated data, that is, a fake sample data generated by the G model.

[0082] As shown in Figure 7 , Figure 7 is a schematic diagram of a training process of a generative adversarial network in an embodiment of the present application. It should be understood that Figure 7 the execution subject can be an electronic device, or a chip or a processor in the electronic device. Exemplarily, the execution subject of Figure 7 is taken as an electronic device for detailed description.

[0083] Figure 7The training process of the generative adversarial network is generated. The RN data is input into the preset generation network to generate fake sample data (Fake Resample, FR), and the discriminant network judges the fake sample generated by the preset generation network and the RS sample. The discriminant network needs to distinguish the true and false nature of the fake sample and the RS as much as possible, and the preset generation network needs to generate realistic samples to deceive the discriminant network. The GAN model is trained in a loop and finally converges stably.

[0084] In one embodiment, step 510: obtaining a second sample of complete charge and discharge data of the battery cell, comprises:

[0085] Step 810: obtaining a random noise sample;

[0086] Step 820: inputting the random noise sample into the preset generation network to obtain the second sample.

[0087] When generating a sample, the discriminant network needs to be separated from the GAN network, and the preset generation network is used to generate a sample data from the random noise to obtain the second sample.

[0088] In one embodiment, step 130: obtaining the SOH of the power battery at the current time according to the first SOH and the second SOH, comprises:

[0089] The first case is that when the first SOH and the second SOH are equal, the first SOH and the second SOH are confirmed as the SOH of the power battery at the current time;

[0090] The second case is that when the first SOH and the second SOH are not equal, the SOH of the power battery at the current time is confirmed based on whether the absolute value of the difference between the first SOH and the second SOH is greater than a preset threshold.

[0091] It can be understood that the second case includes two cases of the first SOH being less than or greater than the second SOH. The above three cases are specifically listed as follows:

[0092] 1) SOH1< SOH2, that is, the value of the first SOH is less than the value of the second SOH;

[0093] 2) SOH1> SOH2, that is, the value of the second SOH is less than the value of the first SOH;

[0094] 3) SOH1=SOH2, that is, the value of the second SOH is equal to the value of the first SOH.

[0095] In one embodiment, when the first SOH and the second SOH are not equal, the SOH of the power battery at the current time is confirmed based on whether the absolute value of the difference between the first SOH and the second SOH is greater than a preset threshold.

[0096] Step 910: in the case that the absolute value of the difference between the first SOH and the second SOH is less than the preset threshold, the smaller one of the first SOH and the second SOH is determined as the SOH of the power battery at the current moment.

[0097] As described above, the preset threshold is SOH0. The absolute value of the difference between the first SOH and the second SOH is ΔSOH, and ΔSOH = |SOH1-SOH2|.

[0098] The SOH of the power battery at the current moment is SOH Final Therefore, in the case that SOH0 is greater than ΔSOH (i.e., the absolute value of the difference between the first SOH and the second SOH is less than the preset threshold), the final result is:

[0099] SOH Final = min(SOH1, SOH2).

[0100] Step 920: in the case that the absolute value of the difference between the first SOH and the second SOH is greater than or equal to the preset threshold, the first SOH and the second SOH are fused by weighting to obtain the SOH of the power battery at the current moment.

[0101] By weighting the SOH at the current moment under different conditions, the dynamic SOH at the current moment can be accurately calculated, and the accuracy of the result is ensured.

[0102] In one embodiment, the step 920 of fusing the first SOH and the second SOH by weighting includes:

[0103] Step 1010: taking the product of the smaller one of the first SOH and the second SOH and the preset threshold to obtain a first weighted term;

[0104] Step 1020: taking the product of the larger one of the first SOH and the second SOH and one minus the preset threshold to obtain a second weighted term;

[0105] Step 1030: obtaining the SOH based on the first weighted term and the second weighted term.

[0106] Therefore, in the case that SOH0 is less than or equal to ΔSOH (i.e., the absolute value of the difference between the first SOH and the second SOH is greater than or equal to the preset threshold), the final result is:

[0107] SOH Final = ΔSOH*min(SOH1, SOH2) + (1-ΔSOH)*max(SOH1, SOH2).

