CN112798962B - Battery hysteresis model training method, battery SOC estimation method and device - Google Patents

Abstract

The application provides a battery hysteresis model training method, which comprises the following steps: collecting working current I _k of the battery from the time t ₀ to the time t _k on line; a current integration quantity Q is calculated from the operating current (t _k), wherein,If the battery meets the open-circuit voltage acquisition condition at the time t _k according to the working current, acquiring the open-circuit voltage OCV (t _k) of the battery at the time t _k; collecting the temperature T _bat(t_k of the battery at the time T _k); constructing a sample set from the current integration quantity Q (T _k), the temperature T _bat(t_k), and the open circuit voltage OCV (T _k); and training the battery hysteresis model according to the sample set. The application also provides a method and a device for estimating the SOC of the battery. The application can obtain the open-circuit voltage of the battery on line according to the current and the temperature acquired in real time.

Description

Battery hysteresis model training method, battery SOC estimation method and device

Technical Field

The present application relates to the field of battery technology, and in particular to a battery hysteresis model training method, and a method and device for estimating battery SOC.

Background technique

The estimation of the battery's state of charge (SOC) is one of the core functions of the battery management system. Accurate SOC estimation can ensure the safe and reliable operation of the battery system, optimize the battery system, and provide a basis for energy management and safety management of electrical devices (such as electric vehicles). Traditional SOC estimation methods mostly use the correspondence between SOC and open circuit voltage (OCV). However, the open circuit voltage and OCV in the battery do not completely correspond one to one, but there is a hysteresis relationship. Therefore, when using the open circuit voltage to estimate the battery's SOC, it is necessary to consider the impact of the hysteresis characteristics on the battery's OC. The existing modeling method of the battery open circuit voltage hysteresis characteristics introduces a lot of simplifications, which makes the modeling accuracy of the hysteresis model low, thereby affecting the SOC estimation. Therefore, the battery SOC that meets the accuracy requirements and can be obtained in real time has always been a problem that needs to be solved in the industry.

Summary of the invention

In view of this, it is necessary to provide a battery hysteresis model training method, a method and device for estimating battery SOC, which can obtain the open circuit voltage of the battery online according to the current and temperature collected in real time.

An embodiment of the present application provides a battery hysteresis model training method, which collects the working current I _k of the battery from the time t ₀ to the time t _k online, where k>0; calculates the current integral Q(t _k ) from the time t ₀ to the time t _k according to the working current, where: If it is determined according to the working current that the battery meets the open circuit voltage collection condition at time t _k , collect the open circuit voltage OCV(t _k ) of the battery at the time t _k ; collect the temperature T _bat (t _k ) of the battery at the time t _k ; construct a sample set according to the current integral Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ); and train the battery hysteresis model according to the sample set.

According to some embodiments of the present application, the sample set includes a positive sample set and a negative sample set, and constructing the sample set according to the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) includes: obtaining the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples in the positive sample set and the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the negative samples in the negative sample set; and labeling the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples with category data so that the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples carry category labels.

According to some embodiments of the present application, training the battery hysteresis model according to the sample set includes: generating a sample training set and a sample test set according to the sample set; training the battery hysteresis model according to the sample training set, and verifying the accuracy of the trained battery hysteresis model according to the sample test set; and if the accuracy is greater than or equal to a preset accuracy, ending the training process of the battery hysteresis model.

According to some embodiments of the present application, training the battery hysteresis model based on the sample set also includes: if the accuracy is less than the preset accuracy, increasing the number of the sample training sets to retrain the battery hysteresis model until the accuracy is greater than or equal to the preset accuracy.

According to some embodiments of the present application, generating a sample training set and a sample test set based on the sample set includes: randomly selecting a first preset number of sample training sets from the generated sample training set for training; and randomly selecting a second preset number of sample test sets from the generated sample test set for verification.

According to some embodiments of the present application, if the operating current remains less than a first preset value within a first preset time period, it is determined that the battery meets the open circuit voltage acquisition condition at time t _k , wherein the start time point of the first preset time period is t _ki and the end time point is t _k , and 0＜i＜k.

