A method for analyzing cell consistency

Abstract

The application discloses a kind of battery consistency analysis methods, belong to battery system management technical field, the method is by obtaining the battery data in a certain time period in target energy storage equipment;The battery data is analyzed, whether the battery data exists exception is judged;If battery data is all normal or partially abnormal, according to the battery data, abnormal battery and normal battery are screened out, and the abnormal battery is excluded, the voltage data of the normal battery is used, the time required for target energy storage equipment to reach full charge and / or full discharge voltage is calculated;Based on the time required for target energy storage equipment to reach full charge and / or full discharge voltage and the power data, the loss capacity of normal battery is calculated, and battery consistency evaluation is carried out;Abnormal battery can be quickly positioned and whether maintenance and equipment maintenance priority order are confirmed, for the same energy storage system array in energy storage equipment, the influence of different energy storage equipment on energy storage system can be analyzed.

Description

Technical Field

[0001] This invention belongs to the field of battery system management technology and is used for cell consistency analysis, specifically relating to a cell consistency analysis method. Background Technology

[0002] Energy storage system arrays typically consist of one or more energy storage devices, which are composed of multiple battery cells connected in series or parallel. The charging and discharging performance of these cells determines the energy storage capacity of the device and affects the overall energy storage capacity of the array. Since the charging and discharging performance of battery cells is usually affected by their manufacturing process, materials, and operating environment, even cells from the same batch may have some performance variation. This will affect the overall energy storage capacity of the array. For example, in actual operation, if a cell in a particular device of an array takes a shorter time to fully charge and discharge than other cells, it may stop charging and discharging before the other cells are fully charged or discharged. This results in the energy storage device failing to reach the expected charge and discharge capacity, ultimately causing the overall energy storage capacity of the array to fall short of expectations. If the performance of the cells within the energy storage system varies significantly, indicating poor cell consistency, this will lead to energy loss and hinder subsequent maintenance of the system.

[0003] Currently, methods for analyzing cell consistency mainly fall into two categories: static sorting and dynamic sorting. Static sorting primarily involves testing cell performance under fixed operating conditions and in a static environment. Dynamic sorting mainly compares reference model data (experimental data under predetermined operating conditions or simulation data based on machine learning) with existing cell performance data. The main purpose of these methods is to screen out cells with poor consistency, but they do not assess the impact of poorly consistent cells on the energy storage system. An existing method for evaluating the consistency of individual cells in a battery system includes the following steps: S1, acquiring data on the voltage, temperature, current, and charging time of the cells during a constant current charging segment; S2, obtaining the voltage-capacity relationship based on the voltage, current, and charging time data from S1; S3, calculating the similarity between cells based on the voltage-capacity relationship from S2 and using this to form initial clusters; S4, classifying the cells using clustering based on the cell similarity data from S3 and the cell temperature data from S1; S5, filtering outliers (abnormal cells) based on the classification results from S4 and evaluating cell consistency based on these results. The following defects exist in practical use: 1) The consistency of battery cells during discharge is not considered; 2) Clustering algorithms have problems with partitioning and initial parameter selection; for example, K-means is uniform classification, but in reality, abnormal battery cell data may not be uniformly distributed, which may lead to unsatisfactory clustering results. In view of the above defects, this application proposes a battery cell consistency analysis method to solve the problems existing in the prior art. Summary of the Invention

[0004] To address the aforementioned problems, the present invention aims to propose a cell consistency analysis method to solve the problem of unsatisfactory analysis results in related technologies.

[0005] The technical solution adopted in this invention is as follows: a cell consistency analysis method, the method comprising the following steps:

[0006] Step 1: Obtain cell data from the target energy storage device within a certain time period, wherein the cell data includes voltage data, charging and / or discharging time period data, and power data;

[0007] Step 2: Analyze the cell data to determine if there are any abnormalities in the cell data;

[0008] If all the cell data is normal or some are abnormal, the abnormal cells and normal cells are filtered out based on the cell data, and the abnormal cells are excluded. The voltage data of the normal cells are used to obtain the polynomial fitting parameters.

[0009] Step 3: Based on the polynomial fitting parameters, use resampled data to perform polynomial curve fitting and calculate the time required for the target energy storage device to reach full charging voltage.

