CN111398837A - Vehicle battery health state estimation method based on data driving - Google Patents
Vehicle battery health state estimation method based on data driving Download PDFInfo
- Publication number
- CN111398837A CN111398837A CN202010249465.6A CN202010249465A CN111398837A CN 111398837 A CN111398837 A CN 111398837A CN 202010249465 A CN202010249465 A CN 202010249465A CN 111398837 A CN111398837 A CN 111398837A
- Authority
- CN
- China
- Prior art keywords
- health
- battery
- voltage
- data
- vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/385—Arrangements for measuring battery or accumulator variables
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention relates to a vehicle battery state of health (SOH) estimation method based on data driving, belonging to the technical field of battery management. The method comprises the following steps: selecting a battery to be tested, and collecting and arranging technical parameters of the battery; and carrying out a battery cyclic aging experiment according to the charging working condition and the constant current or dynamic discharging working condition of the vehicle, and receiving information such as the voltage, the current and the temperature of the battery to establish a battery aging database. And extracting health factors of other batteries according to the screened subsets, and estimating the verification of SOH estimation of the batteries under other different discharge working conditions by using the trained regression model. And embedding the trained model into a vehicle battery management system, extracting the health factors covered by the subsets in the vehicle, and estimating the SOH. The method selects the optimal feature subset for model training by using a fusion method, can effectively reduce the calculated amount and improve the model precision, and provides reference for actual vehicles.
Description
Technical Field
The invention belongs to the technical field of battery management, and relates to a vehicle battery health state estimation method based on data driving.
Background
Successful development of electric vehicles depends largely on the cycling performance, cost, and safety of the battery. Rechargeable lithium ion batteries are currently the best choice for electric vehicles due to their reasonable energy density and cycle life. Lithium ion batteries will lead to higher energy densities and more complex battery dynamics, and the efficiency and safety of such batteries will be the focus of attention. An advanced Battery Management System (BMS), which can monitor and optimize the behavior and safety of the battery, is therefore an essential whole electrification system. Reliable prediction of battery state of health (SOH) by BMS will allow batteries to fully develop their potential and maximum expected life before replacement or disposal. Knowledge of the useful life of used batteries will also enable them to be redeployed in less demanding secondary life applications, such as for power grids.
The existing methods for estimating the SOH of the battery can be mainly divided into two types: model-based methods and data-driven based methods. In which the battery is regarded as a "black box" based on the data-driven method, the regression model training may be performed based on only the extracted health factor without considering the internal complex changes of the battery, and the SOH estimation may be performed by the regression model. The health factor extraction and selection are the basis and key of a machine learning algorithm, and the estimation effect is determined to a great extent. The extraction method of the health factor can be divided into two methods based on measurement parameters and calculation parameters. In general, multiple health factors can be extracted, and selecting the optimal feature subset is particularly important for estimation, which can effectively lower data dimensionality and improve estimation accuracy. The existing method mainly comprises the steps of evaluating correlation coefficients between the health factors and the SOH of the battery, removing some health factors with low correlation coefficients, or selecting a plurality of health factors with the highest correlation coefficients as feature subsets to achieve the purpose of reducing data dimensionality. Then, there may be some health factors filtered out that have complementary effects on the retained factors, or there may be a large correlation between the retained health factors, so that the screening of the subsets may reduce the estimation accuracy. Furthermore, the health factor is extracted on the basis of the charging process, since the charging strategy remains almost unchanged over the lifetime of the electric vehicle. The health factor is extracted based on the charging process, and accurate SOH estimation of all battery cells is still a breakthrough problem faced by the BMS at present.
For the above problems, there is no scheme for extracting health factors based on a charging process and screening an optimal feature subset by using a more effective screening strategy to estimate the SOH of the battery more accurately and reliably.
Disclosure of Invention
In view of the above, the present invention provides a method for estimating a state of health of a vehicle battery based on data driving.
In order to achieve the purpose, the invention provides the following technical scheme:
a vehicle battery state of health estimation method based on data drive, this method is according to extracting the health factor based on charging process of the existing health factor extraction method, including the health factor based on that the measured data is extracted and the health factor based on that the calculation data is extracted; screening the health factor subset by using a fusion method of a filtering method and a packaging method; carrying out regression model training by adopting a data-driven algorithm, estimating the SOH of a tested monomer by using the regression model obtained by training, and loading the regression model into a vehicle-mounted battery management system to carry out SOH estimation on a battery; the method specifically comprises the following steps:
step S1: selecting a battery to be tested, and collecting and arranging technical parameters of the battery; carrying out a battery cycle aging experiment according to a vehicle charging working condition and a constant current or dynamic discharge working condition, and receiving battery voltage, current and temperature information to establish a battery aging database;
step S2: selecting an aging data set of one of the batteries as a training data set, and extracting a plurality of health factors based on charging conditions, wherein the health factors are extracted based on measurement data and calculation data;
step S3: performing data-driven regression model training based on the extracted health factors, and screening an optimal health factor subset by adopting a fusion method of a filtering method and a packaging method;
step S4: extracting health factors of other batteries according to the screened subsets, and estimating the verification of SOH estimation of the batteries under other different discharge working conditions by using a trained regression model; and embedding the trained model into a vehicle battery management system, extracting the health factors covered by the subsets in the vehicle, and estimating the SOH.
