CN115932634A

CN115932634A - Method, device, equipment and storage medium for evaluating health state of battery

Info

Publication number: CN115932634A
Application number: CN202211708578.3A
Authority: CN
Inventors: 谭震; 张越超; 高秀玲; 马华
Original assignee: Tianjin EV Energies Co Ltd
Current assignee: Tianjin EV Energies Co Ltd
Priority date: 2022-12-29
Filing date: 2022-12-29
Publication date: 2023-04-07

Abstract

The invention discloses a method, a device, equipment and a storage medium for evaluating the health state of a battery. The method comprises the following steps: acquiring charge and discharge data corresponding to a battery to be tested, and determining first charge and discharge data in the charge and discharge data and second charge and discharge data which is closest to the first charge and discharge data and meets a preset ampere-hour integral processing condition; processing the first charging data based on the state evaluation model to obtain a first available capacity, and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested; processing the second charge and discharge data through an ampere-hour integration method to obtain a second evaluation attribute; based on the first evaluation attribute and the second evaluation attribute, a target evaluation result is determined. The method solves the problems that the evaluation accuracy is low, and the SOC-OCV curve shifts along with the attenuation of the battery to cause the increase of the error of the evaluation result in the later period and the like when the health state of the battery is evaluated by an ampere-hour integration method, and achieves the effect of improving the evaluation accuracy of the health state of the battery in the whole life cycle.

Description

Method, device, equipment and storage medium for evaluating health state of battery

Technical Field

The present invention relates to the field of computer processing technologies, and in particular, to a method, an apparatus, a device, and a storage medium for evaluating a state of health of a battery.

Background

With the rapid development of the electric automobile market, the accumulated loading capacity of the lithium ion power battery is increased year by year, so that higher requirements on the safety, the service life and other performances of the battery are provided, and how to accurately evaluate the health state of the power battery also becomes a hotspot of research in the industry.

At present, a traditional ampere-hour integral estimation method is generally used for estimating the actual capacity of a battery according to the charging and discharging time and current in a certain charging and discharging section, so as to estimate the health state, however, as the battery ages, the problem that the estimation accuracy of the health state of the battery at the later stage is reduced exists.

Disclosure of Invention

The invention provides a method, a device, equipment and a storage medium for evaluating the health state of a battery, and aims to achieve the technical effect of improving the accuracy of evaluating the health state of the battery in the whole life cycle.

According to an aspect of the present invention, there is provided a method of evaluating a state of health of a battery, the method including:

acquiring charge and discharge data corresponding to a battery to be tested, and determining first charge and discharge data in the charge and discharge data and second charge and discharge data which is closest to the first charge data and meets a preset ampere-hour integral processing condition;

processing the first charging data based on a state evaluation model obtained through pre-training to obtain a first available capacity, and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested;

processing the second charging and discharging data through an ampere-hour integration method to obtain a second evaluation attribute;

and determining a target evaluation result of the state of health evaluation of the battery to be tested based on the first evaluation attribute and the second evaluation attribute.

According to another aspect of the present invention, there is provided an apparatus for battery state of health assessment, the apparatus comprising:

the charging and discharging data acquisition module is used for acquiring charging and discharging data corresponding to a battery to be detected, and determining first charging data in the charging and discharging data and second charging and discharging data which are closest to the first charging data and meet a preset ampere-hour integral processing condition;

the first evaluation attribute determining module is used for processing the first charging data based on a state evaluation model obtained by pre-training to obtain first available capacity and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested;

the second evaluation attribute determining module is used for processing the second charge and discharge data through an ampere-hour integral method to obtain a second evaluation attribute;

and the target evaluation result determining module is used for determining a target evaluation result for evaluating the state of health of the battery to be tested based on the first evaluation attribute and the second evaluation attribute.

According to another aspect of the present invention, there is provided an electronic apparatus including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform a method of battery state of health assessment according to any of the embodiments of the present invention.

According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement a method for battery state of health assessment according to any one of the embodiments of the present invention when executed.

According to the technical scheme of the embodiment of the invention, the first charging data in the charging and discharging data and the second charging and discharging data which is closest to the first charging data and meets the preset ampere-hour integral processing condition are determined by acquiring the charging and discharging data corresponding to the battery to be tested, the first charging data is processed based on a state evaluation model obtained by pre-training to obtain a first available capacity, and a first evaluation attribute is determined based on the first available capacity and the rated capacity of the battery to be tested; processing the second charge and discharge data through an ampere-hour integration method to obtain a second evaluation attribute; the method comprises the steps of determining a target evaluation result of the evaluation of the health state of the battery to be evaluated based on a first evaluation attribute and a second evaluation attribute, solving the problems that the health state of the battery is evaluated by an ampere-hour integration method in the prior art, the evaluation accuracy is low, and the SOC-OCV curve shifts along with the attenuation of the battery, so that the error of a later evaluation result is increased, and the like.

It should be understood that the statements in this section are not intended to identify key or critical features of the embodiments of the present invention, nor are they intended to limit the scope of the invention. Other features of the present invention will become apparent from the following description.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a method for evaluating a state of health of a battery according to an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a state estimation model according to a second embodiment of the present invention;

fig. 3 is a schematic structural diagram of an apparatus for evaluating a state of health of a battery according to a third embodiment of the present invention;

fig. 4 is a schematic structural diagram of an electronic device implementing the method for evaluating the state of health of a battery according to an embodiment of the present invention.