According to some embodiments of the present application, the method further includes: if the operating current is greater than or equal to a second preset value within a second preset time period, ending the collection of the open circuit voltage of the battery; continuing to collect the operating current of the battery, and updating historical data based on the collected operating current.

According to some embodiments of the present application, the first preset value is related to at least one of the battery capacity or the battery temperature.

According to some embodiments of the present application, the battery hysteresis model includes an input layer, a hidden layer and an output layer.

An embodiment of the present application provides a method for estimating a battery SOC by using a hysteresis model trained by the battery hysteresis model training method, the method comprising: online collecting a working current I _k of the battery from the time t ₀ to the time t _k , wherein k>0; calculating a current integral Q(t _k ) from the time t ₀ to the time t _k according to the working current; obtaining a temperature T _bat (t _k ) of the battery at the time t _k ; inputting the current integral Q(t _k ) and the temperature T _bat (t _k ) into the battery hysteresis model to obtain an open circuit voltage of the battery at the time t _k ; and obtaining the state of charge of the battery by querying a SOC-OCV correspondence relationship according to the open circuit voltage.

According to some embodiments of the present application, the SOC-OCV correspondence relationship is obtained by combining the ampere-hour integration method and the open circuit voltage method.

An embodiment of the present application provides an electrical device, which includes a memory and a processor, and the processor is used to implement the above-mentioned battery hysteresis model training method or the above-mentioned method for estimating battery SOC when executing a computer program stored in the memory.

According to some embodiments of the present application, the electrical device includes an energy storage device, an electric vehicle with two or more wheels, a drone, or an electric tool.

The implementation method of the present application collects the working current data of the battery online, and abstracts the working current data into the current integral. The current integral and the battery temperature and open circuit voltage collected at the current moment are used as sample data for training to obtain the battery hysteresis model. During use, the current integral can be calculated according to the current collected in real time, and the current integral and the collected battery temperature are used as input quantities. The corresponding open circuit voltage is obtained through the trained battery hysteresis model, and the core power state of the battery is obtained according to the open circuit voltage. Since offline experiments are not performed during the modeling process of this method, and the input quantity of the battery hysteresis model is simple, the computing power and storage capacity requirements of the electrical device are low. The advantages of low experimental quantity, low calculation quantity and low storage quantity are achieved while the accuracy is acceptable.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic structural diagram of an electric device according to an embodiment of the present application.

FIG2 is a flow chart of a battery hysteresis model training method according to an embodiment of the present application.

FIG. 3 is a schematic diagram of a battery hysteresis model according to an embodiment of the present application.

FIG. 4 is a flow chart of a method for estimating a battery SOC according to an embodiment of the present application.

Main component symbols

Electrical equipment 1

Memory 11

Processor 12

Battery 13

Collection device 14

Timer 15

The following specific implementation methods will further explain the present application in detail in conjunction with the above-mentioned drawings.

Detailed ways

The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

Please refer to Figure 1, which is a schematic diagram of an electric device according to an embodiment of the present application. As shown in Figure 1, the electric device 1 includes, but is not limited to, a memory 11, at least one processor 12, a battery 13, a collection device 14, and a timer 15, and the above components can be connected through a bus or directly connected.

In one embodiment, the battery 13 is a rechargeable battery for providing electrical energy to the electrical device 1. For example, the battery 13 can be a lead-acid battery, a nickel-cadmium battery, a nickel-metal hydride battery, a lithium-ion battery, a lithium polymer battery, and a lithium iron phosphate battery. The battery 13 is logically connected to the processor 12 through a battery management system (Battery Management System, BMS), so that the battery management system can realize functions such as charging and discharging. The battery management system can be connected to the energy storage inverter (Power Conversion System, PCS) via CAN or RS485. The battery 13 includes a battery cell (not shown in the figure), and the battery can be repeatedly charged in a cyclic and rechargeable manner.

In this embodiment, the acquisition device 14 includes an analog-to-digital converter for acquiring the voltage of the battery 13 and the current of the battery 13 and a thermometer for acquiring the temperature of the battery 13. It is understood that the acquisition device 14 can also be other voltage acquisition devices and current acquisition devices. The timer 15 is used to record the working time of the battery 13. It is understood that the electrical device 1 can also include other devices, such as a pressure sensor, a light sensor, a gyroscope, a hygrometer, an infrared sensor, etc.

