A method of battery cluster performing inconsistency anomaly cell detection and localization

Abstract

The application discloses a kind of battery cluster carries out the method for inconsistent abnormal monomer detection and positioning, belong to lithium ion battery detection field, first, based on the charge-discharge strategy of battery cluster and upper and lower cut-off voltage division charge-discharge cycle, second, according to the charge-discharge upper and lower cut-off voltage of battery cluster selects suitable feature extraction voltage to influence the inconsistency monomer of battery pack available capacity, then, further utilize residual chargeable electric quantity and residual dischargeable electric quantity to further diagnose and position inconsistent monomer, by aiming at battery pack available capacity characteristics, based on battery monomer voltage, can carry out real-time detection and positioning, and can preferentially locate the monomer of the relatively large degree of comprehensive inconsistency, without additional parameter detection, convenient and fast, inconsistent monomer battery is handled in time, guarantee the life, performance and safety of battery cluster when using.

Description

Technical Field

[0001] The present invention belongs to the field of lithium-ion battery detection, and in particular relates to a method for detecting and locating inconsistency abnormal cells in a battery cluster. Background Art

[0002] Lithium-ion batteries are rechargeable batteries with advantages such as high energy density, long cycle life, and no memory effect. They are widely used in civilian applications such as electric vehicles and energy storage. However, due to the limitations of current single-cell battery manufacturing technology, the construction of power battery packs usually requires connecting multiple single cells in series or parallel to form battery clusters to meet the actual performance requirements of electric vehicles.

[0003] However, there are often differences between individual cells in a power battery pack. This phenomenon is known as battery consistency. The more individual cells in a battery cluster, the more pronounced this problem becomes. This can lead to a series of negative impacts, including shortened battery life, reduced battery pack performance, and decreased safety.

[0004] Traditional technologies mostly use single parameter evaluation, multi-parameter evaluation and dynamic characteristic evaluation methods to evaluate battery inconsistency parameters;

[0005] The single parameter evaluation method only selects one parameter to evaluate the consistency, which cannot well reflect the actual condition of the battery;

[0006] Although the multi-parameter evaluation method evaluates battery consistency by combining two or more parameters such as voltage, capacity, internal resistance and discharge rate through clustering or information fusion, it has problems such as poor real-time performance, sensitivity to environmental factors, limited applicability and high cost.

[0007] Dynamic characteristic evaluation methods include battery internal dynamic process evaluation and charge and discharge characteristic curve evaluation. Among them, the internal dynamic process evaluation requires special equipment and a long test cycle, which is not conducive to actual production. Although the consistency evaluation based on the charge and discharge characteristic curve can more comprehensively reflect the comprehensive performance of the battery, the charge and discharge voltage is introduced into the consistency evaluation model to make the evaluation parameters a dynamic variable, but this method requires a large number of batteries and experimental data support, is costly, complex to operate, and is not convenient for real-time detection. It is especially inconvenient to use the above method when testing a large number of batteries, such as a battery cluster consisting of more than 400 batteries. Summary of the Invention

[0008] (1) Technical issues to be resolved

[0009] In order to overcome the shortcomings of the existing technology, a method for detecting and locating inconsistent abnormal cells in a battery cluster is proposed. The method solves the problems that the existing technology is not convenient for real-time detection and positioning of inconsistent cells in a battery cluster composed of a large number of batteries when used, as well as the single detection parameters, high cost and poor applicability. As a result, the battery cluster composed of a large number of batteries is prone to problems such as shortened life, decreased battery pack performance and reduced safety due to the inability to promptly handle inconsistency problems.

[0010] (2) Technical solution

[0011] The present invention is achieved through the following technical solutions: The present invention proposes a method for detecting and locating inconsistency-abnormal cells in a battery cluster. This method can comprehensively and accurately evaluate inconsistency-abnormal cells in a lithium-ion battery cluster, facilitate timely location and treatment of abnormal cells, and prevent battery cluster performance degradation and malfunction. The method is particularly suitable for battery clusters consisting of a large number of cells.