Detailed Description

In order to make those skilled in the art better understand the technical solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Moreover, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Example one

Fig. 1 is a flowchart of a method for evaluating a state of health of a battery according to an embodiment of the present invention, where the embodiment is applicable to a case of evaluating a state of health of a battery, and the method may be performed by a device for evaluating a state of health of a battery, where the device for evaluating a state of health of a battery may be implemented in a form of hardware and/or software, and the device for evaluating a state of health of a battery may be configured in a computing device. As shown in fig. 1, the method includes:

s110, acquiring charge and discharge data corresponding to the battery to be tested, and determining first charge data in the charge and discharge data and second charge and discharge data which is closest to the first charge data and meets a preset ampere-hour integral processing condition.

The battery to be tested may be a battery whose State Of Health (SOH) needs to be evaluated, such as a lithium battery, a lead-acid battery, a nickel-metal hydride battery, and the like. The battery to be tested can be deployed in the power electric vehicle to provide electric energy for the power electric vehicle. The first charge data of the charge and discharge data may be used as an input parameter for the subsequent model. The second charge and discharge data in the charge and discharge data may be a segment of charge data or a segment of discharge data. The preset ampere-hour integral processing condition can be a preset factor for detecting whether the second charging and discharging data is processed by the ampere-hour integral method.

In this embodiment, the charge and discharge data corresponding to the battery to be tested may be obtained in the process of charging the battery to be tested, or the charge and discharge data corresponding to the battery to be tested may be stored in a certain cache location in advance, and the charge and discharge data may be called from the cache location through an interface, at this time, the charge and discharge data may include charge and discharge data of multiple time periods, and the charge data in a period of time may be determined as the first charge data based on the charging time and the charging condition. Further, the charge and discharge data which is closest to the first charge data and meets the preset ampere-hour integral processing condition can be extracted as second charge and discharge data. To estimate the state of health of the battery based on the first charge data and the second charge-discharge data.

Optionally, the preset ampere-hour integral processing condition includes at least one of the following: the current integral value corresponding to the second charging and discharging data is larger than the product value between the preset weight value and the rated capacity of the battery to be tested; the charging and discharging state is in a current-free standing state within a preset second time length before the first charging and discharging moment corresponding to the initial voltage in the second charging and discharging data; and the charging and discharging state is in a current-free standing state within a preset second time length before a second charging and discharging moment corresponding to the tail voltage in the second charging and discharging data.

The expression form of the preset weight value may be decimal, fractional or percentage, for example, 30% or 35%. The preset second time period may be 1 hour or 1.5 hours. The specific values of the preset weight value and the preset second duration can be determined by a technician according to the actual working condition. The rated capacity is the minimum amount of electricity that the battery should discharge under rated operating conditions, and may be, for example, 10Ah or 28Ah, depending on the characteristics of the battery to be tested.

In this embodiment, a product of the preset weight value and the rated capacity of the battery to be tested may be obtained. And requiring that the current integral of the charge-discharge section of the second charge-discharge data selected from the charge-discharge data is greater than the product value. Meanwhile, in second charge and discharge data selected from the charge and discharge data, an initial voltage (OCV 1) and a final voltage (OCV 2) are required, and the two OCVs need to be kept standing without current for more than a preset second time period, namely, the two OCVs need to be kept standing without current for a preset second time period before the charge and discharge time corresponding to the two OCVs. The qualified charge and discharge data screened out according to the three conditions can be used as second charge and discharge data, and the second charge and discharge data is selected to be closest to the first charge data.

In order to improve the accuracy of the evaluation of the health state of the battery, in the process of determining the first charging data in the charging and discharging data, the charging time when the battery to be tested is in a full charging state in the charging and discharging data can be determined; and acquiring first charging data within a preset first time length between the charging time and the charging time from the charging and discharging data.

The preset first time period may be 100s or 200s.

Specifically, the charging time when the nearest battery to be tested is charged to the full charge state in the charging and discharging data can be identified, and the charging data within a preset first time period before the charging time is extracted as the first charging data. For example, the latest data of full charge can be identified by retrieving the historical charge and discharge data of the battery to be tested, and the data of the charge voltage, the charge current, the charge temperature and the like in a period of time before the charge cut-off voltage (i.e. a preset first time period, such as 30-600 s) is extracted as the first charge data, and the second charge and discharge data which is the closest to the first charge data and meets the preset ampere-hour integral processing condition is retrieved.

And S120, processing the first charging data based on the state evaluation model obtained through pre-training to obtain a first available capacity, and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested.

The first charging data includes at least one of a first charging voltage, a first charging current, and a first temperature of a charging environment. The state estimation model may be a predetermined model for estimating the state of health of the battery, and for example, may be a neural network model such as a Back Propagation (BP) neural network, a Support Vector Machine (SVM), a Relevance Vector Machine (RVM), or a Gaussian Process Regression (GPR), etc. The first evaluation attribute may be used to characterize the state of health of the battery under test, such as may be expressed in terms of percentage, decimal, fraction, etc.