It should be noted that FIG. 1 is only an example of the electric device 1. In other embodiments, the electric device 1 may also include more or fewer components, or have different component configurations. The electric device 1 may be an energy storage product, or an electric tool, or a cleaning tool, a cargo wheel or an electric vehicle with more than two wheels, a drone, or any other suitable rechargeable device.

Although not shown, the electrical device 1 may also include other components such as a Wireless Fidelity (WiFi) unit, a Bluetooth unit, a speaker, etc., which will not be described in detail here.

Please refer to FIG. 2 , which is a flow chart of a battery hysteresis model training method according to an embodiment of the present application. The battery hysteresis model training method is used to establish a battery hysteresis model, and the state of charge of the battery 13 can be estimated online through the battery hysteresis model. The battery hysteresis model training method may include the following steps:

Step S21: online collecting the working current I _k of the battery 13 from the time t ₀ to the time t _k , where k>0.

In this embodiment, the working current of the battery 13 can be collected by the collection device 14. For example, the collection device 14 is a Hall current sensor.

For example, the collected working current at time _t0 is _I0 , the collected working current at time _t1 is _I1 , the collected working current at time _t2 is _I2 , and the collected working current at time _tk is _Ik . A current information sequence can be generated according to the collected working current and time information, and the current information sequence is { _I0 , _I1 , _I2 ... _Ik }.

Step S22: calculating the current integral Q(t _k ) from the time t ₀ to the time t _k according to the working current.

In order to solve the problem of high time consumption caused by offline acquisition of experimental data in existing methods, or high computing power requirements caused by online training of the battery hysteresis model with a large amount of recent historical current data. The battery hysteresis model training method provided in this application can abstract the real-time collected working current data into a parameter (such as current integral), and use it together with the battery temperature collected at the current moment as data for constructing training samples, so as to train the battery hysteresis model. The technical effect of achieving online calculation and reducing the amount of data is achieved.

In this embodiment, the calculation formula of the current integral Q(t _k ) is: That is, the amount of electricity of the battery from the time t ₀ to the time t _k is calculated as sample data. It should be noted that the discharge direction of the working current I _k is taken as positive.

Step S23: Determine whether the battery 13 meets the open circuit voltage collection condition at time _tk . If it is determined that the battery meets the open circuit voltage collection condition at time _tk , the process proceeds to step S24; if it is determined that the battery does not meet the open circuit voltage collection condition at time _tk , the process returns to step S21.

In this embodiment, by determining whether the working current of the battery remains less than the first preset value within the first preset time period Δt1, it is determined whether the battery 13 meets the open circuit voltage collection condition at time _tk . Wherein, the starting time point of Δt1 is t _i , the ending time point is t _k , and 0＜i＜k. If the working current of the battery 13 remains less than the first preset value within the first preset time period Δt1, it is determined that the battery meets the open circuit voltage collection condition at time _tk , and the process enters step S24; if the working current of the battery 13 does not remain less than the first preset value within the first preset time period, it is determined that the battery does not meet the open circuit voltage collection condition at time _tk , and the process returns to step S21.

In this embodiment, it is determined whether the working current of the battery 13 remains less than the first preset value within the first preset time period to determine whether the battery 13 is in a static state, thereby satisfying the open circuit voltage collection condition. Normally, after the battery 13 is charged and discharged, the battery 13 is left to rest, and the open circuit voltage of the battery 13 can be collected during the process of leaving the battery 13 to rest. In the process of leaving the battery 13 to rest, the electrical working current of the battery 13 will remain at a relatively small value (e.g., zero). In this embodiment, when the working current of the battery 13 is set to remain less than the first preset value within the first preset time period, the battery 13 enters a static state, satisfying the open circuit voltage collection condition of the battery 13. In this embodiment, the first preset value is related to at least one of the battery capacity or the battery temperature.

Step S24: collecting the open circuit voltage OCV(t _k ) of the battery at the time t _k .

