CN113777494A - Lithium battery capacity diving turning point identification method based on geometric feature fusion decision - Google Patents
Lithium battery capacity diving turning point identification method based on geometric feature fusion decision Download PDFInfo
- Publication number
- CN113777494A CN113777494A CN202110980619.3A CN202110980619A CN113777494A CN 113777494 A CN113777494 A CN 113777494A CN 202110980619 A CN202110980619 A CN 202110980619A CN 113777494 A CN113777494 A CN 113777494A
- Authority
- CN
- China
- Prior art keywords
- capacity
- lithium battery
- diving
- point
- turning point
- 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]
- G01R31/367—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]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Algebra (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Secondary Cells (AREA)
Abstract
The invention relates to a lithium battery capacity diving turning point identification method based on geometric feature fusion decision, which comprises the following steps: 1) acquiring a capacity decline curve of a lithium battery to be identified on line, and performing smooth denoising and double-axis normalization processing to obtain a normalized lithium battery capacity decline curve; 2) connecting the initial point C on the normalized lithium battery capacity decline curve1And an end point C2Forming a completely linear aging reference line C1C2(ii) a 3) Calculating the degree of deviation of points on the normalized lithium battery capacity decline curve from an aging reference line, comparing the deviation maximum value with a capacity diving early warning threshold value beta, and judging whether capacity diving occurs in real time; 4) triggering the early warning of the capacity diving of the lithium battery when the capacity diving occurs, and corresponding to the maximum deviation degreeAs the identified lithium ion battery capacity trip turning point. Compared with the prior art, the method has the advantages of accurate identification, automatic batch processing and the like.
Description
Technical Field
The invention relates to the technical field of lithium battery monitoring, in particular to a method for on-line early warning of lithium battery capacity diving and identifying a turning point based on a geometric feature fusion decision, which is used for on-line aging state evaluation and gradient utilization of lithium ion batteries in the fields of electric automobiles, hybrid electric vehicles and energy storage.
Background
The problem of traditional energy exhaustion and environmental pollution causes high attention to the field of new energy all over the world, and China continuously increases the supporting force on new energy automobiles and develops a series of policies of increasing research and development investment, infrastructure construction, market scale expansion and the like. Among various new energy batteries, the lithium ion battery has the excellent comprehensive advantages of small volume, high energy density, long service life, zero emission, no pollution and the like, and is widely applied to various industries and fields including new energy automobiles. However, in practical application, due to the common influence of self chemical change and specific use conditions, the service life of the lithium ion battery often cannot reach a preset service life value, and system faults and safety accidents caused by the failure of the service life of the battery pose great threats to the personal and property safety of people.
The aging of the lithium ion battery exists objectively in the whole life cycle, and is a complex process with various mechanisms influencing the aging mechanism, including the formation and growth of a solid electrolyte intermediate phase, the mechanical rupture of an electrode material, the dissolution of a transition metal on a positive electrode, the precipitation of metal lithium on a negative electrode and the like. In most cases, the capacity retention rate of lithium ion batteries remains linearly dependent on charge throughput during cyclic charge and discharge. However, in recent years, many studies have reported that the aging behavior is contrary to the above aging characteristics, that is, the capacity retention rate curve shape of a part of the battery sample shows linear degradation in a period of time, but shows a behavior that the capacity degradation rate rapidly increases after a certain critical point is exceeded. When the capacity retention rate of the lithium ion battery suddenly and rapidly declines, the battery is considered to have a capacity diving phenomenon, and the critical point of the phenomenon is the turning point of the capacity diving.
In the actual use process of the battery, the on-line early warning of capacity diving and the accurate identification of the turning point should attract the important attention of people, and the specific reasons are as follows:
the existing research shows that when a battery sample generates capacity water jump, large-area lithium precipitation behavior in the battery is often accompanied, and the formed lithium dendrite can cause the tearing of a diaphragm, so that short circuit and instant battery failure are caused, and the lithium dendrite becomes an important cause of thermal runaway. Therefore, it is desirable to prevent the occurrence of thermal runaway by replacing the battery at the time of or shortly after the occurrence of a water jump.