[0010] Step 4: Based on the time required for the target energy storage device to reach full charging voltage and the power data, calculate the lost capacity of the normal battery cell;

[0011] Step 5: Based on the aforementioned capacity loss, conduct a cell consistency evaluation.

[0012] In the charging and / or discharging time period data, the charging time period data refers to the time period from the start of charging to full charge of the target energy storage device, the discharging time period data refers to the time period from the start of discharging to complete discharging of the target energy storage device, and the voltage data refers to the voltage data of the target energy storage device during the time period in which the change in SOC is greater than 99% within the selected time period.

[0013] Analyzing the cell data to determine if there are any anomalies includes:

[0014] Select the sampling time interval;

[0015] The sampling time interval is divided into several segments, and the voltage data of each cell of the target energy storage device is sampled within each sampling time segment to obtain the sampling data for each sampling time segment.

[0016] Determine whether the changes in the sampled data in each segment within the sampling time interval conform to the continuous change of voltage data;

[0017] If the changes in the sampled data in each segment conform to the continuous changes in voltage data, then the cell data during the charging or discharging period is normal; otherwise, the cell data during the charging or discharging period is abnormal.

[0018] The voltage data of each cell of the target energy storage device is sampled multiple times within each sampling period to obtain the sampling data for each sampling period.

[0019] Calculate the average value of the sampled data for each sampling time period, and take the first derivative and absolute value of the cell data for each cell in the sampled data for each sampling time period to obtain the derivative value of the sampled data for each sampling time period.

[0020] Determine whether the change in the derivative value within the sampling time interval conforms to the continuous change of the voltage data;

[0021] Preferably, it is determined whether the change in the derivative value of the sampled data in the last two sampling time intervals within the sampling time interval conforms to the continuous change of voltage data.

[0022] The sampling time interval is the last 5% of the charging or discharging time period.

[0023] The method for screening abnormal and normal battery cells includes: analyzing the battery cell data using the local anomaly factor algorithm and the isolated forest algorithm, respectively, and obtaining a first analysis result. Second analysis results Regarding the first analysis results Second analysis results Perform cross-comparison, and based on the cross-comparison results The abnormal and normal battery cells are filtered out, and their serial numbers are obtained.

[0024] The abnormal battery cell is the first analysis result. Second analysis results Cells with inconsistent cross-comparison results are considered normal cells, while those with consistent results are considered normal cells.

[0025] Using the voltage data of the normal battery cell, polynomial fitting parameters are obtained, including:

[0026] Select voltage data within a specific charging or discharging time period;

[0027] The selected voltage data is sampled at a set sampling interval, and the average value of the sampled data is calculated.

[0028] Based on the average value, the absolute value of the first derivative and the absolute value of the percentage change of the sampled data are obtained, and the polynomial fitting parameters are determined according to the absolute value of the first derivative and the absolute value of the percentage change of the sampled data.

[0029] The method for determining the polynomial fitting parameters is as follows:

[0030] ;

[0031] in, Let be the absolute value of the first derivative of the sampled data. This represents the absolute percentage change in the sampled data.

[0032] Step 5 includes: based on the lost capacity, calculating a score value according to the expected capacity loss and the number of abnormal cells, and judging cell consistency based on the score value;

[0033] The formula for calculating the score is as follows:

[0034]

[0035] in, This is the score; For lost capacity; This represents the expected capacity loss.

[0036] if ,but

[0037] .

[0038] The method for calculating the loss capacity in step 5 is as follows:

[0039] ;

[0040] in The final non-zero power at the end of the selected charging or discharging time period;

[0041] For lost capacity;

[0042] The time required to reach full charging voltage.

[0043] It also includes step 6: based on the results of the cell consistency evaluation, the target energy storage device is given a maintenance priority rating.