Optionally, step S1 specifically includes:
step S11: selecting a battery to be tested, and determining basic parameters of the battery to be tested, such as rated capacity, rated voltage, upper cut-off voltage and lower cut-off voltage;
step S12: carrying out initial capacity calibration on a battery to be tested;
step S13: carrying out charging test according to the actual charging condition of the vehicle to obtain the voltage, current and temperature data of the battery;
step S14: performing discharge test by adopting constant current or dynamic current to obtain the voltage, current and temperature data of the battery;
step S15: standing for 30 min;
step S16: and repeating the steps S14-S15 until the constant-current discharge capacity is less than 80% of the initial calibration capacity, and establishing a battery cycle aging database.
Optionally, in step S14, the charging condition for the vehicle includes one of a common constant-current constant-voltage CC-CV charging condition and a multi-stage constant-current charging condition;
in step S18, the battery cycle aging database includes missing value padding and error value deletion data preprocessing.
Optionally, step S2 specifically includes:
step S21: selecting an aging data set of a single battery in a battery aging database as a training set;
step S22: according to the voltage, current and temperature information obtained by real-time measurement, drawing a curve of the voltage, current and temperature information changing along with time;
step S23: extracting a plurality of health factors based on the measured data according to the curve drawn in the step S21 and the corresponding time recorded by the battery management system;
step S24: drawing a capacity increment IC curve, a voltage difference DV curve and a temperature difference DT curve according to the relation between voltage and temperature and capacity;
step S25: smoothing the original IC, DV and DT by using a filtering method;
step S26: a plurality of health factors based on the calculation data are extracted from the IC, DV, DT curves drawn at step S25.
Optionally, in step S23, the plurality of health factors based on the measurement data includes: constant current charging time, constant voltage charging time, voltage curve slope, voltage cutoff point position, initial voltage value, voltage rise in equal time, voltage curve distance difference of different cycle times, current reduction in equal time, charging cutoff current, current reduction in equal time, highest temperature and position thereof, equal time temperature rise, average temperature, time required for equal voltage rise, time required for equal current reduction and time required for equal temperature rise.
Optionally, in step S25, the filtering method includes moving average filtering, gaussian filtering, differential filtering, and wavelet transform filtering;
in step S26, the plurality of health factors based on the calculation data includes: the peak value and the position of the IC/DV/DT curve, the valley value and the position of the IC/DV/DT curve, the distance between different peak values/valley values and the peak area.
Optionally, step S3 specifically includes:
step S31: removing factors with low correlation with the SOH of the battery from the extracted health factors by adopting a filtering method of a correlation coefficient method;
step S32: performing a health factor selection process based on a forward-sequence packaging method on the health factors obtained after the filtering in the step S31;
step S33: and (4) taking the plurality of health factors obtained in the step (S32) as a final feature set, and performing regression model training by using a machine learning algorithm.
Optionally, in step S31, the correlation coefficient method includes a pearson correlation coefficient, a gray correlation degree correlation coefficient, and a spearman correlation, where the health factor with a lower correlation is a health factor with a correlation coefficient lower than a threshold value of 0.8;
in step S33, the machine learning algorithm includes support vector machine SVM, relevance vector machine RVM, neural network ANN, auto-regressive moving average, and gaussian process regression GPR methods.
Optionally, step S4 specifically includes:
step S41: extracting health factors of other monomers according to the screened health factor subset;
step S42: obtaining a regression model by training, and estimating the SOH of the battery to be tested by using the extracted health factor;
step S43: evaluating the prediction effect of the error of the calculated prediction result by adopting an error evaluation method;
step S44: and embedding the regression model into a battery management system to provide real-vehicle application.
Optionally, in step S4, the error includes a maximum absolute error, a mean absolute error, and a root-mean-square error; the selected subset of health factors is the subset selected in step S3 by the combination of the filtering method and the packaging method.
The invention has the beneficial effects that:
1) a plurality of health factors based on the charging process in the existing method are extracted, and the method is in line with practical application.
2) And the optimal feature subset is screened based on a filtering method and a packaging method, so that the estimation precision and reliability are improved, and the calculated amount is reduced.
3) The regression model training is performed based on a training set, and can be used for SOH estimation of the remaining monomers.
4) The regression model can be installed in a battery management system, and has great potential for online application.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
FIG. 1 is a process flow diagram of the present invention as a whole;
FIG. 2 is a graph of capacity fade for an embodiment; FIG. 2(a) is a complete capacity fade process, and FIG. 2(b) is a graph before 20% capacity fade;
FIG. 3 is a subset screening method using filtration in combination with packaging according to the present invention;
FIG. 4 is a training set fitting effect of an embodiment;
FIG. 5 shows the SOH estimation result of the example.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.
Wherein the showings are for the purpose of illustrating the invention only and not for the purpose of limiting the same, and in which there is shown by way of illustration only and not in the drawings in which there is no intention to limit the invention thereto; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there is an orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", "front", "rear", etc., based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not an indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes, and are not to be construed as limiting the present invention, and the specific meaning of the terms may be understood by those skilled in the art according to specific situations.