[0012] The following steps are involved:

[0013] Step 1: During the charging and operation of the energy storage system, the battery management system collects data from the batteries in the battery cluster in real time and performs data preprocessing;

[0014] First, various data of the batteries in the battery cluster are collected, including time, total voltage, current, single cell maximum voltage, single cell minimum voltage, single cell maximum temperature, single cell minimum temperature, battery cell voltage and other data, and the above data are recorded and stored;

[0015] Finally, the data is preprocessed by removing duplicate values, sorting disordered values, eliminating outliers, and filling missing values, so that the data can be cleaned before fault diagnosis to improve data quality and increase the accuracy and reliability of subsequent algorithms. The screened valid battery cell voltage data is stored and recorded. The battery voltage data is expressed as:

[0016]

[0017] Where m is the number of battery cells in the battery pack, and n is the number of sampling points;

[0018] Step 2: In order to comprehensively consider the inconsistency of abnormal cells in the battery cluster that restrict the available capacity of the battery cluster, the first cell to reach the upper cut-off voltage V upper Battery A1 with discharge cut-off voltage V lower Perform preliminary inconsistency diagnosis on battery A2;

[0019] First, according to the charging cut-off voltage V specified by the battery cluster upper and the discharge cut-off voltage V lower , set a fixed charging high-end voltage Vcharge and the discharge low-end voltage V discharge ;

[0020] Secondly, according to the charging high-end voltage V of the battery cluster charge and the discharge low-end voltage V discharge , find the two cells A1 and A2 in the battery cluster that reach the upper cut-off voltage of charge and the lower cut-off voltage of discharge the fastest, and calculate the voltage difference V between these two cells and the median of all cells at the corresponding moment diff ;

[0021] Furthermore, the voltage difference V of battery A1 diff The calculation formula is:

[0022] V diff1 =V charge -V meidian1

[0023] Among them, V meidian1 It is the median voltage of all battery cells at the sampling moment when charging to the high-end voltage, V diff1 is the voltage difference between battery A1 and the median voltage;

[0024] The voltage difference of battery A2 is V diff The calculation formula is:

[0025] V diff2 =V discharge -V meidian2

[0026] Among them, V meidian2 It is the median voltage of all battery cells at the sampling moment when discharged to the low-end voltage, V diff2 is the voltage difference between battery A2 and the median voltage;

[0027] Finally, fixed thresholds threshold1 and threshold2 are set. The number of times batteries A1 and A2 exceed the thresholds during a specified charge and discharge cycle is used to determine whether the inconsistency between the two cells significantly reduces the available capacity of the battery cluster.

[0028] Furthermore, the method for determining whether the number of times battery A1 and battery A2 exceeds a fixed threshold is as follows:

[0029]

[0030] Alarm1 and Alarm2 represent whether battery A1 and battery A2 exceed thresholds in this charge-discharge cycle, respectively. Based on the data to be tested and the required detection sensitivity, if A1 exceeds threshold 1 for m consecutive charge cycles, and if A2 exceeds threshold 2 for n consecutive charge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

[0031] Step 3: According to the charging high-end voltage V charge and the discharge low-end voltage V discharge , calculate the remaining rechargeable capacity QR of the monomer with larger inconsistency i1 and the remaining discharge capacity QD i2 ;

[0032] First, calculate the remaining charge capacity QR i1 and the remaining discharge capacity QD i2 , calculated as follows:

[0033]

[0034] Where t_charge0 indicates that A1, the first battery, reaches the high-end charging voltage V charge time, and t_charge i1 Indicates that the i1th battery reaches the high-end charging voltage V charge Time, QR i1 Indicates the remaining chargeable capacity of the i1th battery; similarly, t_discharge0 indicates that A2 is the first to reach the low-end discharge voltage V discharge time, and t_discharge i2 Indicates that the i2th reaches the low end discharge voltage V discharge Time, QD i2 Indicates the remaining dischargeable capacity of the i2th battery;

[0035] Finally, fixed thresholds threshold3 and threshold4 are set to determine whether there are other inconsistent abnormal cells that affect the available battery capacity besides battery A1 and battery A2. The determination method is as follows:

[0036]