However, in view of the above problems, the existing analysis and processing methods for the on-line early warning of the capacity of the lithium battery and the identification of the turning point are not complete.
Aiming at the fact whether the volume diving occurs in the diving sample and the identification and marking of the turning point, the existing method is basically offline judgment and mostly depends on manual judgment, namely, after a data curve is visualized, whether the volume diving occurs in the battery is judged by naked eyes, and the judgment method has the following three defects:
firstly, manual identification and marking are greatly influenced by subjective factors, different identification and marking personnel, identification scales and the like can influence the identification result, and interference is brought to identification and classification of the capacity diving turning point.
Secondly, a large amount of labor cost and time cost are needed depending on a manual distinguishing method, batch operation cannot be carried out, and the labeling efficiency of the sample is influenced.
And thirdly, offline judgment cannot timely early warn and process the battery with capacity jump, so that the optimal replacement opportunity is missed, even if timely judgment is not available in severe cases, and the lithium battery has thermal runaway.
Aiming at the capacity diving online early warning and the identification of turning points, no mature method is put into use at present. In the using process of the battery, due to the fact that a real-time detection mechanism is not available, once a capacity water-jumping point occurs, early warning cannot be timely carried out, and the battery can still be continuously used until the battery is completely invalid or fails. Therefore, the echelon utilization value of the battery can be reduced, the battery can even be seriously disabled, and great economic loss is brought.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a method for identifying the jump turning point of the capacity of the lithium battery based on a geometric feature fusion decision.
The purpose of the invention can be realized by the following technical scheme:
a lithium battery capacity diving turning point identification method based on geometric feature fusion decision comprises the following steps:
1) acquiring a capacity decline curve of a lithium battery to be identified on line, and performing smooth denoising and double-axis normalization processing to obtain a normalized lithium battery capacity decline curve;
2) connecting the initial point C on the normalized lithium battery capacity decline curve1And an end point C2Forming a completely linear aging reference line C1C2；
3) Calculating the degree of deviation of points on the normalized lithium battery capacity decline curve from an aging reference line, comparing the deviation maximum value with a capacity diving early warning threshold value beta, and judging whether capacity diving occurs in real time;
4) and triggering the capacity diving early warning of the lithium battery when the capacity diving occurs, and taking a point corresponding to the maximum deviation degree as the identified capacity diving turning point of the lithium battery.
In the step 1), smoothing and denoising are performed in a filtering mode, specifically, polynomial fitting of a sliding window is performed in a time domain through a local least square method based on a convolution principle of least squares.
Performing a sliding window polynomial fit specifically includes the steps of:
11) selecting a window size w, and extracting partial data from original data according to the window size;
12) performing least square fitting on the data in the window by utilizing a p-order polynomial function;
13) replacing the original data point at the center of the window with a predicted point obtained by polynomial fitting, moving the window forward by an interval, and repeating steps 11) -13) until data smoothing is completed.
The control of the data smoothness degree is realized by adjusting the sizes of w and p, and the following conditions are met:
(1) the window size w is odd and does not exceed the length of the original data;
(2) the polynomial order p is less than or equal to w-1.
In the step 1), the biaxial normalization processing specifically comprises:
and uniformly mapping the capacity and the cycle number of the lithium battery to be identified into a [0,1] interval so as to convert the capacity and the cycle number into dimensionless data with the same order of magnitude.
In the step 2), a completely linear aging reference line C1C2The expression of (a) is:
wherein x isnFor the number of cycles in the fully linear aging process,
the capacity corresponding to the corresponding cycle in the complete linear aging process.