Referring to fig. 1, a method for estimating a state of health of a vehicle battery based on data driving may include the following steps:
step S1: selecting a battery to be tested, and collecting and arranging technical parameters of the battery; carrying out a battery cycle aging experiment according to a vehicle charging working condition and a constant current or dynamic discharge working condition, and receiving information such as battery voltage, current, temperature and the like to establish a battery aging database;
step S2: selecting an aging data set of one of the batteries as a training data set, and extracting a plurality of health factors based on charging conditions, wherein the health factors are extracted based on measurement data and calculation data;
step S3: performing data-driven regression model training based on the extracted health factors, and screening an optimal health factor subset by adopting a fusion method of a filtering method and a packaging method;
step S4: and extracting health factors of other batteries according to the screened subsets, and estimating the verification of SOH estimation of the batteries under other different discharge working conditions by using the trained regression model. And embedding the trained model into a vehicle battery management system, extracting the health factors covered by the subsets in the vehicle, and estimating the SOH.
As an alternative embodiment, the above step S1 selects two published experimental data sets: data set of battery aging experiments at university of maryland.
As an alternative example, the maryland university battery experimental data set described in step S1 selects two battery cells CS2_34 and CS2_35, and CS2_34 has the same charging strategy, but different discharging rates, which is closer to the actual use.
The capacity fade curves of the three cells are shown in fig. 2. Where fig. 2(a) is the complete capacity fade process and fig. 2(b) is the curve before 20% capacity fade as the SOH estimation interval of an embodiment of the present invention.
As an alternative embodiment, the step S2 specifically includes steps S21-S26:
step S21: and selecting an aging data set of a battery cell in the battery aging database as a training set.
Step S22: according to the voltage and current information obtained by real-time measurement, drawing a curve of the voltage and current information changing along with time;
step S23: and extracting a plurality of health factors based on the measured data according to the curve drawn in the step S21 and the corresponding time recorded by the battery management system.
Step S24: according to the relationship between voltage and temperature and capacity, an Incremental Capacity (IC) curve and a Differential Voltage (DV) curve are drawn.
Step S25: and smoothing the original IC and DV by using a filtering method.
Step S26: a plurality of health factors based on the calculation data are extracted from the IC, DV curve drawn at step S25.
As an alternative embodiment, in step S23, the plurality of health factors based on the measurement data includes: constant current charging time, constant voltage charging time, voltage curve slope, voltage cut-off point position, initial voltage value, voltage rise in equal time, voltage curve distance difference of different cycle times, current reduction in equal time, charging cut-off current, current reduction in equal time, time required for equal voltage rise and time required for equal current reduction.
As an alternative embodiment, gaussian filtering is selected in step S25, and the formula is:
in the formula, x and y respectively represent parameters corresponding to a coordinate horizontal axis and a coordinate vertical axis.
As an alternative embodiment, in step S25, the filtering method further includes one of moving average filtering, difference filtering, and wavelet transform filtering.
As an alternative embodiment, in step S26, the plurality of health factors based on the calculated data includes: the peak value and the position of the IC/DV curve, the valley value and the position of the IC/DV curve, the distance between different peak values/valley values, the peak area and the like.
As an alternative example, there is no temperature-related health factor in the steps S23 and S26, but the relevant researchers will understand that for the battery with the measured temperature, the temperature-related health factor should be added and the effect of the invention will not be affected.
As an alternative embodiment, the step S3 specifically includes steps S31-S33
Step S31: removing factors with low correlation with the SOH of the battery from the extracted health factors by adopting a filtering method of a correlation coefficient method;
step S32: performing a health factor selection process based on a forward-sequence packaging method on the health factors obtained after the filtering in the step S31;
step S33: and (4) taking the plurality of health factors obtained in the step (S32) as a final feature set, and performing regression model training by using a machine learning algorithm.
The health factor subset selection process is illustrated in fig. 3.
As an alternative embodiment, in step S31, the correlation coefficient method includes one of a Pearson correlation coefficient, a gray correlation coefficient, and a Spearman correlation coefficient, and the health factor with low correlation refers to a health factor with a correlation coefficient lower than a certain threshold (e.g. 0.8). As an alternative embodiment, the pearson correlation coefficient is selected as follows, and the threshold is selected to be 0.
And comparing the correlation coefficient of each health factor with a threshold value, and removing the health factors with lower correlation.
As an alternative embodiment, the specific process of selecting the subset of health factors by the packaging method is as follows:
performing model training by using the health factor subset screened by the filtering method, and performing training effect evaluation by using a self-verification method, wherein the evaluation method adopts a root mean square error as follows:
in the formula
And yiThe fit and true values are represented separately, and N represents the health factor dimension.
And removing one health factor and re-evaluating the training effect of the model.
And adding the health factor removed in the previous step, removing another health factor, and continuing to evaluate the training effect of the model.
The subset with the smallest RMSE is selected and compared with the initial subset. If the RMSE is smaller than the initial subset, the iterative process continues using this subset instead of the initial subset until a subset with a smaller RMSE cannot continue.
And taking the obtained subset with the minimum RMSE as the final input feature.
As an alternative embodiment, the data-driven model in step S33 includes one of Support Vector Machine (SVM), Relevance Vector Machine (RVM), neural network (ANN), Gaussian Process Regression (GPR), and the like. As an alternative embodiment, two algorithms of GPR and SVM are selected for explanation.