In this embodiment, the first charging data may be input to the state evaluation model, and the battery capacity of the battery to be tested, which can be used externally, is obtained as the first available capacity through the processing of the state evaluation model. That is, the first available capacity may be obtained using the first charge data as an input to the state estimation model. Further, the quotient of the first available capacity and the rated capacity of the battery to be tested can be processed to obtain a quotient value, which is used as the first evaluation attribute.

In order to improve the accuracy of model evaluation, the state evaluation model can be trained by acquiring the charging data of the battery under various charging conditions. Before training the state estimation model, training samples required for training the state estimation model may be first determined, so as to obtain the state estimation model based on the training samples. Specifically, the training samples may be determined in the following manner: acquiring historical charging data of a tested battery under a plurality of charging working conditions; and determining a training sample of the training state evaluation model based on the third charging voltage, the third charging current and the third temperature in each third charging data and the corresponding third available capacity.

The battery to be tested can be a battery needing to be subjected to a charging test. The historical charging data comprises third charging data corresponding to a plurality of charging and discharging cycles. The third charging data includes at least one of a third charging voltage, a third charging current, a third temperature of the charging environment, and a third available capacity. The charge and discharge period may be a data acquisition period, e.g., 10s, or 20s.

It should be noted that, in order to make the obtained state estimation model have high accuracy, training samples can be obtained as many as possible and abundantly, so as to obtain the state estimation model after training.

In practical application, the battery to be tested can be charged and discharged by simulating different charging conditions, historical charging data of the battery to be tested under a plurality of charging conditions can be obtained, for example, the battery can be charged by adjusting charging voltage, current, power or temperature, and the like, and in the charging process, data of capacity, voltage, current, power, temperature and the like are collected to obtain charging data under various charging conditions as historical charging data. For example, after the tested battery is selected, the tested battery is subjected to charging tests of 0.2C, 0.5C, 1C and 2C currents and cyclic aging charging conditions at the temperature of 10 ℃, 25 ℃ and 40 ℃, respectively, the charging and discharging period of the charging condition point is set to 10s, namely, charging data (including current, voltage, temperature, capacity and the like) are collected every 10s, charging data under different current and temperature conditions are obtained, and third charging data corresponding to a plurality of charging and discharging periods are correspondingly obtained. The battery aging experiment database can be constructed based on the charging data under various charging conditions, so that the training sample of the training state evaluation model is determined based on the historical charging data in the experiment database. Specifically, a third charging voltage, a third charging current, a third temperature of the charging environment, and the like in the third charging data may be used as input samples, a corresponding third available capacity may be used as an output label corresponding to the input sample, the input sample and the corresponding output label may be used as a training sample, and a plurality of training samples are obtained correspondingly, so as to obtain the state evaluation model based on the training of the training samples.

Illustratively, after a state evaluation model (which may be one of a BP neural network, a support vector machine SVM, a correlation vector machine RVM, or a gaussian process regression GPR method) is selected, a data set conforming to a state evaluation model format may be established, data such as current, voltage, temperature, and the like, which are a period of time (30-600 s) before a charge cut-off voltage of each charge and discharge cycle under different charge conditions, may be extracted through a matlab program, an input characteristic parameter matrix may be established, capacity data of each charge and discharge cycle may be extracted, an output tag matrix may be established, the input characteristic parameter matrix may be used as an input of the state evaluation model, the output tag matrix may be used as an output of the state evaluation model, and the state evaluation model may be trained.

In this embodiment, the implementation manner of training the obtained state estimation model may be: for each training sample, inputting a third charging voltage, a third charging current and a third temperature in third charging data of the current training sample into a state evaluation model to obtain an evaluation available capacity; determining a loss value based on the estimated available capacity and a third available capacity in third charging data of the current training sample to correct the model parameters in the state estimation model based on the loss value; and (5) taking the loss function convergence in the state evaluation model as a training target to obtain the state evaluation model.

It should be noted that, when it is required to determine the estimated available capacity corresponding to each training sample, the determination of the estimated available capacity of any training sample data may be processed as the determination of the estimated available capacity of the current training sample, that is, one of the training samples may be described as the current training sample. The state estimation model to be trained may be one of a BP neural network, a support vector machine SVM, a correlation vector machine RVM, or a gaussian process regression GPR method. The model parameters of the state evaluation model to be trained may be default values. The state estimation model to be trained may be trained on a per training sample basis to obtain a trained state estimation model. The loss value refers to the error between the estimated available capacity and the theoretical third available capacity. The training target is that the model is trained to achieve the convergence of the preset loss function.

It should be noted that, since the model parameter of the state estimation model to be trained is an initial value, or a parameter for which the correction is not completed, the estimated available capacity obtained at this time is inaccurate, and accordingly, there is a certain difference between the estimated available capacity and the third available capacity in the training sample.

In this embodiment, the third charging voltage, the third charging current, and the third temperature in the third charging data of the current training sample may be input into the state evaluation model to be trained to obtain an estimated available capacity corresponding to the current training sample, the currently output estimated available capacity may be compared with the third charging voltage in the third charging data, and a similar error value, that is, a loss value, may be calculated, for example, an error value between the estimated available capacity and the third available capacity may be calculated by using an error evaluation method, where the error includes a Mean Absolute Percentage Error (MAPE) and a Root Mean Square Error (RMSE), and the two terms may be used as evaluation indexes of the model, so that the model parameters in the model may be adjusted based on the loss value. The training error of the loss function, that is, the loss parameter, may be used as a condition for detecting whether the loss function reaches convergence currently, for example, whether the training error is smaller than a preset error or whether an error variation trend tends to be stable, or whether the current iteration number is equal to a preset number. If the detection reaches the convergence condition, for example, the training error of the loss function is smaller than the preset error or the error change tends to be stable, indicating that the training of the state evaluation model to be trained is completed, at this time, the iterative training may be stopped. If the current condition is not met, training samples can be further obtained to train the state evaluation model to be trained until the training error of the loss function is within a preset range. When the training error of the loss function reaches convergence, the state estimation model to be trained can be used as the state estimation model obtained by final training.