In the step 3), the degree of deviation of the point on the normalized lithium battery capacity decline curve from the aging reference line is defined as follows:
connecting a point L and an initial point C on the normalized lithium battery capacity decline curve1And an end point C2Three points form a triangle LC1C2Connecting the point L and the bottom edge C1C2The length α of the vertical line segment LM is defined as the degree of deviation from the aging reference line, and the larger the length α of the vertical line segment LM represents the larger the degree of deviation of the point from the aging reference line.
In the step 3), the step of judging whether the volume diving occurs is specifically as follows:
and setting a capacity diving early warning threshold value beta, judging that the capacity diving occurs in the lithium battery to be identified when the maximum deviation degree is greater than the capacity diving early warning threshold value, and otherwise, judging that the capacity diving does not occur.
Determining a volume diving early warning threshold value alpha according to historical data and expert knowledge, and specifically comprising the following steps:
selecting N lithium ion batteries artificially marked as capacity jump occurring and capacity jump not occurring from historical capacity decline data, and enabling the maximum value alpha of the off-line aging reference line distance in the corresponding capacity decline curve1,α2,…,αNAs a sample, in the sample, the minimum value of the α values corresponding to all occurrences of the volume saltwater is used as the upper range limit, the maximum value of the α values corresponding to all the occurrences of the volume saltwater is used as the lower range limit, and the volume saltwater warning threshold β is the median of the range.
The online capacity diving early warning method specifically comprises the following steps:
when the battery to be identified circulates N0During the circle, the maximum value alpha of the distance between the corresponding actual capacity fade point and the linear aging reference line is obtained0Comparing alpha in real time0And a volume diving early warning threshold value beta when alpha is0When beta is greater than beta, the alarm is triggered directly, and alpha is generated at the moment0The corresponding cycle number is the turning point of the lithium battery capacity diving, when alpha is0If the cell number is less than beta, the cell to be identified continues to circulate, and the termination point C is updated2So as to obtain a new capacity decline curve, and at the moment, the battery to be identified circulates N1Circle and calculate the maximum value alpha of the distance from the linear aging reference line at the moment1Comparison of α1And β, performing the diving judgment again, and repeating the steps.
Compared with the prior art, the invention has the following advantages:
firstly, the invention provides an online deviation linear aging degree value calculated based on a geometric method by adding a completely linear capacity decline reference line, can effectively represent the capacity water-jumping phenomenon in the battery decline process, and has more stable calculation result.
The capacity diving online early warning method provided by the invention can monitor and early warn whether the battery generates capacity diving in real time, and the monitoring method can realize automatic batch processing through an algorithm.
The capacity diving turning point on-line identification method provided by the invention can accurately judge the aging state of the battery, is beneficial to subsequent echelon utilization and improves the full life cycle value of the battery.
Drawings
Fig. 1 is a flow chart of the lithium battery capacity diving online early warning and turning point identification based on geometric feature fusion.
Fig. 2 is a schematic diagram of smoothing noise reduction and biaxial normalization preprocessing and adding reference lines, wherein fig. 2a is a process of smoothing a battery capacity fading data sample, and fig. 2b is a normalized capacity fading curve after adding a complete linear aging reference line.
Fig. 3 is a schematic diagram of a distance maximum value of a deviation of a calculated actual capacity fading point from a complete linear fading curve based on a geometric feature fusion decision according to the present invention.
Fig. 4 is a schematic diagram of the geometric feature fusion online real-time calculation result, threshold comparison and turning point identification of the diving battery sample.
Fig. 5 is a schematic diagram of the geometric feature fusion online real-time calculation result, threshold comparison and turning point identification of a normal battery sample and a diving battery sample based on the strategy of the present invention, wherein fig. 5a is a non-diving sample b3c28, and fig. 5b is a diving sample b1c 24.
