CN114035095B - Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification - Google Patents
Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification Download PDFInfo
- Publication number
- CN114035095B CN114035095B CN202111285782.4A CN202111285782A CN114035095B CN 114035095 B CN114035095 B CN 114035095B CN 202111285782 A CN202111285782 A CN 202111285782A CN 114035095 B CN114035095 B CN 114035095B
- Authority
- CN
- China
- Prior art keywords
- point
- curve
- voltage
- inflection point
- soh
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 32
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 32
- 238000007600 charging Methods 0.000 claims abstract description 30
- 230000008859 change Effects 0.000 claims abstract description 28
- 239000013598 vector Substances 0.000 claims abstract description 24
- 230000036541 health Effects 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 17
- 230000008569 process Effects 0.000 claims abstract description 14
- 238000005070 sampling Methods 0.000 claims abstract description 8
- 238000010277 constant-current charging Methods 0.000 claims abstract description 7
- 238000003066 decision tree Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000007637 random forest analysis Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 description 7
- 238000005259 measurement Methods 0.000 description 5
- 239000007772 electrode material Substances 0.000 description 4
- 238000012937 correction Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000032683 aging Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000003487 electrochemical reaction Methods 0.000 description 2
- 238000000157 electrochemical-induced impedance spectroscopy Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- GELKBWJHTRAYNV-UHFFFAOYSA-K lithium iron phosphate Chemical compound [Li+].[Fe+2].[O-]P([O-])([O-])=O GELKBWJHTRAYNV-UHFFFAOYSA-K 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009831 deintercalation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000009830 intercalation Methods 0.000 description 1
- 230000002687 intercalation Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 229910001416 lithium ion Inorganic materials 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
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]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
- G01R31/3835—Arrangements for monitoring battery or accumulator variables, e.g. SoC involving only voltage measurements
Abstract
The invention relates to a lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification, wherein the method comprises the following steps: collecting charging data around a second voltage platform in the constant current charging process monitored by the BMS system, and obtaining terminal voltage and corresponding sampling time to form a voltage-time curve; identifying inflection points by identifying the angular change of the curve at each point by point, and constructing SOH health factors in a single charging period by the identified inflection points; combining SOH health factors of the current period of the charging process with the average temperature to form a feature vector; and taking the characteristic vector as input of a regression model of the maximum residual capacity of the battery cell obtained through offline training, predicting to obtain a predicted value of the maximum residual capacity of the next cycle period, and finishing SOH estimation. Compared with the prior art, the method has the advantages of high SOH estimation precision and the like.
Description
Technical Field
The invention relates to the technical field of battery SOH estimation, in particular to a lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification.
Background
The state of health SOH estimation of the lithium battery is a precondition for evaluating the aging degree of the battery and guaranteeing the safe, reliable and economic operation of the battery. The variety of lithium battery operating conditions, the complexity of the internal electrochemical mechanism, makes accurate estimation of SOH challenging. Offline testing of internal resistance and maximum remaining capacity of the battery, such as Electrochemical Impedance Spectroscopy (EIS), is the most accurate method of SOH estimation, but is limited by the test conditions and produces unnecessary losses, which is not feasible in most practical application scenarios. While the OCV-SOC curve-based method is accurate, it requires full charge or discharge of the battery at a low rate, or measurement of the open circuit voltage after a long time relaxation (exceeding 2 hours) at SOC levels in the entire range, so that its application range is also greatly limited. The potential of the lithium battery electrode depends on the thermodynamic properties of the electrode materials, so that the potential range and voltage plateau are unique for different electrode materials at different life stages. Although the electrode potential is difficult to measure, the OCV, which represents the potential difference between the cathode and anode, contains valuable information about the material-level phase change over the life cycle of a lithium battery. However, it is challenging to use OCV directly for health diagnostics. First, intercalation and deintercalation of lithium ions occurs only within a very narrow voltage range. Second, all voltage levels are within the threshold voltage range and overlap, and the voltage is susceptible to measurement noise. All these factors lead to difficult to observe characteristics of OCV curves for most lithium battery chemistries.