[0108] The dynamic SOH predicted in the power battery charging and discharging process is fused with the static SOH estimated in the complete power battery charging and discharging process, so that the dynamic SOH of the battery can be monitored in real time during the charging and discharging process, and the problem of large deviation caused by using the static OCV-SOC characteristic curve to estimate the dynamic OCV-SOC is solved. This is helpful for real-time understanding of the health status of the power battery, so that the power battery can be disposed in time in case of safety problems.

[0109] In one embodiment, Figure 8 A module schematic diagram for SOH estimation of a power battery in an embodiment of the present application is shown in FIG. 8. As shown in FIG. 8, the SOH estimation device 800 of the power battery includes a first calculation module 810, a second calculation module 820 and a fusion module 830. Figure 8

[0110] The first calculation module 810 is configured to obtain a first SOH according to data from the start time to the current time of the power battery charging and discharging process; the second calculation module 820 is configured to obtain a second SOH according to data of the complete power battery charging and discharging process; and the fusion module 830 is configured to obtain the SOH of the power battery at the current time according to the first SOH and the second SOH.

[0111] In one embodiment, the first calculation module 810 is further configured to obtain first data from the start time to the current time of the power battery charging; and use a preset first model to predict the first data to obtain the first SOH.

[0112] The SOH estimation device 800 of the power battery further includes a first training module 840, which is configured to obtain a first sample of the cell charging and discharging data, wherein the label of the first sample is the SOH value of different charging and discharging data; and train a preset recurrent neural network model according to the first sample to obtain the first model.

[0113] In one embodiment, the first training module 840 is further configured to obtain a voltage-time curve and a current-time curve of the cell charging and discharging; in a preset time range, at least one voltage segment and at least one current segment are intercepted in the voltage-time curve and the current-time curve; the SOH value corresponding to each voltage segment and each current segment is obtained, and the SOH value is confirmed as label information; and the first sample is obtained according to the at least one voltage segment and the at least one current segment, and the SOH value.

[0114] ​In one embodiment, the second computing module 820 is further configured to obtain second data of a complete charging and discharging process of the power battery; the second data includes temperature and open circuit voltage; and the SOC of the power battery is obtained by using a preset second model to predict the second data; and the second SOH is obtained according to the SOC, a starting voltage and a cut-off voltage of the charging and discharging of the power battery.

[0115] In one embodiment, the second computing module 820 is further configured to calculate the charging and discharging ampere-hour integral of the power battery according to the SOC, the starting voltage and the cut-off voltage of the charging and discharging of the power battery, and obtain the second SOH.

[0116] The SOH estimation device 800 of the power battery further includes a second training module 850, which is configured to obtain a second sample of complete charging and discharging data of a battery cell, wherein the second sample includes complete charging and discharging data of different models of battery cells under at least one temperature and at least one current rate combination; and the second model is obtained by training a preset integrated algorithm model according to the second sample.

[0117] In one embodiment, the second training module 850 is further configured to obtain a random noise sample; and the random noise sample is input into a preset generation network to obtain the second sample.

[0118] The SOH estimation device 800 of the power battery further includes a third training module 860, which is configured to input the random noise sample into a preset generation network to obtain a fake sample; input the real sample and the fake sample into a preset discrimination network to obtain a prediction result; and update the preset generation network and the preset discrimination network based on the prediction result until the discrimination network cannot distinguish the authenticity of the fake sample generated by the preset generation network, and obtain the preset generation network.

[0119] In one embodiment, the fusion module 830 is further configured to, in a case where the first SOH and the second SOH are equal, confirm the first SOH and the second SOH as the SOH of the power battery at the current time; and in a case where the first SOH and the second SOH are not equal, confirm the SOH of the power battery at the current time based on whether an absolute value of a difference between the first SOH and the second SOH is greater than a preset threshold.

[0120] In one embodiment, the fusion module 830 is further configured to, in a case where the absolute value of the difference between the first SOH and the second SOH is less than the preset threshold, confirm the smaller one of the first SOH and the second SOH as the SOH of the power battery at the current time.