Illustratively, a BP neural network is taken as a state evaluation model for illustration, and an input characteristic parameter matrix can be established based on data such as current, voltage, temperature and the like, and an output label matrix can be established based on capacity data. The BP neural network can be composed of an input layer, a hidden layer and an output layer, and an input characteristic parameter matrix can be transmitted from the input layer as an input parameter, processed layer by the hidden layer and transmitted to the output layer to obtain the estimated capacity Y. Error signals of Y and expected output capacity S are used as information, the information is reversely transmitted to an input layer by layer through a hidden layer, so that the weight of each unit is corrected until a convergence condition is met, the training of a state evaluation model is completed, the structural schematic diagram of the state evaluation model can be seen in FIG. 2, X _i The input parameters of the ith node of the input layer are input parameters; omega _ij The weight coefficient from the ith node of the input layer to the jth node of the hidden layer is obtained; theta _j Hiding a threshold at layer j node; omega _jk The weight coefficient from the jth node of the hidden layer to the kth node of the output layer; theta _k A threshold at an output layer k node; s is the expected output; y is the estimated output. i is the number of nodes of an input layer, j is the number of nodes of a hidden layer, k is the number of nodes of an output layer, a tan sig function is selected by an activation function F1 of the hidden layer, a purelin function is selected by an activation function F2 of the output layer, and MSE is selected by a loss function.

And S130, processing the second charge and discharge data through an ampere-hour integral method to obtain a second evaluation attribute.

The ampere-hour integration method estimates the SOC (State-of-Charge) of the battery, which is a State of the remaining capacity of the battery, by accumulating the amount of Charge or discharge of the battery during charging or discharging.

In this embodiment, the second charge-discharge data may be used as an input parameter of an ampere-hour integral method function to obtain a corresponding SOH value as the second evaluation attribute. Specifically, the current data in the second charge-discharge data may be integrated by an ampere-hour integration method, and the second available capacity of the battery to be measured is calculated, where the calculation formula may be:

wherein, C _t Is the second available capacity of the battery to be tested, t is the full charge time, I _t Is the charging current data. Further, the SOC1 and SOC2 may be determined from the voltage (OCV) at the start charge-discharge time and the end charge-discharge time in the second charge-discharge data and the SOC-OCV curve, and the second evaluation property, such as the second evaluation property, may be obtained based on the SOC1, SOC2, and the second available capacity

C ₀ The rated capacity of the battery to be tested.

And S140, determining a target evaluation result for the evaluation of the state of health of the battery to be tested based on the first evaluation attribute and the second evaluation attribute.

In this embodiment, after determining the first evaluation attribute and the second evaluation attribute, the first evaluation attribute and the second evaluation attribute may be integrated, and the final evaluation attribute that may characterize the state of health of the battery to be tested may be determined as the target evaluation result.

It should be noted that if the difference between the first evaluation attribute and the second evaluation attribute is small, the accuracy of the evaluation may be considered to be high, and if the difference between the first evaluation attribute and the second evaluation attribute is large, the accuracy of the evaluation may be considered to be low. In order to improve the accuracy of the evaluation of the state of health of the battery to be tested, when determining the target evaluation result of the evaluation of the state of health of the battery to be tested based on the first evaluation attribute and the second evaluation attribute, the weights of the first evaluation attribute and the second evaluation attribute may be equalized based on the difference between the first evaluation attribute and the second evaluation attribute to obtain the final target evaluation result.

Optionally, determining a target evaluation result of the evaluation of the state of health of the battery to be tested based on the first evaluation attribute and the second evaluation attribute includes: determining a correction parameter corresponding to the evaluation error based on the evaluation error between the first evaluation attribute and the second evaluation attribute; and determining a target evaluation result based on the first evaluation attribute, the second evaluation attribute and the correction parameter.

In this embodiment, the first evaluation attribute and the second evaluation attribute may be compared to obtain a difference therebetween, the difference may be used as an evaluation error, and further, a correction parameter corresponding to the evaluation error may be retrieved based on the evaluation error. For example, if the evaluation error is within a first preset interval, the correction parameter may be A1, and if the evaluation error is within a second preset interval, the correction parameter may be A2. The target evaluation result may be determined based on the correction parameter in combination with the first evaluation attribute and the second evaluation attribute.