In this implementation, if the working current of the battery 13 remains less than the first preset value within the first preset time period, it is determined that the battery 13 is in a static state, and the open circuit voltage of the battery 13 at the time _tk can be collected.

Step S25: obtaining the temperature T _bat (t _k ) of the battery at time t _k .

In this embodiment, the temperature T _bat (t _k ) of the battery at time t _k is collected by the collection device 14. Since the battery temperature has a great influence on the battery performance, the battery temperature is used as sample data for training the battery hysteresis model to obtain a more accurate open circuit voltage value.

Step S26: constructing a sample set according to the current integrated quantity Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ), wherein the sample set includes a positive sample set and a negative sample set.

In this embodiment, in order to train the battery hysteresis model, it is necessary to first construct the training data required by the battery hysteresis model, and then construct a sample set based on the training data. The training data includes the current integral Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the battery 13 .

Specifically, constructing a sample set according to the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the battery 13 includes: obtaining the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples in the positive sample set and the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the negative samples in the negative sample set; marking the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples with category data, so that the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples carry category labels. For example, 500 current integral amounts and open circuit voltage data corresponding to the temperature at different times are selected respectively, and category data are marked for each open circuit voltage data. "1" can be used as the open circuit voltage data label corresponding to the current integral amount and temperature at time _t0 , "2" can be used as the open circuit voltage data label corresponding to the current integral amount and temperature at time _t1 , and "3" can be used as the open circuit voltage data label corresponding to the current integral amount and temperature at time _t2 .

As an optional embodiment, in order to make the battery hysteresis model more intelligent and able to identify negative samples input into the battery hysteresis model, after obtaining the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the positive samples in the positive sample set and the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the negative samples in the negative sample set; the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the negative samples are marked with category data, so that the current integral amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) of the negative samples carry category labels. It can be understood that, for the positive samples in the sample set carrying category labels, the positive samples can be identified by the battery hysteresis model according to the category labels; similarly, for the negative samples in the sample set carrying category labels, the negative samples can be identified by the battery hysteresis model according to the category labels. However, in general, it is only necessary to identify the category labels corresponding to the positive samples input into the battery hysteresis model, so as to obtain accurate output results according to the battery hysteresis model.

Step S27: training the battery hysteresis model according to the sample set.

In this implementation, after the training data is acquired and the sample set is constructed according to the training data, the battery hysteresis model is trained according to the sample set.

Specifically, training the battery hysteresis model according to the sample set includes:

(1) Generate a sample training set and a sample test set based on the sample set.

In this embodiment, the idea of cross validation can be adopted when training the battery hysteresis model, and the constructed sample set can be divided into a sample training set and a sample test set according to a suitable ratio. For example, the suitable division ratio is 6:4.

Furthermore, if the total number of divided sample training sets is still large, that is, all sample training sets are used to train the battery hysteresis model, it will result in a high cost for finding the parameters corresponding to the battery hysteresis model. Therefore, generating the sample training set may also include: randomly selecting a first preset number of sample training sets from the generated sample training set for training.

In this preferred embodiment, in order to increase the randomness of the training sample training set, a random number generation algorithm can be used for random selection.

In this preferred embodiment, the first preset number can be a preset fixed value, for example, 500, that is, 500 samples are randomly selected from the generated sample training set for training the battery hysteresis model. The first preset number can also be a preset ratio value, for example, 1/10, that is, 1/10 of the samples are randomly selected from the generated sample training set for training the battery hysteresis model.

(2) Training the battery hysteresis model according to the sample training set, and verifying the accuracy of the trained battery hysteresis model according to the sample test set.

In this embodiment, the sample training set is used to train the battery hysteresis model, and the sample test set is used to test the performance of the trained battery hysteresis model.

In this embodiment, the training samples in the sample training set are first distributed to different folders. For example, the training samples of the current integral and temperature at time _t0 are distributed to the first folder, the training samples of the current integral and temperature at time _t1 are distributed to the second folder, and the training samples of the current integral and temperature at time _t2 are distributed to the third folder. Then, the training samples of the first preset proportion (for example, 70%) are extracted from different folders as the total training samples for training the battery hysteresis model, and the remaining second preset proportion (for example, 30%) of the training samples are extracted from different folders as the total test samples to verify the accuracy of the trained battery hysteresis model.