On-line SOH estimation methods based on battery operation data are an important aspect of current lithium battery health management. Currently more efficient methods include Incremental Capacity Analysis (ICA), differential voltage analysis (differential voltage analysis, DVA). ICA converts the voltage plateau on the charge/discharge voltage (V-Q) curve, identifying the dQ/dV peak on the delta capacity (IC) curve. ICA is advantageous in that it detects gradual changes in battery behavior during life cycle test, has higher sensitivity than that based on conventional charge-discharge curves, and generates key information of battery behavior related to electrochemical characteristics. The DVA method is similar to the ICA method, calculates the dV/dQ curve and extracts the peak as SOH health factor.
The current ICA analysis method and DV analysis method both adopt characteristic point extraction of characteristic curves, but because all peaks on the IC curve are positioned in a voltage platform area of the V-Q curve and are relatively flat, the analysis method is sensitive to measurement noise, and dQ/dV stability is poor in direct calculation from a data set. The current SOC estimation method based on the voltage dynamic curve extracts the point with the maximum first-order differential or the maximum second-order differential of 0 from the voltage curve as the characteristic point, is sensitive to the voltage measurement error, and directly calculates most of the actual charging data without covering the whole SOC range, so that the off-line identification method cannot be applied. The SOC estimation method given in chinese patent application CN109870655a also uses characteristic points of a voltage curve, but uses data fitting of a front inflection point and a rear inflection point under different conditions of temperature, current multiplying power and cycle times to obtain an inflection point correction coefficient equation under corresponding conditions, and uses correction coefficients to fit different working conditions. The method using the correction coefficient is rough, is mostly suitable for linear conditions, has more nonlinear relations in actual working conditions, and is difficult to achieve good estimation accuracy.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a lithium battery SOH estimation method, medium and equipment with high SOH estimation precision based on voltage curve inflection point identification.
The aim of the invention can be achieved by the following technical scheme:
a lithium battery SOH estimation method based on voltage curve inflection point identification comprises the following steps:
collecting charging data around a second voltage platform in the constant current charging process monitored by the BMS system, and obtaining terminal voltage and corresponding sampling time to form a voltage-time curve;
identifying inflection points by identifying the angular change of the curve at each point by point, and constructing SOH health factors in a single charging period by the identified inflection points;
combining SOH health factors of the current period of the charging process with the average temperature to form a feature vector;
and taking the characteristic vector as input of a regression model of the maximum residual capacity of the battery cell obtained through offline training, predicting to obtain a predicted value of the maximum residual capacity of the next cycle period, and finishing SOH estimation.
Further, the identifying the inflection point by identifying the angular change of the curve at each point is specifically:
step1, normalizing the voltage-time curve coordinates of charge and discharge to ensure that the x and y values of the curve are all between 0 and 1;
step2, calculate each data point (x i ,y i ) The direction of the curve at the point is changed, and the angle change of each data point is obtained;
step3, based on the angular change of each data point on the voltage-time curveA curve identifying said->X at peaks and valleys in the curve i Which is in the voltage-time curve (x i ,y i ) Namely an inflection point;
step4, restoring the normalized coordinates of Step1, and returning the coordinate value of the inflection point after the restoration.
The angle change of each data point is obtained by the following steps:
201 Get the point (x) i ,y i ) Two points before and after (x i-d ,y i-d ) And (x) i+d ,y i+d )；
202 Calculating vector x i+d -x i ,y i+d -y i ]Is at an inclination angle of (2)Sum vector x i -x i-d ,y i -y i-d ]Dip angle of->
203 Calculating the point (x) i ,y i ) At the point (x) i-d ,y i-d ) And (x) i+d ,y i+d ) Angle change of connection line
204 Repeating steps 201) -203) for each point of the charging curve, calculating the angular change of each point at the adjacent d-point
205 Repeating steps 201) to 204) with d being 1 to N, and calculating each point at different points Average value->As the value of the change in angle of the point.
In Step3, a function extremum recognition method is used for screeningPeaks and valleys in the curve.
In Step3, the absolute value of the screening is larger than the threshold value delta phi bound And the width of the peaks and valleys exceeds a threshold Δx bound As inflection points.
And establishing the regression model of the maximum residual capacity of the battery cell by adopting an integrated learning model.