[0121] In one embodiment, the fusion module 830 is further configured to, in a case where the absolute value of the difference between the first SOH and the second SOH is greater than or equal to the preset threshold, perform weighted fusion on the first SOH and the second SOH to obtain the SOH of the power battery at the current time.

[0122] In one embodiment, the fusion module 830 is further configured to take the product of the smaller one of the first SOH and the second SOH and a preset threshold to obtain a first weighted term; take the product of the larger one of the first SOH and the second SOH and one minus the preset threshold to obtain a second weighted term; and obtain the SOH based on the first weighted term and the second weighted term.

[0123] In one embodiment, an electronic device is provided, including a memory and a processor, the memory is configured to store computer executable instructions; the processor is configured to access the memory and execute the computer executable instructions to perform the operations in the detection method of any one of the preceding embodiments.

[0124] In one embodiment, the processor can be an integrated circuit chip with a processing capability of signals. In the implementation process, each step of the method embodiments described above can be completed by the integrated logic circuit or the instruction in the form of software in the processor. The processor described above can be a general processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. Each method and step disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read only memory, a programmable read only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory and combines the hardware to complete the steps of the above method.

[0125] In one embodiment, the memory can be volatile memory or nonvolatile memory, or can include both volatile and nonvolatile memory. By way of illustration, and not limitation, nonvolatile memory can be read-only memory (ROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electrically EPROM (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which acts as external cache. By way of illustration and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous dynamic RAM (SDRAM), double-data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), and direct Rambus RAM (DR RAM). It is to be noted that the system and method described herein are intended to include, and not to be limited to, all possible combinations of the above and any other suitable type of memory.

[0126] In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program, which, when executed by a processor, performs the method of any of the preceding embodiments.

[0127] In one embodiment, a computer program product is provided, characterized by comprising computer readable code, or a non-transitory computer readable storage medium carrying computer readable code, which, when run in an electronic device, causes a processor in the electronic device to perform the method of any of the preceding embodiments.

[0128] It should be understood that, although the steps in the flowcharts involved in the embodiments described above are shown in sequence according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other sequences. Moreover, at least some of the steps in the flowcharts involved in the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of the steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least some of the other steps or the steps or stages in the other steps.

[0129] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions recorded in the above embodiments can still be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the description of the present application. In particular, as long as there is no structural conflict, each technical feature mentioned in each embodiment can be combined in any manner. The present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method of estimating SOH of a power battery, characterized by, The method comprises the following steps: According to the data from the starting time to the current time of the power battery charging and discharging process, a first SOH is obtained, including obtaining first data from the starting time to the current time of the power battery charging; Using a preset first model to predict the first data, the first SOH is obtained, wherein the first model is trained by a first sample with different charging and discharging data SOH values; According to the data of the complete charging and discharging process of the power battery, a second SOH is obtained, including: obtaining second data of the complete charging and discharging process of the power battery; wherein the second data includes temperature and open circuit voltage; using a preset second model to predict the second data, obtaining the SOC of the power battery; according to the SOC, the starting voltage and the cut-off voltage of the power battery charging and discharging, the second SOH is obtained, the second model is trained by a second sample including the complete charging and discharging data of different models of battery cells under at least one temperature and at least one current rate combination; According to the relationship between the absolute value of the difference between the first SOH and the second SOH and the preset threshold, the SOH of the power battery at the current time is obtained, wherein when the absolute value of the difference is greater than or equal to the preset threshold, the first SOH and the second SOH are weighted and fused.

2. The SOH estimation method of a power battery according to claim 1, characterized by, The first model is trained in the following way: Obtain the first sample of the battery charging and discharging data; According to the first sample, a preset recurrent neural network model is trained to obtain the first model.

3. The SOH estimation method of a power battery according to claim 2, characterized by, The first sample of the battery charging and discharging data is obtained, including: Obtain the voltage-time curve and current-time curve of the battery charging and discharging; In a preset time range, at least one voltage segment and at least one current segment are intercepted in the voltage-time curve and the current-time curve; Obtain the SOH value corresponding to each voltage segment and each current segment, and confirm the SOH value as label information; According to the at least one voltage segment and at least one current segment, and the SOH value, the first sample is obtained.