Brief introduction to GPR calculation process:
in general, we can assume that the input and output are functions of probability correlation with gaussian noise:
y＝f(x)+,
where white noise is present in a gaussian distribution. (x) can be written as:
where m (x) and k (x, x') are the mean function and covariance function, respectively:
m(x)＝E[f(x)]
k(x,x')＝E[(f(x)-m(x))(f(x')-m(x'))T]
the kernel function of the GPR model is selected as a square exponential covariance function, as shown in the following formula:
k(x,z)＝sf2*exp(-(x-z)T*inv(P)*(x-z)/2)
where sf and P are the variance and identity matrix of the signal, respectively.
The mean function is chosen to be a null function.
The likelihood function of the GPR model is chosen as gaussian likelihood function as shown in the following equation:
in the formula, m is a likelihood mean value, and sd is a standard deviation.
The input-output relationship can be written as:
in the formula InIs an n-dimensional unit matrix, and a hyperparametric matrix theta is [ sigma [ ]f,σn,l]The following can be found by the maximum likelihood function:
the output mean and error covariance of the GPR can be written as:
the likelihood function, the kernel function and the mean function of the GPR can be selected according to actual requirements.
The SVM calculation process is briefly described as follows:
the regression problem is represented by the following formula
Where w and b represent the weight matrix and the intercept of the hyperplane,
the nonlinear mapping function may also be referred to as a kernel function denoted as k (x, x'). As an alternative embodiment, the kernel function is a radial basis kernel function, as shown in the following formula
In the formula | | | x-x' | non-phosphor2Can be seen as the squared euclidean distance between two feature vectors. Is a hyper-parameter, and sigma is optimized by adopting a particle swarm algorithm. Then, an insensitive loss function is introduced to solve the nonlinear regression problem, and the problem is marked as
Where the sample below which the estimation error falls within the band, the more samples fall within the band, the higher the accuracy of the regression model, and for samples outside the band, through the relaxation variable ξ*And a non-negative penalty coefficient C. The problem of regression optimization using standard SVM can be summarized as follows,
where l represents the number of training samples.
As an alternative embodiment, step S4 specifically includes steps S41-S43
Step S41: extracting health factors of other monomers according to the screened health factor subset;
step S42: obtaining a regression model by training, and estimating the SOH of the battery to be tested by using the extracted health factor;
step S43: and evaluating the prediction effect on the error of the calculated prediction result by adopting an error evaluation method.
Step S44: and embedding the regression model into a battery management system to provide real-vehicle application.
The errors include maximum absolute error, mean absolute error, and root mean square error. The selected subset of health factors is the subset selected in step S3 by combining the filtering method with the packaging method. Specifically, the method comprises the following steps:
maximum absolute error:
mean absolute error:
root mean square error:
as an alternative embodiment, the SOH is estimated by combining the battery aging data of the university of Maryland and the optimal subset screening by applying two machine learning algorithms of GPR and SVM. CS2_35 training, CS2_34 testing. The training set fitting results are shown in fig. 4, and the test monomer estimation results are shown in fig. 5. The result shows that the optimal health factor subset can be found based on the method, the estimation precision is ensured, the calculation amount is reduced, and the method has great potential for online application of the electric automobile.
Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.
Claims (10)
1. A vehicle battery state of health estimation method based on data drive is characterized in that: the method comprises the steps of extracting health factors based on a charging process according to an existing health factor extraction method, wherein the health factors include health factors extracted based on measurement data and health factors extracted based on calculation data; screening the health factor subset by using a fusion method of a filtering method and a packaging method; carrying out regression model training by adopting a data-driven algorithm, estimating the SOH of a tested monomer by using the regression model obtained by training, and loading the regression model into a vehicle-mounted battery management system to carry out SOH estimation on a battery; the method specifically comprises the following steps:
step S1: selecting a battery to be tested, and collecting and arranging technical parameters of the battery; carrying out a battery cycle aging experiment according to a vehicle charging working condition and a constant current or dynamic discharge working condition, and receiving battery voltage, current and temperature information to establish a battery aging database;
step S2: selecting an aging data set of one of the batteries as a training data set, and extracting a plurality of health factors based on charging conditions, wherein the health factors are extracted based on measurement data and calculation data;
step S3: performing data-driven regression model training based on the extracted health factors, and screening an optimal health factor subset by adopting a fusion method of a filtering method and a packaging method;
step S4: extracting health factors of other batteries according to the screened subsets, and estimating the verification of SOH estimation of the batteries under other different discharge working conditions by using a trained regression model; and embedding the trained model into a vehicle battery management system, extracting the health factors covered by the subsets in the vehicle, and estimating the SOH.
2. The vehicle battery state-of-health estimation method based on data driving according to claim 1, characterized in that: the step S1 specifically includes:
step S11: selecting a battery to be tested, and determining basic parameters of the battery to be tested, such as rated capacity, rated voltage, upper cut-off voltage and lower cut-off voltage;
step S12: carrying out initial capacity calibration on a battery to be tested;
step S13: carrying out charging test according to the actual charging condition of the vehicle to obtain the voltage, current and temperature data of the battery;
step S14: performing discharge test by adopting constant current or dynamic current to obtain the voltage, current and temperature data of the battery;
step S15: standing for 30 min;
step S16: and repeating the steps S14-S15 until the constant-current discharge capacity is less than 80% of the initial calibration capacity, and establishing a battery cycle aging database.