The offline training process of the regression model of the maximum residual capacity of the battery cell comprises the following steps:
the constant-current charging mode with the same current is adopted, complete charging monitoring data of n continuous cycle periods of m same lithium batteries of a specific model from an initial stage are acquired offline, the complete charging monitoring data comprise a voltage sequence and a temperature sequence of each period and the maximum residual capacity of the battery of the period, feature vectors and the maximum residual capacity of the n cycle periods of the m same lithium batteries are acquired by utilizing a feature extraction algorithm to form a training sample set D = { (x) i ,y i ) I=1, 2,..m.m }, where x i ＝(H vi ,T i ,ind i ) As a feature vector, y i For the corresponding label, i.e. maximum remaining capacity of battery, H vi ,T i SOH health factor and average temperature, ind respectively i A corresponding cycle number for the battery;
training by using a training sample set to obtain K decision trees, and adding x according to the K decision trees i The corresponding label predicts:wherein f= { F (x) =w q(x) ' represents the decision tree vs. function space, W q(x) Representing each of the handle treesA node maps to a value;
model parameters are adjusted using the loss function.
The ensemble learning model includes a random forest, XGBoost, or LightGBM.
The present invention also provides a computer-readable storage medium comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs comprising instructions for performing a lithium battery SOH estimation method based on voltage curve inflection point identification as described above.
The present invention also provides an electronic device including:
one or more processors;
a memory; and
one or more programs stored in a memory, the one or more programs including instructions for performing a lithium battery SOH estimation method based on voltage curve inflection point identification as described above.
The invention provides an SOH prediction technology which can be more in line with the engineering practice and is combined with offline on line aiming at how to improve the SOH estimation accuracy of a battery in the practical working condition. And in the off-line stage, the voltage, the current, the temperature and the maximum residual capacity of the batteries of a plurality of groups of batteries set in different temperature environments are tested, and the voltage-related health factors are extracted as characteristics capable of representing the attenuation degree of the batteries by utilizing the association relation between the characteristic points of the voltage curve and the attenuation of the batteries. And offline training the mapping relation between the maximum residual capacity of the battery and the voltage characteristic points and the temperature by using an integrated learning technology to obtain a regression model of the maximum residual capacity of the battery. The current SOH estimation can be obtained by only extracting the voltage characteristic point, the temperature characteristic and the cycle number of the current cycle period on line and inputting the voltage characteristic point, the temperature characteristic and the cycle number into an offline regression model. Compared with other SOH estimation methods, the method provided by the invention utilizes electrochemical characteristics and a data driving method, and has stronger practicability. In summary, the invention has the following advantages:
(1) The method is suitable for various SOC charging working intervals covering the second voltage platform.
(2) The offline model is trained by adopting an integrated learning model, and the requirement on the number of samples is low.
(3) The temperature and voltage characteristics are comprehensively considered, and the model has generalization capability for different temperature conditions, so that the model has more practicability.
(4) The existing method generally takes the characteristic point that the first-order differential maximum value or the second-order differential of the voltage dynamic curve is zero as an inflection point, and the method is very sensitive to voltage measurement noise and is easy to sink into local optimum, so that the inflection point identification is wrong. In addition, the prior method ignores the influence of temperature on voltage, so that the applicability of the method to different temperature working conditions is greatly limited. The voltage key characteristic points in a single charge-discharge period extracted by the inflection point identification algorithm can accurately reflect the battery state, so that SOH estimation accuracy is improved.
Drawings
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a schematic diagram illustrating the identification of inflection points of a voltage curve according to the present invention;
FIG. 3 is a voltage profile distribution for multiple cycles;
fig. 4 shows a trend of the fourth inflection point of the voltage with the cycle period.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples. The present embodiment is implemented on the premise of the technical scheme of the present invention, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the following examples.
The lithium battery voltage-time curves of different electrode materials are obviously characterized, and a 'inflection point' of abrupt voltage curvature exists between voltage platforms. Taking lithium iron phosphate as an example, four obvious inflection points exist in a charging voltage-time curve, and when the battery ages, the inflection points show obvious change trend. There is a correlation between the decaying battery capacity and the voltage versus time curve. Based on the finding, the inventor creatively selects charging data around the second voltage platform, the selected data only comprises inflection points in a higher voltage range, a polynomial curve fitting method or other methods are adopted to fit the charging data to obtain a voltage curve, smoothing processing is carried out on the voltage curve, then the inflection points are identified, and feature quantities are constructed based on the identified inflection points, so that SOH (state of health) can be accurately predicted. Fig. 3 and 4 enumerate voltage-time curves of the same cell under different charging cycles of a certain type of lithium iron phosphate battery, it can be seen that there is a significant change related to the cell attenuation in the voltage curve near the second stage, so that the change can be extracted and represented, and a new SOH health factor is established.