[0037] Among them, Alarm3 i1 Represents the i1th reaching the high-end charging voltage V charge Battery cell A_charge i1 In a charging cycle, whether the threshold value threshold3 is exceeded, the battery cell A_charge is set according to the data to be detected and the detection sensitivity.i1 When the number of consecutive charging cycles exceeds the threshold value threshold3, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

[0038]

[0039] Among them, Alarm4 i2 Represents the i2th reaching the high-end discharge voltage V discharge Battery cell A_discharge i2 Whether a discharge cycle exceeds threshold threshold4, according to the data to be detected and the detection sensitivity requirements, it is set that when A_dischargei2 exceeds threshold threshold4 for q consecutive discharge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster;

[0040] Step 4: Set a time window of length L, according to the pressure difference V diff , and the changes in the remaining chargeable capacity and the remaining dischargeable capacity of the battery cells in this time window, to obtain the changes in the inconsistency degree of these cells;

[0041] First, according to the voltage difference V diff , Remaining chargeable capacity QR i1 and the remaining discharge capacity QD i2 The sliding time window L is used to iteratively calculate the respective linear regression slopes as follows:

[0042] First, define the window and variables: Assume that the data sequence is A = [A1, A2, ..., An], with a total of n cycles and a time window length of L. For each starting position j from 1 to n-L+1, calculate the linear regression slope of Aj, Aj+1, ...Aj+L-1 in the window. The linear regression slope is calculated as follows:

[0043]

[0044] Where aj represents the linear regression slope of the jth time window, V diff , QR i1 and QD i2 Substituting into the above formulas respectively, we get the linear regression slope a1 j , a2 j , a3i1 j and a4i2 j ;

[0045] Secondly, set fixed thresholds threshold5, threshold6, threshold7 and threshold8 to determine whether there are cases where the thresholds are exceeded continuously, and then determine whether the degree of inconsistency of these monomers is further deepened.

[0046] Finally, set the number of times a threshold can be exceeded in a specific number of cycles, and trigger an alarm when the number is exceeded;

[0047] Furthermore, the formula for determining whether to further deepen is as follows:

[0048]

[0049] in, is the indicator function, when a1 (t-k) When the threshold is exceeded, its value is 1, otherwise it is 0; when the a1 values ​​of the past z5 cycles are all greater than the threshold, the sum result is z5, which satisfies the condition of triggering the alarm; Alarm7 i1 Represents the i1th reaching the high-end charging voltage V charge The battery cell A_chargei1 exceeds the threshold threshold7 for z7 consecutive charging cycles.

[0050] (3) Beneficial effects

[0051] One of the above technical solutions has the following advantages or beneficial effects:

[0052] In order to solve the problem that the existing technology is not convenient for real-time detection and positioning of inconsistent cells in a battery cluster composed of a large number of batteries when in use, as well as the detection parameters are single, the cost is high and the scope of application is poor, which leads to the situation that a battery cluster composed of a large number of batteries is prone to shortened life, battery pack performance degradation and safety reduction due to the inability to handle inconsistency problems in a timely manner, the battery management system collects various data of the battery cluster in real time, and performs duplicate value removal, disordered value sorting, abnormal value removal and missing value filling cleaning on the data, and then sorts and stores the voltage data, and automatically selects the batteries in the battery cluster for voltage difference calculation according to the charge and discharge strategy and the characteristics of the upper and lower cut-off voltages of charge and discharge, and then calculates the voltage difference through The pressure difference is compared with the set threshold to make a preliminary record of inconsistent cells that affect the available capacity of the battery pack. Then, the remaining rechargeable power and the remaining dischargeable power are used to further diagnose and locate the inconsistent cells, realizing real-time positioning while preventing the situation where the detection parameters are single. At the same time, the detection cost is low and the application range is wide, and it is less affected by the environment. It is compatible with lithium iron phosphate and ternary lithium batteries, which facilitates the timely detection and positioning of inconsistent battery cells in a battery cluster composed of a large number of batteries, and can give priority to locating cells with a relatively large degree of comprehensive inconsistency without the need for additional parameter detection. It is convenient and fast, and facilitates timely processing of inconsistent single cells to ensure the life, performance and safety of the battery cluster during use. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0054] Figure 1 This is a flow chart of the battery cluster inconsistency abnormal single cell detection of the present invention;

[0055] Figure 2 This is a flow chart of the algorithm for detecting inconsistent cells in a battery cluster according to the present invention;

[0056] Figure 3 This is a flow chart of the algorithm for detecting the degree of change of inconsistent monomers in a battery cluster according to the present invention; DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below with reference to the examples, but the embodiments of the present invention are not limited thereto.