It should be noted that, in order to improve the accuracy and the rapidity of the state estimation, an estimation error between an estimation attribute value (denoted as SOH 1) estimated from the state estimation model for the state of health of the battery and an estimation attribute value (denoted as SOH 2) calculated by an ampere-hour integration method may be subdivided, different range intervals to which the estimation error belongs may correspond to different correction parameters, and a lookup table between the estimation error and the corresponding correction parameters is configured in advance. For example, if SOH2 < SOH1, the correction parameter n may take a value between 0 and 0.2; if SOH2-SOH1 is more than 0% and SOH2-SOH1 is less than or equal to 2%, the correction parameter n can take a value between 0.4 and 0.6; if SOH2-SOH1 is more than 2% and SOH2-SOH1 is less than or equal to 5%, the correction parameter n can take a value between 0.5 and 0.7; if the SOH2-SOH1 is more than 5 percent and the SOH2-SOH1 is less than or equal to 10 percent, the correction parameter n can take a certain value between 0.7 and 0.9; if the SOH2-SOH1 is more than 10%, other historical charging and discharging data of the week are required to be selected again, the first charging data and the second charging and discharging data which are closest to the first charging data and meet the preset ampere-hour integral processing condition are determined again, so that the state evaluation model and the ampere-hour integral method evaluation are carried out again on the basis of the data for evaluation calculation, and if the evaluation errors between two evaluation attributes still differ by more than 10%, the specific safety detection of the vehicle battery is required. Illustratively, the look-up table is shown in Table 1 below:

TABLE 1

Optionally, based on the first evaluation attribute, the second evaluation attribute and the modification parameter, the implementation manner of determining the target evaluation result may be: determining a first intermediate value based on the first evaluation attribute and the modification parameter; determining a difference between the preset value and the correction parameter, and determining a second intermediate value based on the difference and the second evaluation attribute; based on the first intermediate value and the second intermediate value, a target evaluation result is determined.

Wherein the preset value may be 1.

Specifically, the correction parameter may be used as a weight value of the first evaluation attribute, and the product of the first evaluation attribute and the correction parameter may be processed to use the product value as a first intermediate value. The difference value may be obtained by subtracting the preset value from the correction parameter, and at this time, the difference value may be used as a weight value of the second evaluation attribute, and the product of the difference value and the second evaluation attribute may be used as a second intermediate value. Further, the first intermediate value and the second intermediate value may be added to obtain a sum as the target evaluation result.

For example, the formula for calculating the target evaluation result may be: SOH3= n SOH1+ (1-n) SOH2, where SOH3 represents a target evaluation result of the state of health evaluation of the battery to be measured, SOH1 represents a first evaluation attribute, SOH2 represents a second evaluation attribute, and n represents a correction parameter.

It should be noted that S120 to S130 may be executed sequentially or in parallel, and a specific execution order is not limited, and the order is only an order explaining a technical solution in each step, and is not an execution order of each step.

According to the technical scheme, the method comprises the steps of obtaining charge and discharge data corresponding to a battery to be tested, determining first charge data in the charge and discharge data and second charge and discharge data which are closest to the first charge data and meet a preset ampere-hour integral processing condition, processing the first charge data based on a state evaluation model obtained through pre-training to obtain a first available capacity, and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested; processing the second charge and discharge data through an ampere-hour integral method to obtain a second evaluation attribute; the method comprises the steps of determining a target evaluation result of the evaluation of the health state of the battery to be tested based on a first evaluation attribute and a second evaluation attribute, solving the problems that the health state of the battery is evaluated by an ampere-hour integral method in the prior art, the evaluation accuracy is low, and the SOC-OCV curve shifts along with the attenuation of the battery, so that the error of the later evaluation result is increased, and the like, realizing the evaluation of the charging data of the charging end of the battery by using a state evaluation model to obtain the first evaluation attribute, processing the second charging and discharging data by the ampere-hour integral method to obtain the second evaluation attribute, further combining the first evaluation attribute and the second evaluation attribute to determine the target evaluation result of the health state of the battery to be tested, making up the problem that the later evaluation accuracy of the ampere-hour integral method is low, and achieving the technical effect of improving the health state evaluation accuracy of the whole life cycle of the battery.

EXAMPLE III

As an alternative embodiment of the above embodiment, in order to make the technical solutions of the embodiments of the present invention further clear to those skilled in the art, a specific application scenario example is given. Specifically, the following details can be referred to.

In practical application, the running working conditions caused by different habits in different areas are found to have great difference by analyzing the running working conditions of customers, the voltage change in the last charging stage in different aging states is found to have great difference and is influenced by temperature, the voltage, the current and the temperature data in the section can be used as the input parameters of the state evaluation model, the output parameters are capacity, the state evaluation model is obtained by training, and the accuracy of the health state evaluation is improved. Before training to obtain a state evaluation model, a battery aging experiment database under different current and temperature charging conditions can be established, and the method for establishing the experiment database can be as follows: the method comprises the steps of selecting a battery to be tested, and respectively carrying out charging tests on the battery to be tested under the conditions of 0.2C current, 0.5C current, 1C current and 2C current and cyclic aging charging conditions at the temperature of 10 ℃, 25 ℃ and 40 ℃, wherein the charging and discharging period of a charging condition point is set to be 10s, namely, charging data (comprising current, voltage, temperature, capacity and the like) are collected every 10s, so that the charging data under different current and temperature conditions are obtained. Data which are in accordance with the monitoring of an actual battery management system are extracted, and the voltage, the current and the temperature of the charging tail end which occur regularly are selected as input parameters of a state evaluation model based on data driving, so that the state evaluation model is obtained based on input parameter training. The method for establishing the state evaluation model can be as follows: firstly, selecting a state evaluation model (which can be one of a BP neural network, a Support Vector Machine (SVM), a Relevance Vector Machine (RVM) or a Gaussian Process Regression (GPR) method); secondly, a data set (namely a training sample) conforming to a model format is established, data such as current, voltage, temperature and the like in a period (30-600 s) before the charge cut-off voltage of each charge-discharge cycle under different charge working conditions can be extracted through a matlab program, an input characteristic parameter matrix is established, capacity data of each charge-discharge cycle is extracted, an output label matrix is established, the input characteristic parameter matrix can be used as input of a state evaluation model, the output label matrix is used as output of the state evaluation model, finally, basic parameters of the model are set, errors of an evaluation result are calculated by adopting an error evaluation method, and an evaluation effect is evaluated based on the errors. And adjusting model parameters according to the evaluation effect until the evaluation error meets the preset requirement, and training to obtain a state evaluation model. The error calculation method may include a Mean Absolute Percentage Error (MAPE) and a Root Mean Square Error (RMSE), the calculation formula of the mean absolute percentage error is shown in formula (1), and the calculation formula of the root mean square error is shown in formula (2):