[0044] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0045] The cell consistency analysis method and apparatus provided by this invention can quickly locate abnormal cells and determine whether maintenance is required and the priority order of equipment maintenance for energy storage devices. For energy storage devices in the same energy storage system array, it can analyze the impact of different energy storage devices on the energy storage system. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating the process of this embodiment;

[0047] Figure 2 This is a schematic diagram of the cell voltage of the A11 device in this embodiment;

[0048] Figure 3 This is a schematic diagram of the cell voltage at the end of the full charge time of the A11 device in this embodiment;

[0049] Figure 4 This is a schematic diagram of the cell voltage during the full discharge time of the A11 device in this embodiment;

[0050] Figure 5 This is a schematic diagram of the abnormal battery cell labeling for the A11 device in this embodiment;

[0051] Figure 6 This is a schematic diagram showing the abnormal cell markings at the end of the full charge time of the A11 device in this embodiment;

[0052] Figure 7 This is a schematic diagram simulating the cell voltage at the end of the full charge time of the A11 device in this embodiment;

[0053] Figure 8 This is a schematic diagram of the abnormal cell voltage during the full discharge time of the A11 device in this embodiment;

[0054] Figure 9 This is a schematic diagram simulating the cell voltage during the full discharge time of the A11 device in this embodiment;

[0055] Figure 10 This is a schematic diagram of the abnormal battery cell labeling for device A21 in this embodiment;

[0056] Figure 11 This is a schematic diagram of the abnormal battery cell labeling for device A21 in this embodiment;

[0057] Figure 12 This is a schematic diagram of the cell voltage at the end of the full charge time of the A21 device in this embodiment;

[0058] Figure 13 This is a schematic diagram showing the abnormal battery cell labeling at the end of the full charge time of the A21 device in this embodiment.

[0059] Figure 14 This is a schematic diagram simulating the cell voltage at the end of the full charge time of the A21 device in this embodiment;

[0060] Figure 15 This is a schematic diagram of the cell voltage during the full discharge time of the A21 device in this embodiment;

[0061] Figure 16This is a schematic diagram showing the abnormal battery cell markings during the full discharge time of the A21 device in this embodiment;

[0062] Figure 17 This is a schematic diagram simulating the cell voltage during the full discharge time of the A21 device in this embodiment;

[0063] Figure 18 This is a schematic diagram of the cell voltage of the A22 device in this embodiment;

[0064] Figure 19 This is a schematic diagram of the abnormal battery cell labeling for device A22 in this embodiment;

[0065] Figure 20 This is a schematic diagram of the cell voltage at the end of the full charge time of the A22 device in this embodiment;

[0066] Figure 21 This is a schematic diagram showing the abnormal cell markings at the end of the full charge time of the A22 device in this embodiment;

[0067] Figure 22 This is a simulation of the cell voltage at the end of the full charge time for the A22 device in this embodiment;

[0068] Figure 23 This refers to the cell voltage during the full discharge time of device A22 in this embodiment;

[0069] Figure 24 This is a labeling of abnormal battery cells within the full discharge time of device A22 in this embodiment;

[0070] Figure 25 This is a simulation of the cell voltage during the full discharge time of the A22 device in this embodiment. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0072] like Figure 1As shown, a cell consistency analysis method includes the following steps: Step 1, acquiring cell data from a target energy storage device within a certain time period, wherein the cell data includes voltage data, charging and / or discharging time period data, and power data; the charging and / or discharging time period data refers to the time period from the start of charging to full charge of the target energy storage device, the discharging time period data refers to the time period from the start of discharging to complete discharging of the target energy storage device, and the voltage data refers to the voltage of the target energy storage device within the selected time period. Step 2: Analyze the cell data to determine if there are any anomalies. If all cell data are normal or partially abnormal, filter out abnormal and normal cells based on the cell data, and exclude the abnormal cells. Use the voltage data of the normal cells to obtain polynomial fitting parameters. Analyzing the cell data to determine if there are any anomalies includes: selecting a sampling time interval; dividing the sampling time interval into several segments, sampling the voltage data of each cell of the target energy storage device within each sampling time segment, and obtaining the sampling data for each sampling time segment; determining whether the changes in the sampling data of each segment within the sampling time interval conform to the continuous change of voltage data; if the changes in the sampling data of each segment conform to the continuous change of voltage data, then the cell data within the charging or discharging time period is normal; otherwise, the charging or discharging time period is considered normal. Abnormal cell data during the discharge period; multiple samplings of voltage data for each cell in the target energy storage device are performed within each sampling period to obtain sampling data for each sampling period; the average value of the sampling data for each sampling period is calculated, and the first derivative of the cell data for each cell in the sampling data for each sampling period is calculated and the absolute value is taken to obtain the derivative value of the sampling data for each sampling period; it is determined whether the change of the derivative value within the sampling time interval conforms to the continuous change of voltage data; preferably, it is determined whether the change of the derivative value of the sampling data in the last two sampling periods within the sampling time interval conforms to the continuous change of voltage data; the sampling time interval is the last 5% of the charging or discharging period; the screening method for abnormal and normal cells includes: analyzing the cell data using the local anomaly factor algorithm and the isolated forest algorithm respectively, and obtaining a first analysis result. Second analysis results Regarding the first analysis results Second analysis results Perform cross-comparison, and based on the cross-comparison results Abnormal and normal battery cells are screened out, and their serial numbers are obtained; the abnormal battery cells are the first analysis result. Second analysis results Cells with inconsistent cross-comparison results are considered normal cells; using the voltage data of the normal cells, polynomial fitting parameters are obtained, including: selecting voltage data within a certain charging or discharging time period; sampling the selected voltage data at a set sampling interval and calculating the average value of the sampled data; based on the average value, obtaining the absolute value of the first derivative and the absolute value of the percentage change of the sampled data; and confirming the polynomial fitting parameters based on the absolute value of the first derivative and the absolute value of the percentage change of the sampled data; the method for determining the polynomial fitting parameters is as follows:

[0073] ;

[0074] in, Let be the absolute value of the first derivative of the sampled data. This represents the absolute percentage change in the sampled data.

[0075] Based on the lost capacity, a score is calculated according to the expected capacity loss and the number of abnormal cells, and the cell consistency is judged based on the score.

[0076] The formula for calculating the score is as follows:

[0077]

[0078] in, This is the score; For lost capacity; This represents the expected capacity loss.

[0079] if ,but

[0080] .

[0081] The method for calculating the loss capacity is as follows: ;

[0082] in The final non-zero power at the end of the selected charging or discharging time period;

[0083] For lost capacity; The time required to reach full charging voltage;

[0084] Step 6: Based on the results of the cell consistency evaluation, the target energy storage device is assigned a maintenance priority rating.

[0085] Polynomial curve fitting, for reference:

[0086] Montgomery DC, Peck EA, Vining G G. Introduction to linearregression analysis[M]. John Wiley & Sons, 2021: 223-235.

[0087] Isolation Forest Algorithm Reference:

[0088] Liu FT, Ting KM, Zhou Z H. Isolation forest[C] / / 2008 eighth ieeeinternational conference on data mining. IEEE, 2008: 413-422.

[0089] Local anomaly factor algorithm reference:

[0090] Breunig MM, Kriegel HP, Ng RT, et al. LOF: identifying density-based local outliers[C] / / Proceedings of the 2000 ACM SIGMOD international.

[0091] Based on the above steps: for energy storage devices, abnormal cells can be quickly located and it can be determined whether maintenance is required and the priority order of equipment maintenance can be determined; for energy storage devices in the same energy storage system array, the impact of different energy storage devices on the energy storage system can be analyzed.

[0092] Figure 1 As shown, the specific implementation is as follows: In order to achieve the purpose of the above technical solution, the cell consistency analysis method proposed in this embodiment has the following specific steps:

[0093] Step 1: Obtain cell voltage data, charging / discharging time, and power information of the energy storage devices in the target energy storage array over a certain period of time;

[0094] Step 2: Based on the battery cell voltage data and charge / discharge time information obtained in Step 1, filter out the time periods of full charge (discharge), that is, the time periods when the change in SOC is greater than 99%.

[0095] Step 3: Based on the cell voltage data from the full charge and / or full discharge periods in Step 2, for each full charge and / or full discharge period, select the cell voltage data from the last 5% of the time. First, resample this data into three segments and calculate the average value. Then, for each cell in each data segment N,... Battery cell data Take the first derivative and absolute value for each:

[0096] ;

[0097] Based on these three derivative values, and by comparing the last two derivative values... and Based on the characteristics of cell voltage data at the end of full charge and / or full discharge, i.e., the voltage data should show a continuous and significant increase (decrease), if... All the battery cells The derivatives of all are greater than If the cell voltage data is normal during the full charge and / or full discharge period, then it indicates that the cell voltage data during that period may be abnormal, such as due to sensor problems.