[0058] The present invention provides a method for detecting and locating inconsistency abnormal cells in a battery cluster, comprising the following steps:

[0059] Step 1: During the charging and operation of the energy storage system, the battery management system collects data from the batteries in the battery cluster in real time and performs data preprocessing;

[0060] First, various data of the batteries in the battery cluster are collected, including time, total voltage, current, single cell maximum voltage, single cell minimum voltage, single cell maximum temperature, single cell minimum temperature, battery cell voltage and other data, and the above data are recorded and stored;

[0061] Finally, the data is preprocessed by removing duplicate values, sorting disordered values, eliminating outliers, and filling missing values, so that the data can be cleaned before fault diagnosis to improve data quality and increase the accuracy and reliability of subsequent algorithms. The screened valid battery cell voltage data is stored and recorded. The battery voltage data is expressed as:

[0062]

[0063] Where m is the number of battery cells in the battery pack, and n is the number of sampling points;

[0064] Step 2: In order to comprehensively consider the inconsistency of abnormal cells in the battery cluster that restrict the available capacity of the battery cluster, the first cell to reach the upper cut-off voltage V upper Battery A1 with discharge cut-off voltage V lower Perform preliminary inconsistency diagnosis on battery A2;

[0065] First, according to the charging cut-off voltage V specified by the battery cluster upper and the discharge cut-off voltage V lower , set a fixed charging high-end voltage V charge and the discharge low-end voltage V discharge ;

[0066] Secondly, according to the charging high-end voltage V of the battery cluster charge and the discharge low-end voltage V discharge , find the two cells A1 and A2 in the battery cluster that reach the upper cut-off voltage of charge and the lower cut-off voltage of discharge the fastest, and calculate the voltage difference V between these two cells and the median of all cells at the corresponding moment diff ;

[0067] Furthermore, the voltage difference V of battery A1 diff The calculation formula is:

[0068] V diff1 =V charge -V meidian1

[0069] Among them, V meidian1 It is the median voltage of all battery cells at the sampling moment when charging to the high-end voltage, V diff1 is the voltage difference between battery A1 and the median voltage;

[0070] The voltage difference of battery A2 is V diffThe calculation formula is:

[0071] V diff2 =V discharge -V meidian2

[0072] Among them, V meidian2 It is the median voltage of all battery cells at the sampling moment when discharged to the low-end voltage, V diff2 is the voltage difference between battery A2 and the median voltage;

[0073] Finally, fixed thresholds threshold1 and threshold2 are set. The number of times batteries A1 and A2 exceed the thresholds during a specified charge and discharge cycle is used to determine whether the inconsistency between the two cells significantly reduces the available capacity of the battery cluster.

[0074] Furthermore, the method for determining whether the number of times battery A1 and battery A2 exceeds a fixed threshold is as follows:

[0075]

[0076] Alarm1 and Alarm2 represent whether battery A1 and battery A2 exceed thresholds in this charge-discharge cycle, respectively. Based on the data to be tested and the required detection sensitivity, if A1 exceeds threshold 1 for m consecutive charge cycles, and if A2 exceeds threshold 2 for n consecutive charge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

[0077] Step 3: According to the charging high-end voltage V charge and the discharge low-end voltage V discharge , calculate the remaining rechargeable capacity QR of the monomer with larger inconsistency i1 and the remaining discharge capacity QD i2 ;

[0078] First, calculate the remaining charge capacity QR i1 and the remaining discharge capacity QD i2 , calculated as follows:

[0079]