where N represents the number of training samples, Y (N) represents the estimated available capacity of the state estimation model output, and S (N) represents the third charging voltage of the theoretical output.

In real vehicle application, real vehicle charging data can be obtained, the latest full charge data is identified, data such as current, voltage and temperature in a period (such as 30-600 s) before the charging is cut off are extracted, the current, voltage and temperature data are input into a state evaluation model, and SOH1 (namely a first evaluation attribute) is obtained based on the state evaluation model; and calling historical data which is the latest time before the charge is ended, looking up a table according to an OCV-SOC curve to obtain SOC, calculating the accumulated charge-discharge capacity to estimate the actual capacity (namely the second available capacity) of the battery cell, and estimating the SOH2 (namely the second estimation attribute) based on the second available capacity. A correction parameter corresponding to an evaluation error between the first evaluation attribute and the second evaluation attribute may be obtained based on the lookup table for two methods, and a final corrected result SOH3 (i.e., a target evaluation result) is obtained based on the first evaluation attribute, the second evaluation attribute, and the correction parameter, and a formula for calculating the target evaluation result may be: SOH3= n SOH1+ (1-n) SOH2, where n denotes a correction parameter.

Based on the above technical solution, an example is given by taking a BP neural network as a state evaluation model, and an input characteristic parameter matrix may be established based on data such as current, voltage, temperature, and the like, and an output label matrix may be established based on capacity data. The BP neural network can be composed of an input layer, a hidden layer and an output layer, and an input characteristic parameter matrix can be transmitted from the input layer as an input parameter, processed layer by the hidden layer and transmitted to the output layer to obtain the estimated capacity Y. And taking error signals of the Y and the expected output capacity S as information, and reversely transmitting the information to the input layer by layer through the hidden layer, so as to correct the weight of each unit until a convergence condition is met, and finishing the training of the state evaluation model. A schematic of the state estimation model can be seen in FIG. 2, X _i The input parameters of the ith node of the input layer are input parameters; omega _ij The weight coefficient from the ith node of the input layer to the jth node of the hidden layer is obtained; theta _j Hiding a threshold at layer j node; omega _jk The weight coefficient from the jth node of the hidden layer to the kth node of the output layer; theta _k A threshold at an output layer k node; s is the desired output; y is the estimated output. i is the number of nodes of the input layer, j is the number of nodes of the hidden layer, k is the number of nodes of the output layer, the tan sig function is selected by the hidden layer activation function F1,the activation function F2 of the output layer selects the purelin function and the loss function selects the MSE.

According to the technical scheme provided by the embodiment of the invention, the state evaluation model is established and trained by adopting the parameter types according with the real vehicle data types and the granularity, the input parameters of the state evaluation model can be obtained in real vehicle application, and the feasibility is realized. Meanwhile, the state evaluation model is combined with the ampere-hour integral method, the problem that the later evaluation accuracy of the ampere-hour integral method is lowered is solved, the problem of result distortion caused by individual data errors of the state evaluation model is avoided, the health state evaluation accuracy of the full life cycle of the battery is improved, no additional test is needed in practical application, the evaluation method is simple and convenient to operate, the problems of the test period and the test cost do not exist, the hardware calculation requirement of the battery system is low, and the practical application is high.

According to the technical scheme, the method comprises the steps that charging and discharging data corresponding to a battery to be tested are obtained, first charging data in the charging and discharging data and second charging and discharging data which are closest to the first charging data and meet preset ampere-hour integral processing conditions are determined, the first charging data are processed based on a state evaluation model obtained through pre-training, a first available capacity is obtained, and a first evaluation attribute is determined based on the first available capacity and the rated capacity of the battery to be tested; processing the second charge and discharge data through an ampere-hour integration method to obtain a second evaluation attribute; the method comprises the steps of determining a target evaluation result of the evaluation of the health state of the battery to be tested based on a first evaluation attribute and a second evaluation attribute, solving the problems that the health state of the battery is evaluated by an ampere-hour integral method in the prior art, the evaluation accuracy is low, and the SOC-OCV curve shifts along with the attenuation of the battery, so that the error of the later evaluation result is increased, and the like, realizing the evaluation of the charging data of the charging end of the battery by using a state evaluation model to obtain the first evaluation attribute, processing the second charging and discharging data by the ampere-hour integral method to obtain the second evaluation attribute, further combining the first evaluation attribute and the second evaluation attribute to determine the target evaluation result of the health state of the battery to be tested, making up the problem that the later evaluation accuracy of the ampere-hour integral method is low, and achieving the technical effect of improving the health state evaluation accuracy of the whole life cycle of the battery.