[0098] Step 4: Based on the cell voltage data anomaly check results in Step 3, if all full charge and / or full discharge time periods are considered to be abnormal, then the device is considered to be abnormal, and the cell consistency assessment results for all time periods will be set to blank values. In other cases, such as only some time periods are abnormal or no time periods are abnormal, the following steps will continue to be performed.

[0099] Step 5: Based on data from all full charge and / or full discharge periods, perform abnormal cell analysis using the Local Anomaly Factor algorithm and the Isolation Forest algorithm respectively, and obtain the abnormal cell analysis results. and Based on the analysis results of abnormal battery cells and Cross-comparison is performed, meaning that the results of two anomaly analyses for each cell must be consistent; otherwise, the cell is considered normal. The final result is the cross-analysis of the anomaly cells. The system obtains the serial numbers of normal and abnormal battery cells, then filters out the voltage data of normal battery cells during the full charge and / or full discharge periods, and performs the following steps for each full charge and / or full discharge period:

[0100] Step 1: Select the cell voltage data for the last 5% of the time period, and select the cell with the largest (smallest) voltage at the end of the time period according to the charging and discharging state.

[0101] Step 2: Based on the cell voltage data with the maximum (minimum) voltage from Step 1, set the sampling interval. Input time Then, based on the gap time, resampling is performed starting from the end time, and the average cell voltage at each time point after resampling is calculated.

[0102] Step 3: Based on the resampled data from Step 2, calculate the mean absolute value of the first derivative of this data segment. and the absolute value of its percentage change ,according to as well as Numerical confirmation of polynomial fitting parameters :

[0103] ;

[0104] Step 4: Fit parameters using the polynomial from Step 3 Polynomial curve fitting is performed using resampled data, and subsequent data simulations are performed based on the sampling interval in step 2 until the voltage at full charge and / or full discharge is reached. Then, the time required to reach the voltage at full charge and / or full discharge is calculated. ;

[0105] Step 5: Select the last non-zero power value based on the time period of the cell voltage data in Step 1. And step 4 Calculate the capacity of loss :

[0106] ;

[0107] Step 6: Based on step 2 Number of abnormal battery cells And step 5 According to the expected capacity and the number of battery cells Calculate the consistency score :

[0108]

[0109] Among them, if ,but

[0110] ;

[0111] Step 6: Based on the results of the cross-abnormal cell analysis in Step 2 The abnormal battery cells are located and their consistency is assessed based on the consistency score from step five. Perform maintenance priority rating:

[0112]

[0113] The rating criteria A, B, C, and D vary depending on the specific project.

[0114] like Figure 2-25As shown, the specific implementation is as follows: An energy storage system array 1 and array 2 are established. Array 1 contains one energy storage device with device ID A11; array 2 contains two energy storage devices with device IDs A21 and A22 respectively. Each device is of the same model, composed of 224 cells, with a fully charged cutoff voltage of 3600mV, a discharged cutoff voltage of 2800mV, and a rated capacity of 172kWh. Historical data acquisition time is from 2023-02-19 00:00:00 to 2023-02-20 00:00:00.

[0115] Calculated using the aforementioned method, taking device A11 as an example:

[0116] Step 1: As Figure 2 As shown, the cell voltage data of device A11 from 00:00:00 on 2023-02-19 to 00:00:00 on 2023-02-20 was obtained. The charging and discharging time information of device A11 was also obtained.

[0117] Step 2: Based on the cell voltage data and charge / discharge time information from Step 1, filter out the cell voltage data for periods of full charge and / or full discharge, i.e., periods when the SOC change is greater than 99%.

[0118]

[0119] Step 3: Based on the cell voltage data for the full charge and / or full discharge periods in Step 2, for each full charge and / or full discharge period, select the cell voltage data for the last 5% of the time, such as... Figure 3 and Figure 4 As shown; firstly, for the last 5% of the charging data, i.e., the cell voltage data from 06:41:16 to 07:02:16 on 2023-02-19, three segments of resampling are performed and their average values ​​are calculated. Then, the first derivative of the cell data for each cell n in each segment is calculated and its absolute value is taken. Based on these three derivative values, the values ​​of the last two segments of derivatives are compared. and , and thus All the battery cells The derivatives of all are greater than This indicates that the cell voltage data is normal during the charging period; similarly, the cell voltage data is normal during the discharging period.