3. The vehicle battery state-of-health estimation method based on data driving according to claim 2, characterized in that: in the step S14, the charging condition for the vehicle includes one of a common constant-current constant-voltage CC-CV charging condition and a multi-stage constant-current charging condition;
in step S18, the battery cycle aging database includes missing value padding and error value deletion data preprocessing.
4. The vehicle battery state-of-health estimation method based on data driving according to claim 1, characterized in that: the step S2 specifically includes:
step S21: selecting an aging data set of a single battery in a battery aging database as a training set;
step S22: according to the voltage, current and temperature information obtained by real-time measurement, drawing a curve of the voltage, current and temperature information changing along with time;
step S23: extracting a plurality of health factors based on the measured data according to the curve drawn in the step S21 and the corresponding time recorded by the battery management system;
step S24: drawing a capacity increment IC curve, a voltage difference DV curve and a temperature difference DT curve according to the relation between voltage and temperature and capacity;
step S25: smoothing the original IC, DV and DT by using a filtering method;
step S26: a plurality of health factors based on the calculation data are extracted from the IC, DV, DT curves drawn at step S25.
5. The vehicle battery state-of-health estimation method based on data driving according to claim 4, wherein: in step S23, the plurality of health factors based on the measurement data includes: constant current charging time, constant voltage charging time, voltage curve slope, voltage cutoff point position, initial voltage value, voltage rise in equal time, voltage curve distance difference of different cycle times, current reduction in equal time, charging cutoff current, current reduction in equal time, highest temperature and position thereof, equal time temperature rise, average temperature, time required for equal voltage rise, time required for equal current reduction and time required for equal temperature rise.
6. The vehicle battery state-of-health estimation method based on data driving according to claim 4, wherein: in the step S25, the filtering method includes moving average filtering, gaussian filtering, differential filtering and wavelet transform filtering;
in step S26, the plurality of health factors based on the calculation data includes: the peak value and the position of the IC/DV/DT curve, the valley value and the position of the IC/DV/DT curve, the distance between different peak values/valley values and the peak area.
7. The vehicle battery state-of-health estimation method based on data driving according to claim 1, characterized in that: the step S3 specifically includes:
step S31: removing factors with low correlation with the SOH of the battery from the extracted health factors by adopting a filtering method of a correlation coefficient method;
step S32: performing a health factor selection process based on a forward-sequence packaging method on the health factors obtained after the filtering in the step S31;
step S33: and (4) taking the plurality of health factors obtained in the step (S32) as a final feature set, and performing regression model training by using a machine learning algorithm.
8. The vehicle battery state-of-health estimation method based on data driving according to claim 7, wherein: in step S31, the correlation coefficient method includes a pearson correlation coefficient, a gray correlation coefficient, and a spearman correlation, and the health factor with a lower correlation is a health factor with a correlation coefficient lower than a threshold value of 0.8;
in step S33, the machine learning algorithm includes support vector machine SVM, relevance vector machine RVM, neural network ANN, auto-regressive moving average, and gaussian process regression GPR methods.
9. The vehicle battery state-of-health estimation method based on data driving according to claim 1, characterized in that: the step S4 specifically includes:
step S41: extracting health factors of other monomers according to the screened health factor subset;
step S42: obtaining a regression model by training, and estimating the SOH of the battery to be tested by using the extracted health factor;
step S43: evaluating the prediction effect of the error of the calculated prediction result by adopting an error evaluation method;
step S44: and embedding the regression model into a battery management system to provide real-vehicle application.
10. The vehicle battery state-of-health estimation method based on data driving according to claim 9, wherein: in step S4, the error includes a maximum absolute error, a mean absolute error, and a root mean square error; the selected subset of health factors is the subset selected in step S3 by the combination of the filtering method and the packaging method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010249465.6A CN111398837A (en) | 2020-04-01 | 2020-04-01 | Vehicle battery health state estimation method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010249465.6A CN111398837A (en) | 2020-04-01 | 2020-04-01 | Vehicle battery health state estimation method based on data driving |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111398837A true CN111398837A (en) | 2020-07-10 |