[0080] Where t_charge0 indicates that A1, the first battery, reaches the high-end charging voltage V charge time, and t_charge i1 Indicates that the i1th battery reaches the high-end charging voltage V charge Time, QR i1Indicates the remaining chargeable capacity of the i1th battery; similarly, t_discharge0 indicates that A2 is the first to reach the low-end discharge voltage V discharge time, and t_discharge i2 Indicates that the i2th reaches the low end discharge voltage V discharge Time, QD i2 Indicates the remaining dischargeable capacity of the i2th battery;

[0081] Finally, fixed thresholds threshold3 and threshold4 are set to determine whether there are other inconsistent abnormal cells that affect the available battery capacity besides battery A1 and battery A2. The determination method is as follows:

[0082]

[0083] Among them, Alarm3 i1 Represents the i1th reaching the high-end charging voltage V charge Battery cell A_charge i1 In a charging cycle, whether the threshold value threshold3 is exceeded, the battery cell A_charge is set according to the data to be detected and the detection sensitivity. i1 When the number of consecutive charging cycles exceeds the threshold value threshold3, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

[0084]

[0085] Among them, Alarm4 i2 Represents the i2th reaching the high-end discharge voltage V discharge Battery cell A_discharge i2 Whether a discharge cycle exceeds threshold threshold4, according to the data to be detected and the detection sensitivity requirements, it is set that when A_dischargei2 exceeds threshold threshold4 for q consecutive discharge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster;

[0086] Step 4: Set a time window of length L, according to the pressure difference V diff , and the changes in the remaining chargeable capacity and the remaining dischargeable capacity of the battery cells in this time window, to obtain the changes in the inconsistency degree of these cells;

[0087] First, according to the voltage difference V diff , Remaining chargeable capacity QR i1 and the remaining discharge capacity QD i2The sliding time window L is used to iteratively calculate the respective linear regression slopes as follows:

[0088] First, define the window and variables: Assume that the data sequence is A = [A1, A2, ..., An], with a total of n cycles and a time window length of L. For each starting position j from 1 to n-L+1, calculate the linear regression slope of Aj, Aj+1, ...Aj+L-1 in the window. The linear regression slope is calculated as follows:

[0089]

[0090] Where aj represents the linear regression slope of the jth time window, V diff , QR i1 and QD i2 Substituting into the above formulas respectively, we get the linear regression slope a1 j , a2 j , a3i1 j and a4i2 j ;

[0091] Secondly, set fixed thresholds threshold5, threshold6, threshold7 and threshold8 to determine whether there are cases where the thresholds are exceeded continuously, and then determine whether the degree of inconsistency of these monomers is further deepened.

[0092] Finally, set the number of times a threshold can be exceeded in a specific number of cycles, and trigger an alarm when the number is exceeded;

[0093] Furthermore, the formula for determining whether to further deepen is as follows:

[0094]

[0095] in, is the indicator function, when a1 (t-k) When the threshold is exceeded, its value is 1, otherwise it is 0; when the a1 values ​​of the past z5 cycles are all greater than the threshold, the sum result is z5, which satisfies the condition of triggering the alarm; Alarm7 i1 Represents the i1th reaching the high-end charging voltage V charge The battery cell A_chargei1 exceeds the threshold threshold7 for z7 consecutive charging cycles.

[0096] The battery management system collects various data of the battery cluster batteries in real time and performs cleaning processing such as deduplication, sorting of disordered values, elimination of outliers, and filling of missing values. The voltage data is then sorted and stored, and the batteries in the battery cluster are automatically selected for voltage difference calculation based on the charge and discharge strategy and the characteristics of the upper and lower cut-off voltages of charge and discharge. The voltage difference is then compared with the set threshold to preliminarily record the inconsistent cells that affect the available capacity of the battery pack. The remaining chargeable capacity and the remaining dischargeable capacity are then used to further diagnose and locate the inconsistent cells, achieving real-time positioning while preventing the situation where the detection parameters are single. At the same time, the detection cost is low, the application range is wide, and it is less affected by the environment. It is compatible with both lithium iron phosphate and ternary lithium batteries, facilitating the timely detection and location of inconsistent battery cells in a battery cluster composed of a large number of batteries, and can prioritize the location of cells with relatively large comprehensive inconsistencies without the need for additional parameter testing. The system is convenient and fast, facilitating the timely processing of inconsistent cells, and ensuring the life, performance, and safety of the battery cluster during use.