[0120] Step 4: Based on the results of the cell voltage data anomaly check in Step 3, all cell voltage data for the charging and discharging time of Device A11 are normal.

[0121] Step 5: Based on data from all full charge and / or full discharge periods, perform abnormal cell analysis using the Local Anomaly Factor algorithm and the Isolation Forest algorithm respectively, and obtain the abnormal cell analysis results. and Based on the analysis results of abnormal battery cells and By performing cross-comparison, the final analysis results of the cross-abnormal cells were obtained. The system obtains the serial numbers of normal battery cells (v1, v2, v3..., v224) and abnormal battery cells (v62). Then, it filters out the voltage data of normal battery cells within the full charge and / or full discharge periods. For each full charge / discharge period, the following steps are performed, taking the full charge period from 2023-02-19 00:00:01 to 2023-02-19 07:02:16 as an example:

[0122] Step 1: Select the cell voltage data in the last 5% of the time period. Since it is in the charging state, select the voltage data of the cell with the largest voltage at the end of the time period (v7).

[0123] Step 2: Based on the data input in Step 1, set the sampling interval gap to 1 / 30 of the input time, i.e., 42.225 seconds; then, based on the gap time, resample from the end time and calculate the average cell voltage at each time point after resampling.

[0124] Step 3: Based on the resampled data from Step 2, calculate the mean absolute value of the first derivative of this data segment. 8, and the absolute value of its percentage change. 0.0019, according to as well as Numerical confirmation of polynomial fitting parameters ;

[0125] Step 4: Fit parameters using the polynomial from Step 3 Polynomial curve fitting is performed using resampled data, and subsequent data simulations are performed based on the sampling interval in step 2 until a full charge voltage of 3600mV is reached. Then, the time required to reach the full charge voltage is calculated. 0.005 hours;

[0126] Step 5: Select the last non-zero power value based on the time period of the cell voltage data in Step 1. And step 4 Calculate the capacity of loss :

[0127]

[0128] Step 6: Based on step 2 Number of abnormal battery cells and step 5 According to the expected capacity That is, 172 * 0.1 = 17.2 kWh and the number of battery cells. Calculate the consistency score :

[0129]

[0130] Step 6: Based on the results of the cross-abnormal cell analysis in Step 2 The abnormal battery cells are located and their consistency is assessed based on the consistency score from step five. The maintenance priority rating for this equipment is low.

[0131] like Figure 2-25 As shown, based on the above steps, calculations are performed on the relevant data for A11, A21, and A22, yielding the following results:

[0132]

[0133] in Figures 2 to 9 Visualize the relevant calculation process and results for A11. Figures 10 to 17 Visualize the correlation calculation process and results for A21. Figures 18 to 25 The calculation process and results related to A22 are visualized.

[0134] Based on the above results, devices A11, A21, and A22 all have abnormal battery cells. First, A11 has one abnormal battery cell (number v62). This abnormal cell has a minor impact on the device's charging and discharging capacity (scores 99.35 and 98.57). Considering potential sensor errors and other factors, this device can be considered to have a low maintenance priority. Second, A21 has three abnormal battery cells (numbers v10, v93, and v177). These abnormal cells have a minor impact on the device's charging capacity (score 97.51), but a significant impact on its discharging capacity (score 80.74). It is recommended to replace these abnormal cells to improve the device's discharging capacity. The maintenance priority is high. Finally, for device A22, there are 6 abnormal cells (numbered v3, v39, v123, v147, v200, v219). These abnormal cells have a small impact on the charging capacity of the device (score 99.46), but they have a certain impact on the discharging capacity (score 90.33). Therefore, the maintenance priority for this device is medium. In addition, for devices A21 and A22 in array 2, based on the above analysis results, we can know that the system power loss of this array mainly occurs in the discharging state, and the impact of device A21 is greater. Therefore, it is necessary to pay attention to whether device A21 is in an abnormal operating condition or whether there are other reasons that cause its device condition to be worse than that of device A22.