Example 1
As shown in fig. 1, the embodiment provides a lithium battery SOH estimation method based on voltage curve inflection point identification, which includes the following steps:
step S1: a voltage-time curve is obtained.
Terminal voltage data of a constant current charging process monitored by the BMS system is collected, a voltage-time curve is formed by the terminal voltage data and corresponding sampling time, a plurality of obvious inflection points exist in the curve, and a specific voltage platform is represented between each inflection point. The range between inflection points is directly related to the phase transition of the electrode material and exhibits high sensitivity to aging of the battery.
Step S2: and extracting voltage curve key points based on inflection point identification.
In the complete charge-discharge curve, the voltage is changed in stages due to different stages of the electrochemical reaction process of the battery cell. The embodiment adopts an inflection point identification algorithm to extract voltage key feature points in a single charge-discharge period. The inflection point recognition algorithm is as follows:
the definition of the inflection point in the invention is the place where the curve direction changes the most in the voltage curve, so the point-by-point identification of the angle change of the curve at the place is adopted. Calculating the inclination angle of the vector relative to the x-axis on a two-dimensional plane by using the vector between two points separated by dBased on the angle change on the data point +.>As a measure, the place where the angle change is largest is identified. As shown in FIG. 2, in particularThe inflection point identification process includes the steps of:
step1, normalizing the coordinates of the charge-discharge voltage curve, so that the x and y values of the curve are both between 0 and 1;
step2, calculate each data point (x i ,y i ) The direction of the curve at the point changes:
(1) Point (x) i ,y i ) Two points before and after (x i-d ,y i-d ) And (x) i+d ,y i+d )；
(2) Calculate vector x i+d -x i ,y i+d -y i ]Is at an inclination angle of (2)Sum vector x i -x i-d ,y i -y i-d ]Dip angle of->
(3) Calculation Point (x) i ,y i ) At the point (x) i-d ,y i-d ) And (x) i+d ,y i+d ) Angle change of connection line
(4) Repeating (1) - (3) for each point of the charging curve, and calculating the angle change of each point at the adjacent d point
(5) Repeating (1) - (4) with d value of 1 to N, calculating each point at different pointsAverage value->As the angle change value of the point。
Step3, according to the angle change of each data point of the voltage curveA curve, identifying x at peaks and valleys in the curve i Which is in the voltage curve (x i ,y i ) Namely an inflection point:
(1)the peak value and the valley in the curve are identified, a function extremum identification method is used, and when the point x is i Is->The value is a peak value when the value is larger than the left and right adjacent values, and is a valley value when the value is smaller than the left and right adjacent values;
(2) Due to data noise, inA large number of peaks and valleys are identified on the curve, and are thus screened for absolute values greater than a threshold value ΔΦ bound And the width of the peaks and valleys exceeds Δx bound As an inflection point;
(3) The identified inflection point is atThe curve is the same as the abscissa of the voltage curve, and the inflection point is returned to the corresponding voltage curve;
step4, restoring the normalized coordinates of Step1, and returning the coordinate value of the inflection point after the restoration.
Step S3: and (6) constructing a feature vector.
In the complete charge-discharge curve, the terminal voltage is changed in stages due to different stages of the electrochemical reaction process of the battery cell. In the method, SOH health factors in a single charging period are constructed by the identified inflection points, and the SOH health factors in the current period of the charging process and the average temperature are combined into feature vectors.
The calculation method of the SOH health factor Hv is as follows:
wherein t4 represents a sampling time corresponding to a fourth inflection point of the full voltage curve of the charging period; t3 represents the sampling time corresponding to the third inflection point, Q t4 -Q t3 Representing the charge amount between the third inflection point and the fourth inflection point.
The temperature characteristic calculation method comprises the following steps:
where i is a cyclic sequence number, T ik Representing the temperature, T, of the kth sample point of the ith cycle i The average temperature for this cycle.
Step S4: and (3) constructing and training an SOH estimation offline model.