In this embodiment, as shown in Figure 3, the battery hysteresis model includes an input layer, a hidden layer and an output layer. In the subsequent charge and discharge cycle of the battery 13, the current integral and temperature at different times are input into the battery hysteresis model, and the corresponding open circuit voltage can be output.

(3) Confirm whether the accuracy is greater than or equal to a preset accuracy.

In this embodiment, if the test accuracy is higher, it indicates that the performance of the trained battery hysteresis model is better; if the test accuracy is lower, it indicates that the performance of the trained battery hysteresis model is poor. Whether a battery hysteresis model with good performance is trained is confirmed by confirming whether the accuracy is greater than or equal to the preset accuracy. If the accuracy is greater than or equal to the preset accuracy, it is confirmed that the trained battery hysteresis model has good performance, and the process enters step (4); if the accuracy is less than the preset accuracy, it is confirmed that the trained battery hysteresis model has poor performance, and the process enters step (5).

(4) End the training process of the battery hysteresis model.

In this embodiment, if the accuracy is greater than or equal to the preset accuracy, it is confirmed that the trained battery hysteresis model has good performance and meets the requirements, and the training process of the battery hysteresis model is terminated.

(5) Increasing the number of the sample training sets to retrain the battery hysteresis model until the accuracy is greater than or equal to the preset accuracy.

In this embodiment, if the accuracy rate is less than the preset accuracy rate, it is confirmed that the performance of the trained battery hysteresis model is poor and does not meet the requirements, and it is necessary to increase the number of the sample training sets to retrain the battery hysteresis model until the accuracy rate is greater than or equal to the preset accuracy rate, and a battery hysteresis model that meets the requirements is obtained.

In this embodiment, the battery hysteresis model training method further includes:

Determine whether the working current of the battery 13 is greater than or equal to the second preset value within the second preset time period Δt2. If the working current of the battery 13 is greater than or equal to the second preset value within the second preset time period Δt2, end the acquisition of the open circuit voltage of the battery 13; continue to acquire the working current of the battery 13, and update the historical data according to the acquired working current. If the working current of the battery 13 continues to be less than the second preset value within the second preset time period Δt2, continue to determine whether the current of the battery 13 is greater than or equal to the second preset value within the third preset time period Δt3.

It should be noted that the second preset time period is the time after the first preset time period, the third preset time period is the time after the second preset time period, and so on. The starting time point of Δt2 is _tj , and the ending time point is _tn , k＜j＜n; the starting time point of Δt3 is _tl , and the ending time point is _tm , n＜l＜m.

In this embodiment, by determining whether the working current of the battery 13 is greater than or equal to the second preset value within the second preset time period, it can be determined whether the exit condition for collecting the open circuit voltage of the battery 13 is met. That is, if the working current of the battery 13 is greater than or equal to the second preset value within the second preset time period, it is determined that the battery 13 enters the charging and discharging process, and the collection of the open circuit voltage of the battery 13 ends; the working current of the battery 13 continues to be collected, and the historical data is calculated and updated according to the collected working current.

It should be noted that the second preset value is a relatively large current value, which can be used to determine whether the battery 13 enters the charging and discharging process.

Please refer to Figure 4, which is a flow chart of a method for estimating the state of charge of a battery 13 according to an embodiment of the present application. The method for estimating the state of charge of a battery 13 specifically includes the following steps. According to different requirements, the order of the steps in the flow chart can be changed, and some steps can be omitted.

Step S31: online collecting the working current I _k of the battery 13 from the time t ₀ to the time t _k , where k>0.

In this embodiment, during the cyclic charge and discharge process of the battery 13 , the operating current of the battery 13 is collected in real time by the collection device 14 .

Step S32: Calculate the current integral Q(t _k ) from the time t ₀ to the time t _k according to the working current. In this embodiment, the calculation formula of the current integral Q(t _k ) is: That is, the amount of electricity of the battery from the time t ₀ to the time t _k is calculated as sample data. It should be noted that the discharge direction of the working current I _k is taken as positive.