Family
ID=71432977
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010249465.6A Pending CN111398837A (en) | 2020-04-01 | 2020-04-01 | Vehicle battery health state estimation method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111398837A (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112051506A (en) * | 2020-08-28 | 2020-12-08 | 北京航空航天大学 | Similar product transferable sample screening method, system and application |
| CN112083337A (en) * | 2020-10-22 | 2020-12-15 | 重庆大学 | Power battery health prediction method oriented to predictive operation and maintenance |
| CN112347692A (en) * | 2020-09-21 | 2021-02-09 | 深圳前海有电物联科技有限公司 | Method and device for realizing predictive maintenance of battery of uninterruptible power supply and electronic device |
| CN112505570A (en) * | 2020-12-03 | 2021-03-16 | 摩登汽车(盐城)有限公司 | Method for estimating battery health state of electric automobile |
| CN112526354A (en) * | 2020-12-22 | 2021-03-19 | 南京工程学院 | Lithium battery health state estimation method |
| CN112540318A (en) * | 2020-12-22 | 2021-03-23 | 武汉理工大学 | Method for estimating health state of lead-acid storage battery for starting internal combustion engine |
| CN112595986A (en) * | 2020-11-26 | 2021-04-02 | 同济大学 | Battery service life diagnosis method based on differential thermal voltammetry curve characteristics |
| CN112611976A (en) * | 2020-12-10 | 2021-04-06 | 华东理工大学 | Power battery state of health estimation method based on double differential curves |
| CN112630662A (en) * | 2020-12-30 | 2021-04-09 | 南京航空航天大学 | Power battery SOH estimation method based on data driving and multi-parameter fusion |
| CN113030744A (en) * | 2021-02-24 | 2021-06-25 | 上海交通大学 | Battery health condition prediction method, system and medium based on health factor extraction |
| CN113158345A (en) * | 2021-04-29 | 2021-07-23 | 浙江吉利控股集团有限公司 | New energy vehicle power battery capacity prediction method and system |
| CN113219343A (en) * | 2021-06-02 | 2021-08-06 | 上海玫克生储能科技有限公司 | Lithium battery health state prediction method, system, equipment and medium based on elastic network |
| CN113296010A (en) * | 2021-04-27 | 2021-08-24 | 上海电力大学 | Battery health state online evaluation method based on differential voltage analysis |
| CN113567875A (en) * | 2021-06-11 | 2021-10-29 | 国电南瑞科技股份有限公司 | Ternary lithium ion battery health state estimation method based on K nearest neighbor regression |
| CN113721159A (en) * | 2021-09-27 | 2021-11-30 | 重庆大学 | Data-driven battery health estimation method oriented to random local charging |
| CN114047445A (en) * | 2021-09-29 | 2022-02-15 | 联合汽车电子有限公司 | Battery aging estimation method, module, chip, battery pack, battery replacement station and vehicle |
| CN114184972A (en) * | 2021-11-02 | 2022-03-15 | 江苏博强新能源科技股份有限公司 | Method and equipment for automatically estimating SOH (state of health) of battery by combining data driving with electrochemical mechanism |
| CN114742369A (en) * | 2022-03-21 | 2022-07-12 | 电子科技大学 | Evaluation method and system for cycle performance of doped lithium cobaltate material and computer equipment |
| CN114839536A (en) * | 2022-04-05 | 2022-08-02 | 北京工业大学 | Lithium ion battery health state estimation method based on multiple health factors |
| CN114910793A (en) * | 2022-04-24 | 2022-08-16 | 广东工业大学 | SOH estimation method for series battery pack of energy storage power station |
| US11485517B1 (en) * | 2021-08-20 | 2022-11-01 | Beta Air, Llc | System and method for communicating a pre-charging package stream of an electric aircraft |
| CN116430244A (en) * | 2023-06-14 | 2023-07-14 | 聊城大学 | Power battery health state estimation method based on voltage and current characteristics |
| CN116593903A (en) * | 2023-07-17 | 2023-08-15 | 中国华能集团清洁能源技术研究院有限公司 | Battery remaining life prediction method and device |
| CN117572250A (en) * | 2024-01-17 | 2024-02-20 | 山东工商学院 | Method for estimating SOH of battery based on multi-feature fusion and XGBoost |
| CN117613430A (en) * | 2024-01-22 | 2024-02-27 | 无锡冠亚恒温制冷技术有限公司 | New energy battery comprehensive test energy management method and system |
| CN117872165A (en) * | 2024-03-01 | 2024-04-12 | 天科新能源有限责任公司 | Solid-state battery performance test system based on data analysis |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105761728A (en) * | 2015-12-02 | 2016-07-13 | 中国传媒大学 | Chinese typical hearing culture symbol characteristic selection method |
| CN106154164A (en) * | 2015-03-12 | 2016-11-23 | 重庆邮电大学 | Battery health state assessment method |
| CN107122863A (en) * | 2017-04-28 | 2017-09-01 | 厦门大学 | Overstock inquiry and the Forecasting Methodology of material correlative attribute |
| CN108855988A (en) * | 2018-05-07 | 2018-11-23 | 新疆农业大学 | Walnut kernel stage division and walnut kernel grading plant based on machine vision |
| CN109633477A (en) * | 2019-01-04 | 2019-04-16 | 深圳市计量质量检测研究院 | The method for real-time monitoring of battery state-of-health based on EKF-GPR and daily fragment data |
| CN110068774A (en) * | 2019-05-06 | 2019-07-30 | 清华四川能源互联网研究院 | Estimation method, device and the storage medium of lithium battery health status |
| CN110398697A (en) * | 2019-07-23 | 2019-11-01 | 北京工业大学 | A kind of lithium ion health status estimation method based on charging process |