Extracting the cycle period of the same charging mode from the battery operation monitoring data, and labeling the corresponding cycle period sequence number to obtain a Hv sequence { Hv }, as shown in FIG. 3 1 ，Hv 2 ,…Hv n Average temperature sequence { T } 1 ，T 2 ,…T n }. Obtaining a characteristic vector F of the charging period i ＝(H vi ,T i ) Feature vector sequence F formed by n cycle periods 1∶n ＝(H v1∶n ,T 1∶n )。
According to the invention, an integrated learning model is adopted to establish a regression model of the maximum residual capacity of the battery cell, and an accurate classification effect is realized through iterative computation of a series of weak classifiers. In this embodiment, taking a random forest as an example, the method can also be replaced by XGBoost, lightGBM, and the algorithm steps of the regression model training are as follows:
step1, acquiring complete charging monitoring data of n cycles of m lithium batteries of the same type in a specific model in an off-line mode in the same constant current charging mode from an initial stage, wherein the complete charging monitoring data comprise a voltage sequence and a temperature sequence of each cycle and the maximum residual capacity of the battery in the cycle.
Step2, obtaining feature vectors and maximum residual capacity of n cycles of m lithium batteries of the same type to form a sample d= { (x) i ,y i ) I=1, 2,..m x n }, where x i ＝(H vi ,T i ,ind i ) And training a regression model of the maximum residual capacity. Wherein ind i For the corresponding number of cycle periods of the battery.
Step3, training by using a sample set to obtain K decision trees, and adding samples x according to the K decision trees i The corresponding label predicts:wherein f= { F (x) =w q(x) ' represents the decision tree vs. function space, W q(x) Representing mapping each node in the tree to a value.
Step4, adjusting model parameters by using a loss function.
Step S5: and estimating the maximum residual capacity of the online battery.
Calculating to obtain a characteristic vector of the current battery in the current charging period, inputting a regression model of the maximum residual capacity of the battery core obtained by offline training to obtain a corresponding model outputWherein->And (5) finishing SOH estimation for the maximum residual capacity predicted value of the nth cycle period of the battery.
The above-described method, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Example 2
The present embodiment provides an electronic device including one or more processors, a memory, and one or more programs stored in the memory, the one or more programs including instructions for performing the lithium battery SOH estimation method based on voltage curve inflection point identification as described in embodiment 1.
The foregoing describes in detail preferred embodiments of the present invention. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the invention by one of ordinary skill in the art without undue burden. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by the person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
Claims (10)
1. The lithium battery SOH estimation method based on voltage curve inflection point identification is characterized by comprising the following steps of:
collecting charging data around a second voltage platform in the constant current charging process monitored by the BMS system, and obtaining terminal voltage and corresponding sampling time to form a voltage-time curve;
identifying inflection points by identifying the angular change of the curve at each point by point, and constructing SOH health factors in a single charging period by the identified inflection points;
combining SOH health factors of the current period of the charging process with the average temperature to form a feature vector;
taking the feature vector as input of a regression model of the maximum residual capacity of the battery cell obtained through offline training, predicting to obtain a predicted value of the maximum residual capacity of the next cycle period, and finishing SOH estimation;
the calculation method of the SOH health factor Hv is as follows:
wherein t4 represents a sampling time corresponding to a fourth inflection point of the full voltage curve of the charging period; t3 represents the sampling time corresponding to the third inflection point, Q t4 -Q t3 Representing the charge quantity between the third inflection point and the fourth inflection point;
the average temperature was calculated as follows:
where i is a cyclic sequence number, T ik Representing the temperature, T, of the kth sample point of the ith cycle i The average temperature for this cycle.
2. The lithium battery SOH estimation method based on voltage curve inflection point identification according to claim 1, characterized in that the identification of inflection point by identifying the angular change of the curve at each point by point is specifically:
step1, normalizing the voltage-time curve coordinates of charge and discharge to ensure that the x and y values of the curve are all between 0 and 1;
step2, calculate each data point (x i ,y i ) The direction of the curve at the point is changed, and the angle change of each data point is obtained;
step3, based on the angular change of each data point on the voltage-time curveA curve identifying said->X at peaks and valleys in the curve i Which is in the voltage-time curve (x i ,y i ) Namely an inflection point;
step4, restoring the normalized coordinates of Step1, and returning the coordinate value of the inflection point after the restoration.