4. The SOH estimation method of a power battery according to claim 1, characterized by, According to the SOC, the starting voltage and the cut-off voltage of the power battery charging and discharging, the second SOH is obtained, including: According to the SOC, the starting voltage and the cut-off voltage of the power battery charging and discharging, the charge and discharge ampere-hour integral of the power battery is calculated to obtain the second SOH.

5. The SOH estimation method of a power battery according to claim 1, characterized by, The second model is trained in the following way: Obtain the second sample of the complete charging and discharging data of the battery cell; According to the second sample, a preset integrated algorithm model is trained to obtain the second model.

6. The SOH estimation method of a power battery according to claim 5, wherein The second sample of the complete charging and discharging data of the battery cell is obtained, including: Obtain a random noise sample; Input the random noise sample into a preset generation network to obtain the second sample.

7. The SOH estimation method of a power battery according to claim 6, characterized by, The preset generation network is trained in the following way: Input the random noise sample into the preset generation network to obtain a fake sample; Input the true sample and the fake sample into a preset discrimination network to obtain a prediction result; Updating the preset generation network and the preset discrimination network based on the prediction result until the preset discrimination network cannot discriminate the authenticity of the fake sample generated by the preset generation network, to obtain the preset generation network.

8. The SOH estimation method of a power battery according to claim 1, characterized by, The SOH of the power battery at the current time is obtained according to the first SOH and the second SOH, including: In the case that the first SOH and the second SOH are equal, it is confirmed that the first SOH and the second SOH are the SOH of the power battery at the current time; In the case that the first SOH and the second SOH are not equal, it is confirmed that the SOH of the power battery at the current time based on whether the absolute value of the difference between the first SOH and the second SOH is greater than a preset threshold.

9. The SOH estimation method of a power battery according to claim 8, characterized by, In the case that the absolute value of the difference between the first SOH and the second SOH is less than the preset threshold, the smaller one of the first SOH and the second SOH is confirmed as the SOH of the power battery at the current time. In the case that the absolute value of the difference between the first SOH and the second SOH is greater than or equal to the preset threshold, the first SOH and the second SOH are weighted and fused to obtain the SOH of the power battery at the current time.

10. The SOH estimation method of a power battery according to claim 8, wherein The weighted fusion of the first SOH and the second SOH includes: Taking the product of the smaller one of the first SOH and the second SOH and the preset threshold to obtain a first weighted term; 11. The SOH estimation method of a power battery according to claim 10, wherein Taking the product of the larger one of the first SOH and the second SOH and one minus the preset threshold to obtain a second weighted term; Obtaining the SOH based on the first weighted term and the second weighted term. Including: The first calculation module is configured to obtain a first SOH according to data from the start time to the current time of the charging and discharging process of the power battery, including obtaining first data from the start time to the current time of the charging of the power battery; 12. A SOH estimation device of a power battery, characterized by, A preset first model is used to predict the first data to obtain the first SOH, wherein the first model is trained by first samples with different charging and discharging data and SOH values of labels; ​ ​ The second calculation module is configured to obtain a second SOH according to data of a complete charging and discharging process of the power battery, including: obtaining second data of the complete charging and discharging process of the power battery; wherein the second data includes temperature and open-circuit voltage; using a preset second model to predict the second data to obtain the SOC of the power battery; and obtaining the second SOH according to the SOC, a starting voltage and a cutoff voltage of the charging and discharging of the power battery, wherein the second model is trained by second samples including the complete charging and discharging data of different types of battery cells under at least one temperature and at least one current rate combination; The fusion module is configured to obtain the SOH of the power battery at the current time according to a relationship between an absolute value of a difference between the first SOH and the second SOH and a preset threshold, wherein when the absolute value of the difference is greater than or equal to the preset threshold, the first SOH and the second SOH are fused by weighting.

13. An electronic device comprising a memory and a processor, the memory storing a computer program operable to run on the processor, characterized in that, The processor executes the program to implement the steps in the method of any one of claims 1 to 11.

14. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps in the method of any one of claims 1 to 11.

15. A computer program product, characterised in that, The computer program is executed by the processor to implement the steps in the method of any one of claims 1 to 11. The computer program is executed by the processor to implement the steps in the method of any one of claims 1 to 11.

Images