| CN110632528A (en) * | 2019-11-04 | 2019-12-31 | 桂林电子科技大学 | Lithium battery SOH estimation method based on internal resistance detection |
| CN110782950A (en) * | 2019-09-23 | 2020-02-11 | 江苏大学 | Tumor key gene identification method based on preference grid and Levy flight multi-target particle swarm algorithm |
| CN110873841A (en) * | 2019-11-05 | 2020-03-10 | 北京匠芯电池科技有限公司 | Battery life prediction method based on combination of data driving and battery characteristics |
-
2020
- 2020-04-01 CN CN202010249465.6A patent/CN111398837A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106154164A (en) * | 2015-03-12 | 2016-11-23 | 重庆邮电大学 | Battery health state assessment method |
| CN105761728A (en) * | 2015-12-02 | 2016-07-13 | 中国传媒大学 | Chinese typical hearing culture symbol characteristic selection method |
| CN107122863A (en) * | 2017-04-28 | 2017-09-01 | 厦门大学 | Overstock inquiry and the Forecasting Methodology of material correlative attribute |
| CN108855988A (en) * | 2018-05-07 | 2018-11-23 | 新疆农业大学 | Walnut kernel stage division and walnut kernel grading plant based on machine vision |
| CN109633477A (en) * | 2019-01-04 | 2019-04-16 | 深圳市计量质量检测研究院 | The method for real-time monitoring of battery state-of-health based on EKF-GPR and daily fragment data |
| CN110068774A (en) * | 2019-05-06 | 2019-07-30 | 清华四川能源互联网研究院 | Estimation method, device and the storage medium of lithium battery health status |
| CN110398697A (en) * | 2019-07-23 | 2019-11-01 | 北京工业大学 | A kind of lithium ion health status estimation method based on charging process |
| CN110782950A (en) * | 2019-09-23 | 2020-02-11 | 江苏大学 | Tumor key gene identification method based on preference grid and Levy flight multi-target particle swarm algorithm |
| CN110632528A (en) * | 2019-11-04 | 2019-12-31 | 桂林电子科技大学 | Lithium battery SOH estimation method based on internal resistance detection |
| CN110873841A (en) * | 2019-11-05 | 2020-03-10 | 北京匠芯电池科技有限公司 | Battery life prediction method based on combination of data driving and battery characteristics |
Non-Patent Citations (1)
| Title |
|---|
| 李国和等: "决策信息系统的连续型特征选取方法", 《信息与控制》 * |
Cited By (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112051506A (en) * | 2020-08-28 | 2020-12-08 | 北京航空航天大学 | Similar product transferable sample screening method, system and application |
| CN112347692A (en) * | 2020-09-21 | 2021-02-09 | 深圳前海有电物联科技有限公司 | Method and device for realizing predictive maintenance of battery of uninterruptible power supply and electronic device |
| CN112347692B (en) * | 2020-09-21 | 2024-02-06 | 深圳有电物联科技有限公司 | Method and device for realizing battery predictive maintenance of uninterruptible power supply and electronic device |
| CN112083337A (en) * | 2020-10-22 | 2020-12-15 | 重庆大学 | Power battery health prediction method oriented to predictive operation and maintenance |
| CN112083337B (en) * | 2020-10-22 | 2023-06-16 | 重庆大学 | Predictive operation and maintenance-oriented power battery health prediction method |
| CN112595986A (en) * | 2020-11-26 | 2021-04-02 | 同济大学 | Battery service life diagnosis method based on differential thermal voltammetry curve characteristics |
| CN112505570A (en) * | 2020-12-03 | 2021-03-16 | 摩登汽车(盐城)有限公司 | Method for estimating battery health state of electric automobile |
| CN112611976A (en) * | 2020-12-10 | 2021-04-06 | 华东理工大学 | Power battery state of health estimation method based on double differential curves |
| CN112540318B (en) * | 2020-12-22 | 2022-01-04 | 武汉理工大学 | Method for estimating health state of lead-acid storage battery for starting internal combustion engine |
| CN112526354A (en) * | 2020-12-22 | 2021-03-19 | 南京工程学院 | Lithium battery health state estimation method |
| CN112540318A (en) * | 2020-12-22 | 2021-03-23 | 武汉理工大学 | Method for estimating health state of lead-acid storage battery for starting internal combustion engine |
| CN112630662A (en) * | 2020-12-30 | 2021-04-09 | 南京航空航天大学 | Power battery SOH estimation method based on data driving and multi-parameter fusion |
| CN113030744B (en) * | 2021-02-24 | 2022-06-28 | 上海交通大学 | Battery health condition prediction method, system and medium based on health factor extraction |
| CN113030744A (en) * | 2021-02-24 | 2021-06-25 | 上海交通大学 | Battery health condition prediction method, system and medium based on health factor extraction |
| CN113296010A (en) * | 2021-04-27 | 2021-08-24 | 上海电力大学 | Battery health state online evaluation method based on differential voltage analysis |
| CN113158345A (en) * | 2021-04-29 | 2021-07-23 | 浙江吉利控股集团有限公司 | New energy vehicle power battery capacity prediction method and system |
| CN113219343A (en) * | 2021-06-02 | 2021-08-06 | 上海玫克生储能科技有限公司 | Lithium battery health state prediction method, system, equipment and medium based on elastic network |
| WO2022253038A1 (en) * | 2021-06-02 | 2022-12-08 | 上海玫克生储能科技有限公司 | Method and system for predicting state of health of lithium battery on basis of elastic network, and device and medium |
| CN113567875A (en) * | 2021-06-11 | 2021-10-29 | 国电南瑞科技股份有限公司 | Ternary lithium ion battery health state estimation method based on K nearest neighbor regression |
| US11485517B1 (en) * | 2021-08-20 | 2022-11-01 | Beta Air, Llc | System and method for communicating a pre-charging package stream of an electric aircraft |