3. The method for estimating SOH of a lithium battery based on voltage curve inflection point identification of claim 2, wherein said each data point angle change is obtained by:
201 Get the point (x) i ,y i ) Two points before and after (x i-d ,y i-d ) And (x) i+d ,y i+d )；
202 Calculating vector x i+d -x i ,y i+d -y i ]Is at an inclination angle of (2)Sum vector x i -x i-d ,y i -y i-d ]Dip angle of->
203 Calculating the point (x) i ,y i ) At the point (x) i-d ,y i-d ) And (x) i+d ,y i+d ) Angle change of connection line
204 Repeating steps 201) -203) for each point of the voltage-time curve, calculating the angular change of each point at the adjacent d-point
205 Repeating steps 201) to 204) with d being 1 to N, and calculating each point at different points Average value->As the value of the change in angle of the point.
4. The method for estimating SOH of lithium battery based on voltage curve inflection point identification according to claim 2, wherein Step3 uses a function extremum identification method for screeningPeaks and valleys in the curve.
5. The method for estimating SOH of lithium battery based on inflection point identification of voltage curve according to claim 2, wherein in Step3, the absolute value of the screening is greater than a threshold ΔΦ bound And the width of the peaks and valleys exceeds a threshold Δx bound As inflection points.
6. The lithium battery SOH estimation method based on voltage curve inflection point identification according to claim 1, wherein an integrated learning model is adopted to build the regression model of the maximum remaining capacity of the battery cell.
7. The lithium battery SOH estimation method based on voltage curve inflection point identification of claim 1, wherein the offline training process of the battery cell maximum remaining capacity regression model comprises:
the constant-current charging mode with the same current is adopted, complete charging monitoring data of n continuous cycle periods of m same lithium batteries of a specific model from an initial stage are acquired offline, the complete charging monitoring data comprise a voltage sequence and a temperature sequence of each period and the maximum residual capacity of the battery of the period, feature vectors and the maximum residual capacity of the n cycle periods of the m same lithium batteries are acquired by utilizing a feature extraction algorithm to form a training sample set D = { (D) i ,z i ) I=1, 2,..m.n }, where d i ＝(H vi ,T i ,ind i ) As a feature vector, z i For the corresponding label, i.e. maximum remaining capacity of battery, H vi ,T i SOH health factor and average temperature, ind respectively i A corresponding cycle number for the battery;
training by using a training sample set to obtain K decision trees, and adding d according to the K decision trees i The corresponding label predicts:wherein f= { F (d) =w q(d) ' represents the decision tree vs. function space, W q(d) Representing mapping each node in the tree to a value;
model parameters are adjusted using the loss function.
8. The method of claim 6, wherein the ensemble learning model comprises a random forest, XGBoost, or LightGBM.
9. A computer-readable storage medium comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for performing the lithium battery SOH estimation method based on voltage curve inflection point identification of any of claims 1-8.
10. An electronic device, comprising:
one or more processors;
a memory; and
one or more programs stored in a memory, the one or more programs comprising instructions for performing the lithium battery SOH estimation method based on voltage curve inflection point identification of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111285782.4A CN114035095B (en) | 2021-11-02 | 2021-11-02 | Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111285782.4A CN114035095B (en) | 2021-11-02 | 2021-11-02 | Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114035095A CN114035095A (en) | 2022-02-11 |
| CN114035095B true CN114035095B (en) | 2023-12-22 |
Family
ID=80135972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111285782.4A Active CN114035095B (en) | 2021-11-02 | 2021-11-02 | Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114035095B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117388734A (en) | 2022-07-05 | 2024-01-12 | 通用汽车环球科技运作有限责任公司 | Health-based operation for a vehicle power supply |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108333518A (en) * | 2017-01-18 | 2018-07-27 | 福特全球技术公司 | Battery health is assessed |
| CN109031153A (en) * | 2018-10-16 | 2018-12-18 | 北京交通大学 | A kind of health status On-line Estimation method of lithium ion battery |
| 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 |
| CN111323719A (en) * | 2020-03-18 | 2020-06-23 | 北京理工大学 | Method and system for online determination of health state of power battery pack of electric automobile |
| CN111381178A (en) * | 2018-12-28 | 2020-07-07 | 奥动新能源汽车科技有限公司 | Method and system for estimating health state of lithium ion battery |
| CN111443293A (en) * | 2020-03-30 | 2020-07-24 | 青岛大学 | Lithium battery state of health (SOH) estimation method based on data driving |