Detailed Description
In order to make the objects, technical solutions and novel points of the present invention more clear, the present invention is further explained with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the invention provides a method for on-line early warning of lithium battery capacity diving and turning point identification based on geometric feature fusion decision, which comprises the following steps:
firstly, based on a filtering method and a biaxial normalization method, the filtering method takes Savitzky-Golay as an example, filtering smoothing and biaxial normalization processing are carried out on a lithium battery capacity fading curve to obtain a lithium battery after pretreatmentPool capacity decline curve, including starting point C1And an end point C2；
Then, on the basis of the normalized capacity decline curve, adding a starting point C1And an end point C2A fully linear capacity fade curve of composition;
then, calculating the degree of the actual capacity fading point deviating from the complete linear fading curve, finding out the maximum value of the deviation, and comparing the maximum value with a capacity diving early warning threshold value to determine whether capacity diving occurs;
and finally, triggering the early warning of the lithium battery capacity diving when the lithium battery capacity diving is determined to occur, wherein the point corresponding to the maximum deviation degree is the turning point for identifying the lithium battery capacity diving, updating the capacity data termination point when the lithium battery capacity diving is determined not to occur, and repeating the steps.
The above method is described in detail with reference to fig. 2 to 5.
Step 1: data preprocessing and reference line addition
The pre-processing aims at two main points, firstly, due to the influence of errors of a measuring instrument and a test environment, in the process of carrying out a lithium ion battery life test, collected original data can inevitably generate certain random noise, data which do not meet quality requirements in a data set are found out through mathematical statistics or defined rules to be removed, interference brought to subsequent research due to data quality problems is reduced, secondly, due to the fact that different battery data samples have great inconsistency of the capacity and the cycle number, certain calculated characteristic values are not comparable, meanwhile inconsistency is brought to the selection of an early warning threshold, in order to carry out unification processing, the capacity and the cycle number are mapped into a [0,1] interval through a double-axis normalization method, on the basis of not changing the change trend of an original curve, so that the same order of magnitude is obtained, which facilitates batch operation.
The specific operation steps for smoothing are as follows:
defining the original battery sample data set as DC, and arranging according to the sequence of cycle, i.e. arranging
DC＝{(x0,y0),(x1,y1),…(xN,yN)|0＜x1＜x2＜…＜xN}
Wherein, for
xiCorresponding to the number of cycles in the aging process, yiCharging or discharging capacity of the battery for the corresponding cycle.
Since the capacity retention rate of the battery is continuously degraded along with time, certain correlation exists between data, in order to fully consider the time correlation between data points, the example takes Savitzky-Golay filtering as an example to smooth a battery capacity attenuation data sample, and selects the window size w and the polynomial order p according to the data sample so as to give consideration to better smoothing effect and capture curve shape characteristics. The smoothed capacity fade data set is denoted DCfAs shown in fig. 2 a.
Wherein, for
xiCorresponding to the number of cycles in the aging process,
and charging or discharging capacity of the battery after corresponding cycle smoothing.
The steps for biaxial normalization were as follows:
to improve the applicability of the algorithm to different batteries, the smoothed battery samples DCfRespectively normalizing the cycles and the capacities to convert the cycles and the capacities into data between 0 and 1, thereby obtaining a capacity attenuation data set NDC after biaxial normalizationfI.e. by
Wherein the content of the first and second substances,
is normalized for the number of cycles in the aging process after the aging, and
is the charge or discharge capacity of the corresponding cycle after normalization, and
the steps for adding the fully linear aged reference line are as follows:
starting point C of battery capacity after normalization1Coordinate (0,1), end point C2The coordinate is (1,0), and the added complete linear aging reference line is C1C2As shown in FIG. 2b, reference line C1C2The expression of (a) is as follows,
wherein x isnFor the number of cycles in the additive fully linear aging process,
to addCorresponding to the corresponding cycle during the full linear aging process.
Step 2: on-line real-time calculation of actual capacity fading point deviating from maximum value of complete linear fading curve distance
As shown in FIG. 3, a dual-axis normalized capacity-decay data set NDC of a diving battery samplefThe corresponding curve and the added fully linear aging reference line construct an approximately arched structure, and both can be regarded as the structure from the starting point C1(0,1) to termination point C2(1,0), the capacity degradation is completely linear if the battery sample is aged according to the dotted path. According to the fading of the actual aging path, the capacity fading is stable but slower in the early period compared with the reference aging path, and the obvious capacity advantage is accumulated, but the capacity fading rate is obviously increased in the later period, so that the capacity advantage is rapidly faded away, and finally the end point C is reached at the same time2。
From the geometric features, the value of LM can be calculated for each point on the curve, i.e.
LM＝{LM1,LM2,…LMN}
And the actual capacity fade point deviates from the fully linear fade curve by the maximum value alpha as calculated below,
α＝max{LM1,LM2,…LMN}
and step 3: sample diving risk real-time early warning
Step 31: the selection of the quantitative early warning threshold value beta comprises the following steps:
the quantitative threshold value depends on the result of manual marking, therefore, firstly, in all samples to be marked, a part of samples are randomly sampled to be used as a training set
Wherein alpha isiFor the actual capacity fade point of the ith training sample deviating from the maximum of the distance of the perfect linear fade curve,
for the data set of the ith sample, N is trainingNumber of training samples.
Finally determining the range of the threshold value (a hollow area exists between the maximum values alpha of the distances between all the diving batteries and the non-diving batteries, the area is taken as the range of the threshold value), and taking the median value of the range interval of the threshold value as beta by comparing the deviation of the actual capacity fading points of the diving batteries and the non-diving batteries in the sample from the maximum values alpha of the distances between the completely linear fading curves.
Step 32: the real-time early warning step of the risk of diving is as follows:
fig. 4 is a schematic diagram of the on-line real-time water-diving risk early warning of the present invention. When the cell sample has cycled N, as shown in FIG. 40During the circle, the distance maximum value alpha of the corresponding actual capacity fading point deviating from the complete linear fading curve at the moment can be obtained through the steps0Alpha obtained by calculating in real time0Comparing with the early warning threshold value beta when alpha is0When beta is greater than beta, the alarm is triggered directly, and alpha is generated at the moment0The corresponding cycle number is the turning point of the lithium battery capacity diving; when alpha is0If beta is less than beta, the battery continues to circulate, and the end point C of the battery sample is updated again2To obtain a new battery sample, which is now cycled by N1Circle and according to the calculation, alpha can be obtained1And repeating the steps at the moment, judging the diving of the battery again, and so on.
Examples
The A123 data set disclosed by Severson et al is adopted to verify the feasibility and effectiveness of the provided method for early warning of the capacity of the lithium ion battery in the case of diving and identifying the turning point.
Two cases of cell samples, numbered b3c28 and b1c24, were selected for analysis in the data set, including one non-saltwater sample, b3c28, and one saltwater sample, b1c 24.
After the steps of data preprocessing, reference line adding, calculation of the distance maximum value alpha of the actual capacity fading point deviating from the complete linear fading curve and the like are carried out, the real-time dynamic results of the two samples are shown in fig. 5. Wherein the threshold value of the lithium battery capacity diving is set to 0.18 by considering the historical data condition and expert knowledge of the selected data set.
As shown in fig. 5 (a), the dashed line indicates the selected pre-warning threshold, no significant diving behavior of the battery b3c28 has occurred, and the calculated α has not exceeded the set threshold of 0.18. The result shows that the capacity diving online early warning method based on the geometric feature fusion decision can effectively monitor the diving phenomenon of the battery.
As shown in fig. 5 (b), the dashed line also represents the selected early warning threshold, when the battery b1c24 circulates to 583 circles, its actual capacity degradation point deviates from the full linear degradation curve by a distance maximum value α greater than the set threshold value 0.18, indicating that the battery has capacity diving, and the corresponding 583 th circle is also the turning point of the battery capacity diving. The results show that the lithium battery capacity diving online early warning and turning point identification method based on the geometric feature fusion decision can effectively and timely find and early warn the capacity diving behavior of the battery, thereby stopping the continuous operation of the battery in advance, preventing the continuous deterioration of the service life of the battery, and improving the use safety of the battery and the echelon utilization value of the battery.
Although the present invention has been described in detail hereinabove, the present invention is not limited thereto, and various modifications can be made by those skilled in the art in light of the principle of the present invention. Thus, modifications made in accordance with the principles of the present invention should be understood to fall within the scope of the present invention.
Claims (10)
1. A lithium battery capacity diving turning point identification method based on geometric feature fusion decision is characterized by comprising the following steps:
1) acquiring a capacity decline curve of a lithium battery to be identified on line, and performing smooth denoising and double-axis normalization processing to obtain a normalized lithium battery capacity decline curve;
2) connecting the initial point C on the normalized lithium battery capacity decline curve1And an end point C2Forming a completely linear aging reference line C1C2；
3) Calculating the degree of deviation of points on the normalized lithium battery capacity decline curve from an aging reference line, comparing the deviation maximum value with a capacity diving early warning threshold value beta, and judging whether capacity diving occurs in real time;
4) and triggering the capacity diving early warning of the lithium battery when the capacity diving occurs, and taking a point corresponding to the maximum deviation degree as the identified capacity diving turning point of the lithium battery.
2. The method for identifying the jump turning point of the capacity of the lithium battery based on the geometric feature fusion decision as claimed in claim 1, wherein in the step 1), smoothing and denoising are performed in a filtering mode, specifically, based on a least square convolution principle, and polynomial fitting of a sliding window is performed in a time domain by a local least square method.
3. The method for identifying the jump turning point of the capacity of the lithium battery based on the geometric feature fusion decision as claimed in claim 2, wherein the polynomial fitting of the sliding window specifically comprises the following steps:
11) selecting a window size w, and extracting partial data from original data according to the window size;
12) performing least square fitting on the data in the window by utilizing a p-order polynomial function;
13) replacing the original data point at the center of the window with a predicted point obtained by polynomial fitting, moving the window forward by an interval, and repeating steps 11) -13) until data smoothing is completed.
4. The method for identifying the jump turning point of the capacity of the lithium battery based on the geometric feature fusion decision as claimed in claim 3, wherein the control of the data smoothness degree is realized by adjusting the sizes of w and p, and the following conditions are satisfied:
(1) the window size w is odd and does not exceed the length of the original data;
(2) the polynomial order p is less than or equal to w-1.
5. The method for identifying the jump turning point of the capacity of the lithium battery based on the geometric feature fusion decision as claimed in claim 1, wherein in the step 1), the biaxial normalization processing specifically comprises:
and uniformly mapping the capacity and the cycle number of the lithium battery to be identified into a [0,1] interval so as to convert the capacity and the cycle number into dimensionless data with the same order of magnitude.
6. The method for identifying the jump turning point of the capacity of the lithium battery based on the geometric feature fusion decision as claimed in claim 1, wherein in the step 2), the aging reference line C is completely linear1C2The expression of (a) is:
wherein x isnFor the number of cycles in the fully linear aging process,
the capacity corresponding to the corresponding cycle in the complete linear aging process.
7. The method for identifying the jump turning point of the lithium battery capacity based on the geometric feature fusion decision as claimed in claim 1, wherein in the step 3), the degree of deviation of a point on the normalized lithium battery capacity degradation curve from the aging reference line is defined as follows:
connecting a point L and an initial point C on the normalized lithium battery capacity decline curve1And an end point C2Three points form a triangle LC1C2Connecting the point L and the bottom edge C1C2The length α of the vertical line segment LM is defined as the degree of deviation from the aging reference line, and the larger the length α of the vertical line segment LM represents the larger the degree of deviation of the point from the aging reference line.
8. The method for identifying the jump turning point of the lithium battery capacity based on the geometric feature fusion decision as claimed in claim 7, wherein the step 3) of judging whether the jump of the capacity occurs is specifically as follows:
and setting a capacity diving early warning threshold value beta, judging that the capacity diving occurs in the lithium battery to be identified when the maximum deviation degree is greater than the capacity diving early warning threshold value, and otherwise, judging that the capacity diving does not occur.
9. The method for identifying the lithium battery capacity diving turning point based on the geometric feature fusion decision as claimed in claim 8, wherein the capacity diving early warning threshold β is determined according to historical data and expert knowledge, and specifically comprises the following steps:
selecting N lithium ion batteries artificially marked as capacity jump occurring and capacity jump not occurring from historical capacity decline data, and enabling the maximum value alpha of the off-line aging reference line distance in the corresponding capacity decline curve1,α2,…,αNAs a sample, in the sample, the minimum value of the α values corresponding to all occurrences of the volume saltwater is used as the upper range limit, the maximum value of the α values corresponding to all the occurrences of the volume saltwater is used as the lower range limit, and the volume saltwater warning threshold β is the median of the range.
10. The method for identifying the lithium battery capacity diving turning point based on the geometric feature fusion decision as claimed in claim 9, wherein the online capacity diving early warning specifically comprises the following steps:
when the battery to be identified circulates N0During the circle, the maximum value alpha of the distance between the corresponding actual capacity fade point and the linear aging reference line is obtained0Comparing alpha in real time0And a volume diving early warning threshold value beta when alpha is0When beta is greater than beta, the alarm is triggered directly, and alpha is generated at the moment0The corresponding cycle number is the turning point of the lithium battery capacity diving, when alpha is0If the cell number is less than beta, the cell to be identified continues to circulate, and the termination point C is updated2So as to obtain a new capacity decline curve, and at the moment, the battery to be identified circulates N1Circle and calculate the maximum value alpha of the distance from the linear aging reference line at the moment1Comparison of α1And β, performing the diving judgment again, and repeating the steps.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110980619.3A CN113777494A (en) | 2021-08-25 | 2021-08-25 | Lithium battery capacity diving turning point identification method based on geometric feature fusion decision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110980619.3A CN113777494A (en) | 2021-08-25 | 2021-08-25 | Lithium battery capacity diving turning point identification method based on geometric feature fusion decision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113777494A true CN113777494A (en) | 2021-12-10 |
Family
ID=78839227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110980619.3A Pending CN113777494A (en) | 2021-08-25 | 2021-08-25 | Lithium battery capacity diving turning point identification method based on geometric feature fusion decision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113777494A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116541718A (en) * | 2023-07-07 | 2023-08-04 | 江苏正力新能电池技术有限公司 | Training method, device, equipment and medium for battery capacity jump recognition model |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103439666A (en) * | 2013-09-02 | 2013-12-11 | 北京航空航天大学 | Geometric method for evaluating capacity recession of lithium ion battery |
| CN106295003A (en) * | 2016-08-11 | 2017-01-04 | 北京航空航天大学 | A kind of based on the reconstruct of Degradation path coordinate and the service life of lithium battery Forecasting Methodology of multiple linear regression |
| CN109596983A (en) * | 2018-11-19 | 2019-04-09 | 清华大学 | The prediction technique that capacity is dived during a kind of cell degradation |
| CN109946612A (en) * | 2019-04-01 | 2019-06-28 | 北京交通大学 | A kind of ternary capacity of lithium ion battery acceleration decline turning point recognition methods |