[0135] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing the consistency of battery cells, characterized in that, The method includes the following steps: Step 1: Obtain cell data from the target energy storage device within a certain time period, wherein the cell data includes voltage data, charging and / or discharging time period data, and power data; Step 2: Analyze the cell data to determine if there are any abnormalities in the cell data; If all the cell data is normal or some are abnormal, the abnormal cells and normal cells are filtered out based on the cell data, and the abnormal cells are excluded. The voltage data of the normal cells are used to obtain the polynomial fitting parameters. Step 3: Based on the polynomial fitting parameters, use resampled data to perform polynomial curve fitting and calculate the time required for the target energy storage device to reach full charge / or full discharge voltage. Step 4: Based on the time required for the target energy storage device to reach full charge / discharge voltage and the power data, calculate the lost capacity of the normal battery cell; Step 5: Based on the aforementioned capacity loss, conduct a cell consistency evaluation; Using the voltage data of the normal battery cell, polynomial fitting parameters are obtained, including: Select voltage data within a specific charging or discharging time period; The selected voltage data is sampled at a set sampling interval, and the average value of the sampled data is calculated. Based on the average value, the absolute value of the first derivative and the absolute value of the percentage change of the sampled data are obtained, and the polynomial fitting parameters are determined according to the absolute value of the first derivative and the absolute value of the percentage change of the sampled data. The method for determining the polynomial fitting parameters is as follows: ； in, For the number of terms, Let be the absolute value of the first derivative of the sampled data. This represents the absolute percentage change in the sampled data. Step 5 includes: based on the lost capacity, calculating a score value according to the expected capacity loss and the number of abnormal cells, and judging cell consistency based on the score value; The formula for calculating the score is as follows: ； in, For cross-comparison results Quantity, This represents the total number of battery cells. This is the score; For lost capacity; This represents the expected capacity loss. if ,but .

2. The method according to claim 1, characterized in that, In the charging and / or discharging time period data, the charging time period data refers to the time period from the start of charging to full charge of the target energy storage device, the discharging time period data refers to the time period from the start of discharging to complete discharging of the target energy storage device, and the voltage data refers to the voltage data of the target energy storage device during the time period in which the change in SOC is greater than 99% within the selected time period.

3. The method according to claim 1, characterized in that, Analyzing the cell data to determine if there are any anomalies includes: Select the sampling time interval; The sampling time interval is divided into several segments, and the voltage data of each cell of the target energy storage device is sampled within each sampling time segment to obtain the sampling data for each sampling time segment. Determine whether the changes in the sampled data in each segment within the sampling time interval conform to the continuous change of voltage data; If the changes in the sampled data in each segment conform to the continuous changes in voltage data, then the cell data during the charging or discharging period is normal; otherwise, the cell data during the charging or discharging period is abnormal.

4. The method according to claim 3, characterized in that, The voltage data of each cell of the target energy storage device is sampled multiple times within each sampling period to obtain the sampling data for each sampling period. Calculate the average value of the sampled data for each sampling time period, and take the first derivative and absolute value of the cell data for each cell in the sampled data for each sampling time period to obtain the derivative value of the sampled data for each sampling time period. Determine whether the change in the derivative value within the sampling time interval conforms to the continuous change of voltage data.

5. The method according to claim 4, characterized in that, Determine whether the changes in the derivative values ​​of the sampled data in the last two sampling time intervals within the sampling time interval conform to the continuous change of voltage data.

6. The method according to claim 4, characterized in that, The sampling time interval is the last 5% of the charging or discharging time period.

7. The method according to claim 1, characterized in that, The method for screening abnormal and normal battery cells includes: analyzing the battery cell data using the local anomaly factor algorithm and the isolated forest algorithm, respectively, and obtaining a first analysis result. Second analysis results Regarding the first analysis results Second analysis results Perform cross-comparison, and based on the cross-comparison results The abnormal and normal battery cells are filtered out, and their serial numbers are obtained.

8. The method according to claim 7, characterized in that, The abnormal battery cell is the first analysis result. Second analysis results Cells with inconsistent cross-comparison results are considered normal cells, while those with consistent results are considered normal cells.

9. The method according to claim 1, characterized in that, In step 5, the method for calculating the loss capacity is as follows: ； in The final non-zero power at the end of the selected charging or discharging time period; For lost capacity; The time required to reach full charging voltage.

10. The method according to any one of claims 1-7, characterized in that, It also includes step 6: based on the results of the cell consistency evaluation, the target energy storage device is given a maintenance priority rating.

Images