| CN112034356A (en) * | 2020-09-09 | 2020-12-04 | 哈尔滨工业大学 | GP-UKF-based online SOH estimation method for electric vehicle power battery |
| 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 |
| 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 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10393813B2 (en) * | 2013-08-27 | 2019-08-27 | The Regents Of The University Of Michigan | On-board state of health monitoring of batteries using incremental capacity analysis |
| WO2020130422A1 (en) * | 2018-12-21 | 2020-06-25 | Samsung Electronics Co., Ltd. | Method and system for predicting onset of capacity fading in a battery |
| CN110967631B (en) * | 2019-05-17 | 2021-03-30 | 宁德时代新能源科技股份有限公司 | SOH correction method and apparatus, battery management system, and storage medium |
| WO2020251854A1 (en) * | 2019-06-14 | 2020-12-17 | Cummins Inc. | Methods and devices for determining battery state of health using incremental capacity analysis and support vector regression |
-
2021
- 2021-11-02 CN CN202111285782.4A patent/CN114035095B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108333518A (en) * | 2017-01-18 | 2018-07-27 | 福特全球技术公司 | Battery health is assessed |
| CN109031153A (en) * | 2018-10-16 | 2018-12-18 | 北京交通大学 | A kind of health status On-line Estimation method of lithium ion battery |
| CN111381178A (en) * | 2018-12-28 | 2020-07-07 | 奥动新能源汽车科技有限公司 | Method and system for estimating health state of lithium ion battery |
| 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 |
| CN111323719A (en) * | 2020-03-18 | 2020-06-23 | 北京理工大学 | Method and system for online determination of health state of power battery pack of electric automobile |
| CN111443293A (en) * | 2020-03-30 | 2020-07-24 | 青岛大学 | Lithium battery state of health (SOH) estimation method based on data driving |
| CN112034356A (en) * | 2020-09-09 | 2020-12-04 | 哈尔滨工业大学 | GP-UKF-based online SOH estimation method for electric vehicle power battery |
| 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 |
| 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 |
Non-Patent Citations (1)
| Title |
|---|
| 过充电对锂离子电池安全的影响研究;柳升龙;武国良;;黑龙江科技信息(第13期);第46-47页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114035095A (en) | 2022-02-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tian et al. | Deep neural network battery charging curve prediction using 30 points collected in 10 min | |
| Wei et al. | Remaining useful life prediction and state of health diagnosis for lithium-ion batteries using particle filter and support vector regression | |
| Ma et al. | A novel data-model fusion state-of-health estimation approach for lithium-ion batteries | |
| CN110031770B (en) | Method for rapidly obtaining capacity of all single batteries in battery pack | |
| US11422192B2 (en) | Method and apparatus for estimating state of health of battery | |
| CN110058178A (en) | A kind of lithium battery method for detecting health status and system | |
| CN110133525B (en) | Lithium ion battery health state estimation method applied to battery management system | |
| CN114167284B (en) | Lithium battery RUL prediction method and equipment based on BMS big data and integrated learning | |
| EP3721499A1 (en) | Method and system for online assessing state of health of a battery | |
| CN113924502A (en) | Method and apparatus for determining battery state of health using incremental capacity analysis and support vector regression | |
| WO2015031437A1 (en) | On-board state of health monitoring of batteries using incremental capacity analysis | |
| CN110673037B (en) | Battery SOC estimation method and system based on improved simulated annealing algorithm | |
| Jiang et al. | An aging-aware soc estimation method for lithium-ion batteries using xgboost algorithm | |
| Gasper et al. | Predicting battery capacity from impedance at varying temperature and state of charge using machine learning | |
| CN111142025A (en) | Battery SOC estimation method and device, storage medium and electric vehicle | |
| CN113777510A (en) | Lithium battery state of charge estimation method and device | |
| CN114184972B (en) | Automatic estimation method and equipment for SOH of battery by combining data driving and electrochemical mechanism | |
| CN115201686B (en) | Lithium ion battery health state assessment method under incomplete charge and discharge data | |
| CN114035095B (en) | Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification | |
| CN114397577A (en) | New energy automobile lithium battery health state assessment method based on ASTUKF-GRA-LSTM model | |
| CN113743664B (en) | Lithium battery state estimation method and system based on random fragment data | |
| CN113125965B (en) | Method, device and equipment for detecting lithium separation of battery and storage medium | |
| CN113820615A (en) | Battery health degree detection method and device | |
| Mills et al. | Estimating state of charge and state of health of rechargable batteries on a per-cell basis | |
| CN117686937B (en) | Method for estimating health state of single battery in battery 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |