CN117805649A - Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation - Google Patents
Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation Download PDFInfo
- Publication number
- CN117805649A CN117805649A CN202311353668.XA CN202311353668A CN117805649A CN 117805649 A CN117805649 A CN 117805649A CN 202311353668 A CN202311353668 A CN 202311353668A CN 117805649 A CN117805649 A CN 117805649A
- Authority
- CN
- China
- Prior art keywords
- battery
- soh
- abnormal
- cluster
- soc
- 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
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 85
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 24
- 238000004458 analytical method Methods 0.000 claims abstract description 17
- 238000009826 distribution Methods 0.000 claims abstract description 7
- 230000005856 abnormality Effects 0.000 claims abstract description 5
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000007599 discharging Methods 0.000 claims description 8
- 238000007726 management method Methods 0.000 claims description 8
- 238000005259 measurement Methods 0.000 claims description 7
- 101000872083 Danio rerio Delta-like protein C Proteins 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 230000015556 catabolic process Effects 0.000 claims description 4
- 238000006731 degradation reaction Methods 0.000 claims description 4
- 238000013139 quantization Methods 0.000 claims description 4
- 238000005520 cutting process Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 2
- 238000012983 electrochemical energy storage Methods 0.000 abstract description 11
- 238000001514 detection method Methods 0.000 abstract description 4
- 238000003745 diagnosis Methods 0.000 abstract description 2
- 238000002474 experimental method Methods 0.000 description 11
- 238000004146 energy storage Methods 0.000 description 9
- 230000006872 improvement Effects 0.000 description 9
- 238000012706 support-vector machine Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 230000036541 health Effects 0.000 description 6
- 238000012423 maintenance Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000005562 fading Methods 0.000 description 4
- 230000010354 integration Effects 0.000 description 4
- 229910001416 lithium ion Inorganic materials 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000011002 quantification Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000007621 cluster analysis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 229910052744 lithium Inorganic materials 0.000 description 2
- 239000000178 monomer Substances 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 101001063878 Homo sapiens Leukemia-associated protein 1 Proteins 0.000 description 1
- 102100030893 Leukemia-associated protein 1 Human genes 0.000 description 1
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Landscapes
- Secondary Cells (AREA)
Abstract
The invention relates to the field of abnormal detection of battery cell levels in a battery cluster of an electrochemical energy storage power station, in particular to a method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation. The method comprises the following steps: step 1, acquiring capacity attenuation data of each single battery in a battery cluster; step 2, calculating SOH values of battery cells of the single batteries by using capacity attenuation data of the single batteries; step 3, outlier analysis is carried out by combining a DBSCAN density clustering algorithm according to the SOH value time sequence distribution rule to remove abnormal SOH values which cannot be clustered into stacks; the generation cause of the abnormal SOH value includes detecting an abnormality of the apparatus; and extracting SOH standard deviation and variation coefficient indexes in the battery cluster for a period of time, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values so as to identify abnormal battery cells. The invention provides powerful support for improving the accurate identification rate of the abnormal battery cells of the battery and accurately identifying and identifying the abnormal battery cells in the electrochemical energy storage system, and enhances the accuracy and reliability of diagnosis.
Description
Technical Field
The invention relates to the field of abnormal detection of battery cell levels in a battery cluster of an electrochemical energy storage power station, in particular to a method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation.
Background
Electrochemical energy storage systems have been widely used in the fields of electric automobiles, wind power generation, solar power generation, and the like. Among them, lithium ion batteries are one of the most commonly used battery technologies at present. However, due to the limited life of the battery, its capacity may gradually decay over time, which may lead to reduced performance or failure of the electrochemical energy storage system. In order to ensure the safety and reliability of the electrochemical energy storage system, it is important to identify abnormal cells in time. SOH (State of Health) is an important indicator of battery health and can be used to evaluate the capacity fade of a battery. Thus, a method of quantifying battery capacity degradation based on SOH may be used to identify abnormal cells.
The method for identifying the abnormal battery cells of the electrochemical energy storage system by SOH quantitative battery capacity attenuation is to quantitatively analyze the capacity attenuation of the battery to judge whether the battery cells are abnormal or not. Typically, a normal battery will gradually decrease in capacity during use, but the rate of decay is relatively steady. And abnormal cells may have an abnormally fast or slow rate of capacity fade.
Prior art related to the invention
The technical scheme in the first prior art is as follows: abnormal cell identification method based on capacity fade rate (Wang, z., huang, w., xiong, r., & Chen, z. (2018) & gt-of-Health Quantification Method for Lithium-Ion Batteries Based on Capacity Fading rate.ieee Transactions on Industrial Informatics,14 (10), 4519-4528.) the method judges abnormal cells by calculating the fade rate of the battery capacity. In general, a normal battery will gradually decrease in capacity during use, but the rate of decay is relatively steady. Thus, by measuring and calculating the change in battery capacity over time, the rate of capacity fade can be obtained. If the decay rate of a certain cell is faster or slower, there is a significant deviation from the normal decay rate, then the cell can be determined to be an abnormal cell.
Advantages and disadvantages of the first prior art: the method has the advantages of simplicity, intuitiveness and no need of complex models or algorithms to identify abnormal cells. However, this approach has a disadvantage in that the non-linear nature of the capacity fade is ignored. Thus, in some cases, erroneous judgment may be caused, and a normal cell is erroneously judged as an abnormal cell or vice versa.
Prior art II related to the invention
The technical scheme of the second prior art is as follows: abnormal cell identification method based on capacity fading curve fitting (Li, b., xiong, r., gao, j., & Liu, z. (2017) & An adaptive state-of-health quantification method for lithium-ion batteries based on capacity fade curve fitting.applied Energy,185, 1893-1902.) the method uses a capacity fading curve fitting method to describe characteristics of battery capacity fading, thereby identifying abnormal cells. First, data of the change of the battery capacity with time is collected and preprocessed. The capacity fade curve is then fitted to the existing model using a suitable fitting algorithm or model (e.g., polynomial fit, exponential fit, etc.). The health status of the cell can be assessed by fitting parameters of the curve or fitting residuals. Abnormal cells typically result in a significant difference in fit results from normal cells.
Advantages and disadvantages of the second prior art: the method has the advantages that the characteristic of battery capacity attenuation is described more accurately, and the identification accuracy of abnormal battery cells is improved. However, this approach has the disadvantage of requiring modeling of the capacity fade curve, which increases the complexity of implementation and requires development and adjustment of the model for different types of batteries.
Prior art III related to the invention
The technical scheme of the third prior art is as follows: abnormal cell identification method (Jin, C., zhang, C., chen, Z., & Li, J. (2019) & An abnormal cell identification method for lithium-ion batteries based on support vector machine. Journal of Energy Storage,23, 285-294.) based on support vector machine, the method utilizes support vector machine (Support Vector Machine, SVM) algorithm to identify abnormal cells. The support vector machine is a commonly used machine learning algorithm that can be used to classify and identify problems. For the identification of abnormal cells of a battery, a training data set is first constructed, which contains battery samples of known state of health and abnormal state. These samples are then used for training to build a support vector machine model. The model can determine whether an unknown battery sample is an abnormal cell by measuring the similarity between the unknown battery sample and the sample in the training dataset.
Advantages and disadvantages of the third prior art: the method has the advantages of higher accuracy and robustness, and is suitable for different types of batteries and conditions with larger change of working conditions. However, this approach has the disadvantage of requiring a large training data set to build the support vector machine model and may require adaptation for a particular application scenario.
Disclosure of Invention
The invention aims to solve the problems of accurate identification and recognition of abnormal cells in an electrochemical energy storage system. The invention can improve the accurate identification capability of the abnormal battery cell, and can more accurately identify the abnormal battery cell by adopting an SOH quantization method and combining the characteristic of battery capacity attenuation. This helps to discover and handle abnormal cells in time, avoiding negative impact on the performance of the overall energy storage system. Secondly, can realize automatic and quick unusual electric core discernment: the invention utilizes the data acquisition and analysis technology and combines advanced algorithm and model to realize the rapid detection and identification of the battery capacity attenuation characteristics. This reduces the need for manual intervention and judgment, improving the degree of automation and efficiency of recognition. By accurately identifying the abnormal battery cells, scientific basis can be provided for supporting decision making, reliable data and information are provided, and scientific basis is provided for decision makers to make reasonable maintenance strategies, optimize operation strategies and energy storage system planning. This helps to improve the reliability, economy and sustainability of the energy storage system.
In order to achieve the above purpose, the present invention is realized by the following technical scheme.
The invention provides a method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation, which comprises the following steps:
step 1, acquiring capacity attenuation data of each single battery in a battery cluster;
step 2, calculating SOH values of battery cells of the single batteries by using capacity attenuation data of the single batteries;
and 3, extracting SOH standard deviation and variation coefficient indexes in the battery cluster for a period of time, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values so as to identify abnormal battery cells.
As one of the improvements of the above technical solution, the step 1 includes:
and acquiring capacity attenuation data of each single battery, including the temperature, voltage, current and charge and discharge time of the single battery, through a battery management system or a battery cluster overall control unit.
As one of the improvements of the above technical solution, the step 2 includes:
step 2-1, cutting data according to the running state of current positioning charging and discharging: starting from a certain charge data and recording the period as continuous charge data before the next discharge data; starting from a certain discharge data and recording the period as continuous discharge data before the next charge data;
step 2-2, screening a time ta meeting the post-charge standing time length T0 and a time tb meeting the post-discharge standing time length T0; if ta and tb cannot be screened out simultaneously, returning to the step 1; if ta and tb are screened out simultaneously, enter step 2-3;
step 2-3, calculating full charge SOC (ta) and full discharge SOC (tb) by using ta and tb based on the OCV-SOC charge curve and the OCV-SOC discharge curve respectively, and judging whether the calculated full charge SOC (ta) and full discharge SOC (tb) respectively meet a full charge threshold and a full discharge threshold; if the full charge amount SOC (ta) does not meet the full charge threshold value or the full discharge amount SOC (tb) does not meet the full discharge threshold value, ending the method; if both the two conditions are satisfied, the step 2-4 is carried out;
step 2-4. Calculating the battery charge/discharge amount DeltaC between ta and tb cell SOC difference DeltaSOC cell And the initial battery capacity Cnom, and using the calculated DeltaC cell 、△SOC cell And Cnom calculates SOH value SOH cell ；
Step 2-5, repeating the steps 2-2 to 2-4, calculating an SOH value every time a pair of adjacent ta and tb are selected, and taking all SOH average values as SOH final values of the battery cells.
As one of the improvements of the above technical scheme, SOH in the steps 2-4 cell The calculation formula of (2) is as follows:
as one of the improvements of the technical proposal, deltaC cell By using ampere-hour integral calculation, the formula is as follows:
wherein t is 1 Calculating the initial time of the charge and discharge quantity of the battery; t is t n Calculating the termination time of the charge and discharge quantity of the battery; i (t) i ) At t i Time series loop current, t i-1 At t i-1 Time n is the total time.
As one of the improvements of the above technical scheme, ΔSOC cell The calculation formula of (2) is as follows:
ΔSOC cell ＝SOC end -SOC start
wherein SOC is end And SOC (System on chip) start The electric quantity of the battery at the end time and the start time of the charge-discharge process is respectively represented.
As an improvement of the foregoing technical solution, the method further includes, before step 3:
performing outlier analysis by combining a DBSCAN density clustering algorithm according to an SOH value time sequence distribution rule to remove abnormal SOH values which cannot be clustered into stacks; the cause of the abnormal SOH value includes detecting an abnormality of the apparatus.
As one of the improvements of the above technical solution, the performing outlier analysis by combining with a DBSCAN density clustering algorithm includes:
(1) input: sample set d= { x 1 ,x 2 ,...,x j ...,x m J=1, 2,..m, m is the total number of samples, where x 1 Representing the SOH average value of the No. 1 battery in the day, neighborhood parameters (epsilon, minPts), and adopting a Euclidean distance in a sample distance measurement mode;
(2) the output cluster is divided into C, and the process is as follows:
I. initializing a set of core objectsInitializing cluster number k=0, initializing unvisited sample set Γ=d, cluster division +.>
II for sample x of j j The clustering core object is found out according to the following steps:
a. finding out a sample x by a distance measurement mode j Is-neighborhood subsampled set N ε (x j )；
b. The number of the sub-sample set samples satisfies |N ε (x j ) I is not less than MinPts, sample x j Adding a core object sample set: Ω=Ω { x } U } j }；
III, e.g. core object setThe algorithm is ended, otherwise, the step IV is carried out;
randomly selecting one core object o from the core object set omega, and initializing the current core object queue omega cur = { o }, initializing a class sequence number k=k+1, initializing a current cluster sample set Ω k = { o }, update unvisited set Γ = Γ - { o };
v. if the current cluster core objectThen the current cluster C k After the generation is completed, updating cluster division C= { C 1 ,C 2 ,...,C k Update core object set Ω=Ω -C k Turning to the step III; otherwise update the core object set Ω=Ω -C k ；
VI. core object queue Ω in current cluster cur A core object o' is taken out, and all epsilon-neighborhood subset sample sets N are found out through neighborhood distances ε (o') let Δ=n ε (o')Γ, updating the current cluster sample set C k ＝C k Update non-accessed sample set Γ=Γ - Δ, update Ω cur ＝Ω cur U (delta U omega) -o', turning to step V;
VII, outputting the following results: cluster division c= { C 1 ,C 2 ,...,C k And defining SOH data points which are not identified as clusters as abnormal SOH value points.
The invention also provides a system for identifying abnormal battery cells based on SOH quantized battery capacity attenuation, which comprises:
the data acquisition and processing module is used for acquiring capacity attenuation data of each single battery in the battery cluster and calculating SOH values of the battery cells of each single battery by utilizing the capacity attenuation data of each single battery;
the abnormal battery cell identification module is used for extracting SOH standard deviation and variation coefficient indexes in a battery cluster for a period of time, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values, so that the abnormal battery cells can be identified.
As an improvement of the foregoing technical solution, the system further includes: the SOH value eliminating module is used for carrying out outlier analysis and eliminating abnormal SOH values which cannot be clustered into a pile through a SOH value time sequence distribution rule and a DBSCAN density clustering algorithm; the cause of the abnormal SOH value includes detecting an abnormality of the apparatus.
Compared with the prior art, the invention has the advantages that:
according to the invention, SOH quantization and battery capacity attenuation characteristics are comprehensively considered, and are combined, so that the capacity attenuation is fully utilized as an important index for evaluating the health state of the battery, and the abnormal battery cells are more comprehensively and accurately identified. Meanwhile, various technical means are introduced, and the method not only adopts the traditional capacity attenuation rate analysis, but also adopts various technical means such as ampere-hour integration and DBSCAN algorithm, and the like, thereby improving the identification accuracy and reliability of abnormal battery cells. In summary, the method for identifying the abnormal battery cells of the electrochemical energy storage system based on SOH quantitative battery capacity attenuation aims to improve the accurate identification capability of the abnormal battery cells of the battery, realize an automatic and rapid identification process, and provide scientific basis for decision making so as to ensure the safety and performance of the energy storage system.
The invention provides powerful support for improving the accurate identification rate of the abnormal battery cells of the battery and accurately identifying and identifying the abnormal battery cells in the electrochemical energy storage system, and enhances the accuracy and reliability of diagnosis.
Drawings
Fig. 1 is a method of calculating SOH based on BMS data.
Detailed Description
The technical scheme of the invention is described in detail below with reference to the accompanying drawings and examples.
Example 1
The method for identifying abnormal cells based on SOH quantized battery capacity fade of the present invention generally comprises the steps of:
(1) and (3) data acquisition: capacity fade data of the battery, i.e., cell voltage, current, charge-discharge time, etc., is acquired through a battery management system (Battery Management System, abbreviated as BMS).
(2) Data preprocessing: and filtering, denoising and the like are carried out on the acquired capacity attenuation data so as to eliminate the influence of interference factors on analysis results.
(3) SOH quantification: based on ampere-hour integration and an SOC and OCV interpolation method, the collected data such as the battery voltage value, the current value and the like are calculated and converted into an SOH value through a formula, and the SOH value is used for representing the health state of the battery.
(4) Abnormal cell identification: and establishing a model for identifying abnormal cells through DBSCAN density cluster analysis, dividing SOH values of more than half a year of each single cell into clusters connected with different densities, and marking the rest unviewed objects as noise or belonging to other clusters so as to judge whether the cells are abnormal.
(5) And (3) outputting results: and outputting the identification result to the BMS or other control systems so as to take corresponding measures, such as eliminating abnormal battery cells, maintaining and the like.
The embodiment 1 of the invention provides a complete technical scheme:
1 data preprocessing
The invention needs the collected data: and acquiring real-time running data of a battery cluster total control unit (BCMS) of the whole station battery, wherein the real-time running data comprise parameters such as current, voltage, temperature and the like, and test experimental data corresponding to charge and discharge SOC-OCV used for estimating the SOC by a linear interpolation method. Preprocessing the acquired data, including operations such as field screening, missing value, noise removal, data alignment and interpolation, and the like, so as to ensure the accuracy and the continuity of the data. The charge/discharge state is determined uniformly by determining that the total current is negative, the charge is determined, the discharge is determined by a positive value, and the rest state is determined by 0.
SOC value (State of Charge): this is the state of charge of the battery, typically expressed in percent, e.g., 0% for full discharge and 100% for full charge. In experiments, it is often expressed in a series of discrete SOC values, e.g. 0%, 10%, 20%, 100%.
OCV value (Open Circuit Voltage): OCV is the voltage of a battery without charging or discharging, typically expressed in volts (V). In the experiment, it is necessary to measure or record the OCV value corresponding to each SOC value. These OCV values will constitute data points of the SOC-OCV curve.
Charge and discharge experimental data: and obtaining OCV values associated with different SOC values through a series of charging and discharging experiments, and generating an SOC-OCV curve. These experimental data should include OCV values of the battery at different SOCs while recording environmental parameters such as time point, current, temperature, etc. at the time of the experiment to ensure reproducibility and comparability of the experiment.
Environmental conditions: recording the environmental conditions at the time of the experiment is very important, including factors such as temperature, humidity, atmospheric pressure, etc. These conditions can affect the performance of the battery and therefore need to be recorded in experimental data.
Measuring equipment information: details of instruments and equipment for measuring SOC and OCV, including model number, accuracy specifications, etc., are recorded. This helps to ensure accuracy and repeatability of the experiment.
Experimental methods and procedures: detailed methods and procedures of experiments are described, including experimental conditions of rate of charge and discharge, cut-off voltage, sampling frequency, etc. This helps to replicate the experiment and verify the data accuracy.
The linear interpolation method may estimate the OCV value corresponding to the SOC between two known points based on the known SOC-OCV data points, thereby achieving SOC estimation. Linear interpolation assumes that the relationship between SOC and OCV is linear, so that a sufficient number of experimental data points are required to ensure accuracy of the estimation.
2SOH online estimation
And calculating the SOH value of the battery according to the acquired data. SOH is an indicator describing the state of health of a battery, and is calculated by estimating based on capacity fade, i.e., comparing the difference between the current capacity and the rated capacity of the battery. The battery capacity is the amount of charge that the battery can store, and gradually decreases with time and through charge and discharge cycles. Accurate estimation of SOH is critical to battery management and maintenance. It may help determine when the battery needs to be serviced, replaced, or optimally managed to ensure reliability and performance of the battery system. In addition, accurate assessment of SOH is also of great importance for secondary use and recycling of batteries.
FIG. 1 is a flow chart for on-line SOH estimation:
3 monomer SOC calculation
Firstly, judging continuous operation data, and cutting the data according to the operation state of current positioning charging and discharging: starting from a certain charge data and recording the period as continuous charge data before the next discharge data; the period from the start of a certain discharge data to the next charge data is referred to as continuous discharge data. The time when the charge data is continuously kept and before the charge data is kept and still for 3 hours is ta meeting the calculated SOC condition, and the time when the discharge data is continuously kept and before the charge data is kept and still for 3 hours is tb meeting the calculated SOC condition. The cell voltage at the penultimate point of the fully set state is referred to as the open circuit voltage (Open Circuit Voltage, OCV). The OCV of each cell is designated OCVcell.
Since the linear relation of the SOC-OCV of the lithium battery is not good, the battery OCV jumps greatly when the SOC is between 100% and 95%, the change of the OCV is very small and tends to be smooth when the SOC is between 95% and 65%, and the battery OCV starts to change obviously when the SOC is below 65%, particularly 5%. Therefore, considering that the data is recorded at intervals of 5%, a piecewise linear method is adopted, a linear relation is made between two adjacent points, two parameters of K, B are calculated, a binary table is searched for on which line the OCV voltage of the current static battery is, and then a corresponding function is substituted, so that the initial value of the current SOC is calculated. And (3) performing interpolation calculation on the OCVcell to obtain the SOC of each single cell, and marking the SOC as SOCcell (a boundary value is adopted beyond the boundary during interpolation).
Then selecting and using a charging or discharging SOC-OCV interpolation table according to the running state before the standing state; a stationary state after a certain time ta is a charged state exists: SOC calculation result > 95%; a stationary state after a certain time tb is a discharge state exists: SOC calculation results need < =25%.
Every time a pair of adjacent ta and tb are detected (the adjacent points are selected to avoid the excessive accumulated error of more than 10 circles of current measurement), and the average change rate of the SOC calculated based on the SOC-OCV curve is smaller than 0.2%/min in 5 minutes, an SOH value is calculated according to the following general expression, and finally all SOH average values are taken as the final SOH value of the battery cell.
4 Battery cell capacity throughput calculation
And regarding the judged continuous operation data, if the net accumulated operation time meets the operation time threshold value and the net accumulated standing time meets the standing time threshold value, the judged continuous operation data is regarded as effective operation data, and the capacity throughput calculation of the battery cell can be carried out. Calculating the capacity throughput of the battery unit of continuous operation data through ampere-hour integration, and recording the capacity throughput as delta C cell The unit Ah, formula is as follows: i is current, unit A, and direction is distinguished; t is time, in h.
Wherein t is 1 Calculating the initial time of the charge and discharge quantity of the battery; t is t n Calculating the termination time of the charge and discharge quantity of the battery;
I(t i ) At t i A time-of-day series loop current; deltaC cell And charging and discharging the battery.
5 monomer SOH calculation
And for the calculated system capacity throughput, if the calculated system capacity throughput is the last time continuous operation data and the continuous operation data of the system capacity throughput, the calculated SOCcell meets the SOC threshold value, and the SOH of each single cell is recorded as SOHcell, and the calculation formula is as follows:
wherein DeltaC cell The capacity throughput of a battery cell is recorded as the previous SOC of a certain battery cell cell Is SOC (State of charge) start Last SOC cell Is SOC (State of charge) end ，C nom The battery capacity at normal temperature is generally 280Ah.
6 abnormal cell outlier identification
1) SOH values of the batteries are calculated based on the collected data (wherein 1 represents that the batteries are completely healthy and 0 represents that the batteries are invalid), and the method and the device are used for carrying out cluster analysis to remove outliers by analyzing SOH value time sequence distribution rules of the battery cells in each battery cluster of the container and collecting DBSCAN density clustering algorithm.
(1) Input: sample set d= { x 1 ,x 2 ,...,x m X, where x 1 Representing the SOH average value of the No. 1 battery in the day, neighborhood parameters (epsilon, minPts), and adopting a Euclidean distance in a sample distance measurement mode;
(2) and (3) outputting: cluster partition C.
I. Initializing a set of core objectsInitializing cluster number k=0, initializing unvisited sample set Γ=d, cluster division +.>
Ii. for j=1, 2,..m, the clustering core object is found out according to the following steps:
a. finding out a neighborhood sub-sample set N of a sample xj by a distance measurement mode ε (x j )；
b. The number of the sub-sample set samples satisfies |N ε (x j ) I is not less than MinPts, sample x j Joining core object samplesAggregation:
Ω＝Ω∪{x j }；
III, e.g. core object setThe algorithm is ended, otherwise, the step IV is carried out;
randomly selecting one core object o from the core object set omega, and initializing the current core object queue omega cur = { o }, initializing a class sequence number k=k+1, initializing a current cluster sample set Ω k = { o }, update unvisited set Γ = Γ - { o };
v. if the current cluster core objectThen the current cluster C k After the generation is completed, updating cluster division C= { C 1 ,C 2 ,...,C k Update core object set Ω=Ω -C k And (3) turning to the step III. Otherwise update the core object set Ω=Ω -C k ；
VI. core object queue Ω in current cluster cur A core object o' is taken out, and all epsilon-neighborhood subset sample sets N are found out through neighborhood distances ε (o') let Δ=n ε (o')Γ, updating the current cluster sample set C k ＝C k Update non-accessed sample set Γ=Γ - Δ, update Ω cur ＝Ω cur U (delta. U.OMEGA) -o', and step V.
VII, outputting the following results: cluster division c= { C 1 ,C 2 ,...,C k And defining data points which are not identified as clusters as outliers. The data points refer to SOH data of the battery cells, and SOH value of one battery cell is one data point.
2) And (3) identifying the SOH discrete index outliers of the battery in the cluster: the method comprises the steps of respectively extracting SOH standard deviation and variation coefficient indexes in each cluster of data over half a year, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values, wherein the outlier SOH values are identified by selecting eps=0.01 and min_sample=5, so that abnormal battery cells are positioned.
In the embodiment, firstly, abnormal data points are removed through 1) a density clustering algorithm, and abnormal data/bad data points which cannot be clustered into piles are removed, wherein the abnormal data/bad data points possibly are electrical core SOH data points with extremely-outliers, which are caused by detection equipment problems (voltage sampling errors caused by BMS voltage acquisition point contact problems for acquiring original data or SOH values calculated by current integration in a certain period of time become outliers), and the like; and then clustering analysis is carried out by 2) setting the distance and the clustering number of the minimum clustering points to find out the truly outlier battery cells.
The invention has the technical effects that:
realize timely early warning and fault handling: after the abnormal battery cells are accurately identified, corresponding measures such as maintenance, replacement or isolation can be timely taken to prevent the abnormal battery cells from negatively affecting the performance of the whole energy storage system. This helps to improve the safety, reliability and lifetime of the energy storage system.
Optimizing maintenance and operation management policies: through the identification and evaluation of the abnormal battery cells, scientific basis can be provided for maintenance strategies and operation management. For example, a patrol plan can be formulated and a charging and discharging strategy can be optimized according to the distribution and the characteristics of the abnormal battery cells, so that the efficiency and the economy of the energy storage system are improved.
Promoting the development and application and popularization of battery technology: the high-efficiency identification and evaluation capability of the method is beneficial to the improvement and optimization of battery technology, promotes the technical development and application popularization in the field of batteries, and further improves the performance and sustainability of the battery energy storage system.
In summary, the method for identifying the abnormal battery cells of the electrochemical energy storage system based on SOH quantitative battery capacity attenuation improves the accurate identification rate of the abnormal battery cells on the basis of comprehensively considering various indexes and methods, realizes timely early warning and fault handling, optimizes maintenance and operation management strategies, promotes the development and application popularization of battery technology, and has obvious innovation and beneficial application effects.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and are not limiting. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the present invention, which is intended to be covered by the appended claims.
Claims (10)
1. A method of identifying abnormal cells based on SOH quantified battery capacity fade, the method comprising:
step 1, acquiring capacity attenuation data of each single battery in a battery cluster;
step 2, calculating SOH values of battery cells of the single batteries by using capacity attenuation data of the single batteries;
and 3, extracting SOH standard deviation and variation coefficient indexes in the battery cluster for a period of time, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values so as to identify abnormal battery cells.
2. The method for identifying abnormal cells based on SOH quantization of battery capacity degradation according to claim 1, wherein the step 1 comprises:
and acquiring capacity attenuation data of each single battery, including the temperature, voltage, current and charge and discharge time of the single battery, through a battery management system or a battery cluster overall control unit.
3. The method for identifying abnormal cells based on SOH quantization of battery capacity degradation according to claim 2, wherein the step 2 comprises:
step 2-1, cutting data according to the running state of current positioning charging and discharging: starting from a certain charge data and recording the period as continuous charge data before the next discharge data; starting from a certain discharge data and recording the period as continuous discharge data before the next charge data;
step 2-2, screening a time ta meeting the post-charge standing time length T0 and a time tb meeting the post-discharge standing time length T0; if ta and tb cannot be screened out simultaneously, returning to the step 1; if ta and tb are screened out simultaneously, enter step 2-3;
step 2-3, calculating full charge SOC (ta) and full discharge SOC (tb) by using ta and tb based on the OCV-SOC charge curve and the OCV-SOC discharge curve respectively, and judging whether the calculated full charge SOC (ta) and full discharge SOC (tb) respectively meet a full charge threshold and a full discharge threshold; if the full charge amount SOC (ta) does not meet the full charge threshold value or the full discharge amount SOC (tb) does not meet the full discharge threshold value, ending the method; if both the two conditions are satisfied, the step 2-4 is carried out;
step 2-4. Calculating the battery charge/discharge amount DeltaC between ta and tb cell SOC difference DeltaSOC cell And the initial battery capacity Cnom, and using the calculated DeltaC cell 、△SOC cell And Cnom calculates SOH value SOH cell ；
Step 2-5, repeating the steps 2-2 to 2-4, calculating an SOH value every time a pair of adjacent ta and tb are selected, and taking all SOH average values as SOH final values of the battery cells.
4. The method for identifying abnormal cells based on SOH-quantified battery capacity fade as recited in claim 3, wherein SOH is determined in steps 2-4 cell The calculation formula of (2) is as follows:
5. the method for identifying abnormal cells based on SOH quantified battery capacity fade of claim 3 or 4, wherein Δc cell By using ampere-hour integral calculation, the formula is as follows:
wherein t is 1 Calculating the initial time of the charge and discharge quantity of the battery; t is t n Calculating the termination time of the charge and discharge quantity of the battery; i (t) i ) At t i Time series loop current, t i-1 At t i-1 Time of dayN is the total number of times.
6. The method for identifying abnormal cells based on SOH quantified battery capacity fade of claim 3 or 4, wherein Δsoc cell The calculation formula of (2) is as follows:
ΔSOCcell＝SOCend-SOCstart
wherein SOC is end And SOC (System on chip) start The electric quantity of the battery at the end time and the start time of the charge-discharge process is respectively represented.
7. The method for identifying abnormal cells based on SOH quantified battery capacity fade of any one of claims 1-6, further comprising, prior to step 3:
performing outlier analysis by combining a DBSCAN density clustering algorithm according to an SOH value time sequence distribution rule to remove abnormal SOH values which cannot be clustered into stacks; the cause of the abnormal SOH value includes detecting an abnormality of the apparatus.
8. The method for identifying abnormal cells based on SOH quantified battery capacity fade of claim 1 or 6, wherein the performing outlier analysis in combination with DBSCAN density clustering algorithm comprises:
(1) input: sample set d= { x 1 ,x 2 ,...,x j ...,x m J=1, 2,..m, m is the total number of samples, where x 1 Representing the SOH average value of the No. 1 battery in the day, neighborhood parameters (epsilon, minPts), and adopting a Euclidean distance in a sample distance measurement mode;
(2) the output cluster is divided into C, and the process is as follows:
I. initializing a set of core objectsInitializing cluster number k=0, initializing unvisited sample set Γ=d, cluster division +.>
II. for the j-th sampleThe X is j The clustering core object is found out according to the following steps:
a. finding out a sample x by a distance measurement mode j Is-neighborhood subsampled set N ε (x j )；
b. The number of the sub-sample set samples satisfies |N ε (x j ) I is not less than MinPts, sample x j Adding a core object sample set: Ω=Ω { x } U } j }；
III, e.g. core object setThe algorithm is ended, otherwise, the step IV is carried out;
randomly selecting one core object o from the core object set omega, and initializing the current core object queue omega cur = { o }, initializing a class sequence number k=k+1, initializing a current cluster sample set Ω k = { o }, update unvisited set Γ = Γ - { o };
v. if the current cluster core objectThen the current cluster C k After the generation is completed, updating cluster division C= { C 1 ,C 2 ,...,C k Update core object set Ω=Ω -C k Turning to the step III; otherwise update the core object set Ω=Ω -C k ；
VI. core object queue Ω in current cluster cur A core object o' is taken out, and all epsilon-neighborhood subset sample sets N are found out through neighborhood distances ε (o') let Δ=n ε (o')Γ, updating the current cluster sample set C k ＝C k Update non-accessed sample set Γ=Γ - Δ, update Ω cur ＝Ω cur U (delta U omega) -o', turning to step V;
VII, outputting the following results: cluster division c= { C 1 ,C 2 ,...,C k And defining SOH data points which are not identified as clusters as abnormal SOH value points.
9. A system for identifying abnormal cells based on SOH quantified battery capacity degradation, the system comprising:
the data acquisition and processing module is used for acquiring capacity attenuation data of each single battery in the battery cluster and calculating SOH values of the battery cells of each single battery by utilizing the capacity attenuation data of each single battery;
the abnormal battery cell identification module is used for extracting SOH standard deviation and variation coefficient indexes in a battery cluster for a period of time, performing outlier analysis based on a DBSCAN clustering algorithm, and identifying outlier SOH values, so that the abnormal battery cells can be identified.
10. The SOH-quantified battery capacity fade-based system of claim 9, wherein the system further comprises: the SOH value eliminating module is used for carrying out outlier analysis and eliminating abnormal SOH values which cannot be clustered into a pile through a SOH value time sequence distribution rule and a DBSCAN density clustering algorithm; the cause of the abnormal SOH value includes detecting an abnormality of the apparatus.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311353668.XA CN117805649A (en) | 2023-10-18 | 2023-10-18 | Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311353668.XA CN117805649A (en) | 2023-10-18 | 2023-10-18 | Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117805649A true CN117805649A (en) | 2024-04-02 |
Family
ID=90434033
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311353668.XA Pending CN117805649A (en) | 2023-10-18 | 2023-10-18 | Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117805649A (en) |
-
2023
- 2023-10-18 CN CN202311353668.XA patent/CN117805649A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110224192B (en) | Method for predicting service life of power battery by gradient utilization | |
| Liu et al. | State-of-health estimation of lithium-ion batteries based on electrochemical impedance spectroscopy: a review | |
| CN109655754B (en) | Battery performance evaluation method based on multi-dimensional grading of charging process | |
| CN113391213B (en) | Battery micro-fault diagnosis method based on battery pack consistency evolution | |
| CN113075554A (en) | Lithium ion battery pack inconsistency identification method based on operation data | |
| CN116609676B (en) | Method and system for monitoring state of hybrid energy storage battery based on big data processing | |
| CN113821976A (en) | Lithium battery fault diagnosis modeling method based on integrated algorithm | |
| CN113030761A (en) | Method and system for evaluating health state of battery of super-large-scale energy storage power station | |
| CN114839536A (en) | Lithium ion battery health state estimation method based on multiple health factors | |
| CN115219905A (en) | On-line detection method and device for short circuit in battery and storage medium | |
| CN110927609B (en) | Decline evaluation method and device for battery energy storage system by utilizing battery in echelon | |
| CN115480180A (en) | New energy battery health diagnosis and analysis method | |
| CN115061049A (en) | Method and system for rapidly detecting UPS battery fault of data center | |
| Zhou et al. | Online state of health estimation for seriesconnected LiFePO 4 battery pack based on differential voltage and inconsistency analysis | |
| CN110045291B (en) | Lithium battery capacity estimation method | |
| CN115327417A (en) | Early warning method and system for abnormity of power battery monomer and electronic equipment | |
| Li et al. | SOC estimation and fault diagnosis framework of battery based on multi-model fusion modeling | |
| CN114646888A (en) | Assessment method and system for capacity attenuation of power battery | |
| US11959969B2 (en) | Apparatus and method for diagnosing battery | |
| CN117054892B (en) | Evaluation method, device and management method for battery state of energy storage power station | |
| CN113687251A (en) | Dual-model-based lithium ion battery pack voltage abnormity fault diagnosis method | |
| CN114859249B (en) | Method and device for detecting battery pack capacity | |
| CN116842464A (en) | Battery system SOC estimation method | |
| CN117805649A (en) | Method for identifying abnormal battery cells based on SOH quantized battery capacity attenuation | |
| CN116256642A (en) | Lithium iron phosphate battery fault diagnosis method suitable for sparse data |
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 |