EXAMPLE III

Fig. 3 is a schematic structural diagram of an apparatus for evaluating a state of health of a battery according to a third embodiment of the present invention. As shown in fig. 3, the apparatus includes: a charge and discharge data acquisition module 310, a first evaluation attribute determination module 320, a second evaluation attribute determination module 330, and a target evaluation result determination module 340.

The charging and discharging data acquisition module 310 is configured to acquire charging and discharging data corresponding to a battery to be tested, and determine first charging data in the charging and discharging data and second charging and discharging data that is closest to the first charging data and meets a preset ampere-hour integral processing condition; a first evaluation attribute determining module 320, configured to process the first charging data based on a state evaluation model obtained through pre-training to obtain a first available capacity, and determine a first evaluation attribute based on the first available capacity and a rated capacity of the battery to be tested; the second evaluation attribute determining module 330 is configured to process the second charge and discharge data by an ampere-hour integration method to obtain a second evaluation attribute; a target evaluation result determining module 340, configured to determine a target evaluation result for evaluating the state of health of the battery to be tested based on the first evaluation attribute and the second evaluation attribute.

According to the technical scheme, the method comprises the steps that charging and discharging data corresponding to a battery to be tested are obtained, first charging data in the charging and discharging data and second charging and discharging data which are closest to the first charging data and meet preset ampere-hour integral processing conditions are determined, the first charging data are processed based on a state evaluation model obtained through pre-training, a first available capacity is obtained, and a first evaluation attribute is determined based on the first available capacity and the rated capacity of the battery to be tested; processing the second charge and discharge data through an ampere-hour integration method to obtain a second evaluation attribute; the method comprises the steps of determining a target evaluation result of the evaluation of the health state of the battery to be evaluated based on a first evaluation attribute and a second evaluation attribute, solving the problems that the evaluation accuracy is low, and the SOC-OCV curve shifts along with the attenuation of the battery to cause the error increase of a later evaluation result and the like in the evaluation of the health state of the battery by an ampere-hour integral method in the prior art, realizing the evaluation of the charging data of the charging end of the battery by using a state evaluation model to obtain the first evaluation attribute, processing the second charging and discharging data by the ampere-hour integral method to obtain the second evaluation attribute, further combining the first evaluation attribute and the second evaluation attribute to determine the target evaluation result of the evaluation of the health state of the battery to be evaluated, making up the problem that the later evaluation accuracy is low by the ampere-hour integral method, and achieving the technical effect of improving the accuracy of the evaluation of the health state of the whole life cycle of the battery.

On the basis of the above device, optionally, the charging and discharging data obtaining module 310 includes: a charging timing determination unit and a first charging data determination unit.

A charging time determining unit, configured to determine a charging time when the battery to be tested is in a fully charged state in the charging and discharging data;

and the first charging data determining unit is used for acquiring the charging time and first charging data within a preset first time length before the charging time from the charging and discharging data.

On the basis of the above device, optionally, the preset ampere-hour integral processing condition includes at least one of the following:

the current integral value corresponding to the second charging and discharging data is larger than the product value between a preset weight value and the rated capacity of the battery to be tested;

the charging and discharging state is in a current-free standing state within a preset second time length before the first charging and discharging moment corresponding to the initial voltage in the second charging and discharging data;

and the charging and discharging state is in a current-free standing state within a preset second time period before a second charging and discharging moment corresponding to the tail voltage in the second charging and discharging data.

On the basis of the foregoing apparatus, optionally, the first evaluation attribute determining module 320 is specifically configured to obtain the first available capacity by using the first charging data as an input of the state evaluation model; wherein the first charging data includes at least one of a first charging voltage, a first charging current, and a first temperature of a charging environment.

On the basis of the above device, optionally, the target evaluation result determining module 340 includes a modification parameter determining unit and a target evaluation result determining unit.

A correction parameter determination unit configured to determine a correction parameter corresponding to an evaluation error between the first evaluation attribute and the second evaluation attribute based on the evaluation error;

a target evaluation result determination unit configured to determine the target evaluation result based on the first evaluation attribute, the second evaluation attribute, and the correction parameter.

On the basis of the above device, optionally, the target evaluation result determination unit includes a first intermediate value determination subunit and a second intermediate value determination subunit.

A first intermediate value determining subunit configured to determine a first intermediate value based on the first evaluation property and the correction parameter;

a second intermediate value determining subunit, configured to determine a difference between a preset value and the correction parameter, and determine a second intermediate value based on the difference and the second evaluation attribute;

a target evaluation result determination subunit configured to determine the target evaluation result based on the first intermediate value and the second intermediate value.

On the basis of the above device, optionally, the device further includes a training sample determination module, where the training sample determination module includes a historical charging data determination unit and a training sample determination unit.

The historical charging data determining unit is used for acquiring historical charging data of the battery to be tested under a plurality of charging working conditions; the historical charging data comprises third charging data corresponding to a plurality of charging and discharging cycles, and the third charging data comprises at least one of a third charging voltage, a third charging current, a third temperature of a charging environment and a third available capacity;

and the training sample determining unit is used for determining a training sample for training the state evaluation model based on the third charging voltage, the third charging current and the third temperature in each third charging datum and the corresponding third available capacity.