Step S33: obtaining the temperature T _bat (t _k ) of the battery at time t _k .

Step S34: inputting the current integral Q(t _k ) and the temperature T _bat (t _k ) into the battery hysteresis model to obtain the open circuit voltage of the battery at time t _k .

In this embodiment, the current integral Q(t _k ) and the temperature T _bat (t _k ) are input to the trained battery hysteresis model, and the battery hysteresis model outputs the open circuit voltage OCV(t _k ) at time t _k according to the current.

Step S35: querying the SOC-OCV correspondence relationship according to the open circuit voltage to obtain the state of charge of the battery 13.

In this embodiment, the SOC-OCV correspondence is pre-stored in the power consumption device 1. It should be noted that once the battery system is determined, the SOC-OCV correspondence of the battery is usually fixed. Even if the battery has been used for several cycles of charge and discharge, its SOC-OCV correspondence will not change.

Specifically, the state of charge (SOC)-open circuit voltage (OCV) correspondence relationship of the battery can be obtained by the following method:

1) Take a battery, charge the battery to a fully charged state, and then discharge the battery to an empty state using a first preset current; in this embodiment, the first preset current is a low rate current, such as 0.01C, or other currents.

2) Recording the voltage and capacity changes of the battery during the above-mentioned charge and discharge process.

3) Obtaining the state of charge of the battery during the discharge process. For example, the maximum discharge capacity of the battery is used as the full load capacity of the battery, and the capacity value of the battery that changes with time during the discharge process is divided by the full load capacity to obtain the state of charge of the battery during the discharge process.

4) Establishing the corresponding relationship between the battery voltage at different charge states during the discharge process of the battery, and obtaining the SOC-OCV corresponding relationship of the battery.

In another embodiment, the SOC-OCV correspondence relationship may be obtained by combining the ampere-hour integration method and the open circuit voltage method.

The present application provides a battery hysteresis model training method with low experimental and computational load under the premise of acceptable accuracy. The method collects the working current data of the battery online and abstracts the working current data into a current integral. The battery hysteresis model is obtained by training the current integral and the battery temperature and open circuit voltage collected at the current moment as sample data. During use, the current integral can be calculated according to the current collected in real time, and the current integral and the collected battery temperature are used as input quantities. The corresponding open circuit voltage is obtained through the trained battery hysteresis model, and the nuclear power state of the battery is obtained according to the open circuit voltage. Since offline experiments are not performed during the modeling process of this method, and the input of the battery hysteresis model is simple, the computing power and storage capacity of the electrical device are required to be low. The advantages of low experimental load, low computational load and low storage capacity are achieved while achieving acceptable accuracy.

Please continue to refer to Figure 1. In this embodiment, the memory 11 can be an internal memory of the electric device, that is, a memory built into the electric device. In other embodiments, the memory 11 can also be an external memory of the electric device, that is, a memory externally connected to the electric device.

In some embodiments, the memory 11 is used to store program codes and various data, and to achieve high-speed and automatic access to programs or data during the operation of the electrical device.

The memory 11 may include a random access memory and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

In one embodiment, the processor 12 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor, or the processor may be any other conventional processor, etc.

If the program code and various data in the memory 11 are implemented in the form of a software functional unit and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, such as implementing the steps in the battery internal short circuit detection method, and can also be completed by instructing the relevant hardware through a computer program. 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 a computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), etc. that can carry the computer program code.

It is understandable that the module division described above is a logical function division, and there may be other division methods in actual implementation. In addition, the functional modules in each embodiment of the present application may be integrated in the same processing unit, or each module may exist physically separately, or two or more modules may be integrated in the same unit. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of hardware plus software functional modules.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical solution of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present application.