| CN113721159A (en) * | 2021-09-27 | 2021-11-30 | 重庆大学 | Data-driven battery health estimation method oriented to random local charging |
| CN114047445A (en) * | 2021-09-29 | 2022-02-15 | 联合汽车电子有限公司 | Battery aging estimation method, module, chip, battery pack, battery replacement station and vehicle |
| CN114184972A (en) * | 2021-11-02 | 2022-03-15 | 江苏博强新能源科技股份有限公司 | Method and equipment for automatically estimating SOH (state of health) of battery by combining data driving with electrochemical mechanism |
| CN114184972B (en) * | 2021-11-02 | 2023-12-22 | 江苏博强新能源科技股份有限公司 | Automatic estimation method and equipment for SOH of battery by combining data driving and electrochemical mechanism |
| CN114742369A (en) * | 2022-03-21 | 2022-07-12 | 电子科技大学 | Evaluation method and system for cycle performance of doped lithium cobaltate material and computer equipment |
| CN114839536A (en) * | 2022-04-05 | 2022-08-02 | 北京工业大学 | Lithium ion battery health state estimation method based on multiple health factors |
| CN114839536B (en) * | 2022-04-05 | 2024-06-04 | 北京工业大学 | Lithium ion battery health state estimation method based on multiple health factors |
| CN114910793A (en) * | 2022-04-24 | 2022-08-16 | 广东工业大学 | SOH estimation method for series battery pack of energy storage power station |
| CN116430244A (en) * | 2023-06-14 | 2023-07-14 | 聊城大学 | Power battery health state estimation method based on voltage and current characteristics |
| CN116430244B (en) * | 2023-06-14 | 2023-08-15 | 聊城大学 | Power battery health state estimation method based on voltage and current characteristics |
| CN116593903B (en) * | 2023-07-17 | 2023-10-20 | 中国华能集团清洁能源技术研究院有限公司 | Battery remaining life prediction method and device |
| CN116593903A (en) * | 2023-07-17 | 2023-08-15 | 中国华能集团清洁能源技术研究院有限公司 | Battery remaining life prediction method and device |
| CN117572250A (en) * | 2024-01-17 | 2024-02-20 | 山东工商学院 | Method for estimating SOH of battery based on multi-feature fusion and XGBoost |
| CN117613430A (en) * | 2024-01-22 | 2024-02-27 | 无锡冠亚恒温制冷技术有限公司 | New energy battery comprehensive test energy management method and system |
| CN117613430B (en) * | 2024-01-22 | 2024-04-12 | 无锡冠亚恒温制冷技术有限公司 | New energy battery comprehensive test energy management method and system |
| CN117872165A (en) * | 2024-03-01 | 2024-04-12 | 天科新能源有限责任公司 | Solid-state battery performance test system based on data analysis |
| CN117872165B (en) * | 2024-03-01 | 2024-05-10 | 天科新能源有限责任公司 | Solid-state battery performance test system based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111398837A (en) | Vehicle battery health state estimation method based on data driving | |
| CN112092675B (en) | Battery thermal runaway early warning method, system and server | |
| Song et al. | Intelligent state of health estimation for lithium-ion battery pack based on big data analysis | |
| Deng et al. | Battery health estimation with degradation pattern recognition and transfer learning | |
| CN110752410B (en) | Method for rapidly sorting and recombining retired lithium battery | |
| CN110224192B (en) | Method for predicting service life of power battery by gradient utilization | |
| Zhang et al. | A machine learning-based framework for online prediction of battery ageing trajectory and lifetime using histogram data | |
| CN110058178A (en) | A kind of lithium battery method for detecting health status and system | |
| Xiong et al. | A data-driven method for extracting aging features to accurately predict the battery health | |
| CN106556802A (en) | A kind of accumulator battery exception cell recognition methodss and device | |
| Xuan et al. | State-of-charge prediction of battery management system based on principal component analysis and improved support vector machine for regression | |
| CN107422266A (en) | A kind of method for diagnosing faults and device of high capacity cell energy-storage system | |
| US20230305065A1 (en) | Apparatus and method for diagnosing state of battery | |
| CN112904219A (en) | Big data-based power battery health state prediction method | |
| CN105911476A (en) | Battery energy storage system SOC predication method based on data mining | |
| Chen et al. | Capacity prediction and validation of lithium-ion batteries based on long short-term memory recurrent neural network | |
| CN113657360A (en) | Lithium battery health state estimation method, device, equipment and readable storage medium | |
| CN111983474A (en) | Lithium ion battery life prediction method and system based on capacity decline model | |
| CN115015789A (en) | Lithium ion battery health state estimation method and device based on incremental capacitance curve | |
| CN115327403A (en) | Power battery safety evaluation method and device based on new energy automobile big data | |
| CN115994441A (en) | Big data cloud platform online battery life prediction method based on mechanism information | |
| Zhang et al. | State of health estimation for lithium-ion batteries under arbitrary usage using data-driven multimodel fusion | |
| Chen et al. | State of health estimation for lithium-ion battery based on long short term memory networks | |
| CN117113232A (en) | Thermal runaway risk identification method for lithium ion battery pack of electric automobile | |
| CN118191628A (en) | Machine learning-based mining lithium ion battery SOC prediction method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200710 |