On the basis of the above device, optionally, the device further includes a model training module, where the model training module includes an available capacity evaluation determination unit, a loss value determination unit, and a model training unit.

The evaluation available capacity determining unit is used for inputting a third charging voltage, a third charging current and a third temperature in third charging data of the current training sample into the state evaluation model to obtain an evaluation available capacity for each training sample;

a loss value determination unit, configured to determine a loss value based on the estimated available capacity and a third available capacity in third charging data of the current training sample, so as to correct a model parameter in the state estimation model based on the loss value;

and the model training unit is used for converging the loss function in the state evaluation model as a training target to obtain the state evaluation model.

The device for evaluating the battery health state provided by the embodiment of the invention can execute the method for evaluating the battery health state provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.

Example four

Fig. 4 is a schematic structural diagram of an electronic device implementing the method for evaluating the state of health of a battery according to the embodiment of the present invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 4, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, and the like, wherein the memory stores a computer program executable by the at least one processor, and the processor 11 can perform various suitable actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from a storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data necessary for the operation of the electronic apparatus 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

A number of components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, or the like; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

Processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, or the like. Processor 11 performs the various methods and processes described above, such as a method of battery state of health assessment.

In some embodiments, the method of battery state of health assessment may be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When loaded into RAM 13 and executed by processor 11, the computer program may perform one or more of the steps of the method of battery state of health assessment described above. Alternatively, in other embodiments, the processor 11 may be configured by any other suitable means (e.g., by means of firmware) to perform the method of battery state of health assessment.

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

A computer program for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be performed. A computer program can execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical host and VPS service are overcome.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present invention may be executed in parallel, sequentially, or in different orders, and are not limited herein as long as the desired results of the technical solution of the present invention can be achieved.

The above-described embodiments should not be construed as limiting the scope of the invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method of evaluating a state of health of a battery, comprising:

2. The method of claim 1, wherein determining the first charging data of the charging and discharging data comprises:

determining the charging time of the battery to be tested in the charging and discharging data when the battery to be tested is in a fully charged state;

and acquiring the charging time and first charging data within a preset first time length before the charging time from the charging and discharging data.

3. The method of claim 1, wherein the predetermined amp-hour integration processing condition comprises at least one of:

the current integral value corresponding to the second charging and discharging data is larger than the product value between the preset weight value and the rated capacity of the battery to be tested;

the charging and discharging state is in a current-free standing state within a preset second time length before the first charging and discharging time corresponding to the initial voltage in the second charging and discharging data;

4. The method of claim 1, wherein processing the first charging data based on a pre-trained state assessment model to obtain a first available capacity comprises:

taking the first charging data as an input of the state evaluation model to obtain the first available capacity; wherein the first charging data includes at least one of a first charging voltage, a first charging current, and a first temperature of a charging environment.

5. The method of claim 1, wherein determining a target evaluation result for the state of health evaluation of the battery under test based on the first evaluation attribute and the second evaluation attribute comprises:

determining a correction parameter corresponding to an evaluation error between the first evaluation attribute and the second evaluation attribute based on the evaluation error;

determining the target evaluation result based on the first evaluation attribute, the second evaluation attribute and the correction parameter.

6. The method of claim 5, wherein said determining the target evaluation result based on the first evaluation attribute, the second evaluation attribute, and the modification parameter comprises:

determining a first intermediate value based on the first evaluation attribute and the modification parameter;

determining a difference between a preset value and the correction parameter, and determining a second intermediate value based on the difference and the second evaluation attribute;

determining the target evaluation result based on the first intermediate value and the second intermediate value.

7. The method of claim 1, further comprising:

acquiring historical charging data of a tested battery under a plurality of charging working conditions; the historical charging data comprises third charging data corresponding to a plurality of charging and discharging cycles, and the third charging data comprises at least one of a third charging voltage, a third charging current, a third temperature of a charging environment and a third available capacity;

determining training samples for training the state evaluation model based on the third charging voltage, the third charging current, and the third temperature in each third charging data and the corresponding third available capacity.

8. The method of claim 7, further comprising:

for each training sample, inputting a third charging voltage, a third charging current and a third temperature in third charging data of the current training sample into the state evaluation model to obtain an evaluation available capacity;

determining a loss value based on the estimated available capacity and a third available capacity in third charging data of the current training sample to correct model parameters in the state estimation model based on the loss value;

and taking the loss function convergence in the state evaluation model as a training target to obtain the state evaluation model.

9. An apparatus for battery state of health assessment, comprising:

the charging and discharging data acquisition module is used for acquiring charging and discharging data corresponding to a battery to be detected, determining first charging data in the charging and discharging data and second charging and discharging data which are closest to the first charging data and meet a preset ampere-hour integral processing condition;

the first evaluation attribute determining module is used for processing the first charging data based on a state evaluation model obtained by pre-training to obtain a first available capacity and determining a first evaluation attribute based on the first available capacity and the rated capacity of the battery to be tested;

10. A computer-readable storage medium storing computer instructions for causing a processor to perform the method of assessing battery state of health of any one of claims 1-8 when executed.