Claims (11)

1. A battery hysteresis model training method, characterized in that the method comprises: The working current I _k of the battery from time t ₀ to time t _k is collected online, where k>0; The current integral Q(t _k ) from the time t ₀ to the time t _k is calculated according to the working current, wherein: If it is determined according to the working current that the battery meets the open circuit voltage collection condition at the time t _k , collecting the open circuit voltage OCV(t _k ) of the battery at the time t _k ; Obtaining a temperature T _bat (t _k ) of the battery at time t _k ; constructing a sample set according to the current integrated amount Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ); Training the battery hysteresis model according to the sample set; If the working current remains less than the first preset value within the first preset time period, it is determined that the battery meets the open circuit voltage acquisition condition at the time t _k , wherein the starting time point of the first preset time period is t _ki and the ending time point is t _k , 0<i<k; If the operating current is greater than or equal to a second preset value within a second preset time period, the acquisition of the open circuit voltage of the battery is terminated; and The operating current of the battery continues to be collected, and the historical data is updated according to the collected operating current. 2. The battery hysteresis model training method according to claim 1, characterized in that the sample set includes a positive sample set and a negative sample set, and constructing the sample set according to the current integral Q(t _k ), the temperature T _bat (t _k ) and the open circuit voltage OCV(t _k ) includes: Acquire the current integrated quantity Q(t _k ), temperature T _bat (t _k ) and open circuit voltage OCV(t _k ) of the positive samples in the positive sample set and the current integrated quantity Q(t _k ), temperature T _bat (t _k ) and open circuit voltage OCV(t _k ) of the negative samples in the negative sample set; The current integrated quantity Q(t _k ), temperature T _bat (t _k ) and open circuit voltage OCV(t _k ) of the positive sample are annotated with category data so that the current integrated quantity Q(t _k ), temperature T _bat (t _k ) and open circuit voltage OCV(t _k ) of the positive sample carry category labels. 3. The battery hysteresis model training method according to claim 2, wherein training the battery hysteresis model according to the sample set comprises: Generate a sample training set and a sample test set according to the sample set; Training the battery hysteresis model according to the sample training set, and verifying the accuracy of the trained battery hysteresis model according to the sample test set; and If the accuracy is greater than or equal to the preset accuracy, the training process of the battery hysteresis model is terminated. 4. The battery hysteresis model training method according to claim 3, wherein training the battery hysteresis model according to the sample set further comprises: If the accuracy rate is less than the preset accuracy rate, the number of the sample training sets is increased to retrain the battery hysteresis model until the accuracy rate is greater than or equal to the preset accuracy rate. 5. The battery hysteresis model training method according to claim 3, characterized in that generating a sample training set and a sample test set according to the sample set comprises: Randomly selecting a first preset number of sample training sets from the generated sample training sets for training; A second preset number of sample test sets are randomly selected from the generated sample test sets for verification. 6. The battery hysteresis model training method according to claim 1, characterized in that the first preset value is related to at least one of the battery capacity or the battery temperature. 7. The battery hysteresis model training method as described in claim 1 is characterized in that the battery hysteresis model includes an input layer, a hidden layer and an output layer. 8. A method for estimating battery SOC using a hysteresis model trained by the method according to any one of claims 1 to 7, characterized in that the method comprises: Online collecting the working current I _k of the battery from the time t ₀ to the time t _k , where k>0; Calculating the current integral Q(t _k ) from the time t ₀ to the time t _k according to the working current; Obtaining a temperature T _bat (t _k ) of the battery at time t _k ; Inputting the current integral Q(t _k ) and temperature T _bat (t _k ) into the battery hysteresis model to obtain the open circuit voltage of the battery at time t _k ; and The state of charge of the battery is obtained by querying the SOC-OCV correspondence relationship according to the open circuit voltage. 9. The method for estimating the battery SOC as claimed in claim 8, characterized in that the SOC-OCV correspondence relationship is obtained by combining the ampere-hour integration method and the open circuit voltage method. 10. An electrical device, characterized in that the electrical device comprises: Memory; and A processor, wherein the processor is used to implement the battery hysteresis model training method as described in any one of claims 1 to 7 or the battery SOC estimation method as described in any one of claims 8 to 9 when executing the computer program stored in the memory. 11. The electrical device according to claim 10, characterized in that the electrical device comprises an energy storage device, or an electric vehicle with more than two wheels, or a drone, or an electric tool.