| CN112327191A (en) * | 2020-10-21 | 2021-02-05 | 北京航空航天大学 | Battery diving probability evaluation method based on geometric feature fusion decision |
| CN112327193A (en) * | 2020-10-21 | 2021-02-05 | 北京航空航天大学 | Lithium battery capacity diving early warning method |
-
2021
- 2021-08-25 CN CN202110980619.3A patent/CN113777494A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103439666A (en) * | 2013-09-02 | 2013-12-11 | 北京航空航天大学 | Geometric method for evaluating capacity recession of lithium ion battery |
| CN106295003A (en) * | 2016-08-11 | 2017-01-04 | 北京航空航天大学 | A kind of based on the reconstruct of Degradation path coordinate and the service life of lithium battery Forecasting Methodology of multiple linear regression |
| CN109596983A (en) * | 2018-11-19 | 2019-04-09 | 清华大学 | The prediction technique that capacity is dived during a kind of cell degradation |
| CN109946612A (en) * | 2019-04-01 | 2019-06-28 | 北京交通大学 | A kind of ternary capacity of lithium ion battery acceleration decline turning point recognition methods |
| CN112327191A (en) * | 2020-10-21 | 2021-02-05 | 北京航空航天大学 | Battery diving probability evaluation method based on geometric feature fusion decision |
| CN112327193A (en) * | 2020-10-21 | 2021-02-05 | 北京航空航天大学 | Lithium battery capacity diving early warning method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116541718A (en) * | 2023-07-07 | 2023-08-04 | 江苏正力新能电池技术有限公司 | Training method, device, equipment and medium for battery capacity jump recognition model |
| CN116541718B (en) * | 2023-07-07 | 2023-09-22 | 江苏正力新能电池技术有限公司 | Training method, device, equipment and medium for battery capacity jump recognition model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114559819B (en) | Electric automobile battery safety early warning method based on signal processing | |
| CN112327167B (en) | Battery capacity diving risk assessment method and system | |
| CN112327192B (en) | Battery capacity diving phenomenon identification method based on curve form | |
| CN111707958B (en) | Battery internal short circuit detection method based on capacity increment curve characteristics | |
| CN111123148B (en) | Method and equipment for judging short circuit in metal secondary battery | |
| CN115327425B (en) | Multi-stage anomaly detection method and system for lithium battery energy storage power station based on statistics | |
| CN112327193A (en) | Lithium battery capacity diving early warning method | |
| CN113777494A (en) | Lithium battery capacity diving turning point identification method based on geometric feature fusion decision | |
| WO2019174152A1 (en) | Current-based electrode short circuit prediction method and system | |
| CN111641236B (en) | Dynamic threshold power battery charging voltage state judgment method based on big data AI | |
| CN110927609B (en) | Decline evaluation method and device for battery energy storage system by utilizing battery in echelon | |
| CN114062932B (en) | Battery lithium precipitation detection method | |
| CN106325258B (en) | Relay protection device state evaluation method based on online monitoring information | |
| US20240027529A1 (en) | Method and system for detecting lithium precipitation of lithium batteries | |
| CN113777495A (en) | Lithium battery capacity diving online multi-stage early warning method and system based on characteristic area | |
| CN116577345A (en) | Method and system for detecting number of tabs of lithium battery | |
| CN115469226A (en) | Real-time safety early warning method for electric vehicle power battery based on operation big data | |
| CN115792634A (en) | Battery monomer voltage sampling fault identification method based on cloud online data | |
| CN115166564A (en) | Method for online quantitative evaluation of micro short circuit degree of lithium iron phosphate battery | |
| CN115856682B (en) | Lithium battery capacity jump point personalized early warning method based on critical point transition | |
| CN114814588B (en) | Method for rapidly estimating capacity of storage battery | |
| You et al. | Lithium-Ion Battery Online Capacity Diving Multilevel Evaluation and Early Warning Method Based on State of Nonlinear Aging | |
| CN111007418B (en) | Lithium battery residual life prediction method | |
| CN117310517A (en) | Lithium battery voltage state evaluation and early warning method based on big data analysis | |
| CN117879087A (en) | Dynamic control method of lithium battery energy storage 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 |