[0097] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, from all perspectives, the embodiments should be regarded as illustrative and non-restrictive. The scope of the present invention is defined by the appended claims, not the foregoing description, and it is intended that all variations that come within the meaning and range of equivalents of the claims be included within the present invention. Any reference signs in the claims should not be construed as limiting the claim to which they relate.

[0098] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for detecting and locating inconsistent abnormal cells in a battery cluster, characterized by: The following steps are involved: Step 1: During the charging and operation of the energy storage system, the battery management system collects data from the batteries in the battery cluster in real time and performs data preprocessing; First, various data of the batteries in the battery cluster are collected, recorded and stored; Finally, the data is preprocessed by removing duplicate values, sorting out of order values, removing outliers, and filling in missing values, and the filtered valid battery cell voltage data is stored and recorded; Step 2: The first one to reach the upper cut-off voltage V upper The battery A1 has a discharge cut-off voltage V lower Perform preliminary inconsistency diagnosis on battery A2; First, according to the charging cut-off voltage V specified by the battery cluster upper and the discharge cut-off voltage V lower , set a fixed charging high-end voltage V charge and the discharge low-end voltage V discharge ; Secondly, according to the charging high-end voltage V of the battery cluster charge and the discharge low-end voltage V discharge , find the two cells A1 and A2 in the battery cluster that reach the upper cut-off voltage of charge and the lower cut-off voltage of discharge the fastest, and calculate the voltage difference V between these two cells and the median of all cells at the corresponding moment diff ; Finally, fixed thresholds threshold1 and threshold2 are set. The number of times batteries A1 and A2 exceed the thresholds during a specified charge and discharge cycle is used to determine whether the inconsistency between the two cells significantly reduces the available capacity of the battery cluster. Step 3: According to the charging high-end voltage V charge and the discharge low-end voltage V discharge , calculate the remaining rechargeable capacity QR of the monomer with larger inconsistency i1 and the remaining discharge capacity QD i2 ; First, calculate the remaining chargeable capacity and the remaining dischargeable capacity; Finally, fixed thresholds threshold3 and threshold4 are set to determine whether there are other inconsistent abnormal cells that affect the available battery capacity besides battery A1 and battery A2; Step 4: Set a time window of length L, according to the pressure difference V diff , and the changes in the remaining chargeable capacity and the remaining dischargeable capacity of the battery cells in this time window, to obtain the changes in the inconsistency degree of these cells; First, according to the voltage difference V diff , Remaining chargeable capacity QR i1 and the remaining discharge capacity QD i2 The respective linear regression slopes are iteratively calculated using a sliding time window L; Secondly, set fixed thresholds threshold5, threshold6, threshold7 and threshold8 to determine whether the thresholds are exceeded continuously, and then determine whether the inconsistency of these monomers is further deepened; Finally, set the number of times a threshold can be exceeded within a specific number of cycles, triggering an alarm.

2. The method for detecting and locating inconsistent abnormal cells in a battery cluster according to claim 1, characterized in that: The voltage difference V of battery A1 in step 2 diff The calculation formula is: V diff1 =V charge -V meidian1 Among them, V meidian1 It is the median voltage of all battery cells at the sampling moment when charging to the high-end voltage, V diff1 is the voltage difference between battery A1 and the median voltage; The voltage difference of battery A2 is V diff The calculation formula is: V diff2 =V discharge -V meidian2 Among them, V meidian2 It is the median voltage of all battery cells at the sampling moment when discharged to the low-end voltage, V diff2 is the voltage difference between the A2 cell and the median voltage.

3. The method for detecting and locating inconsistent abnormal cells in a battery cluster according to claim 1, wherein: In step 2, the method for determining whether the number of times battery A1 and battery A2 exceeds the fixed threshold is as follows: Among them, Alarm1 and Alarm2 respectively represent whether battery A1 and battery A2 exceed the threshold value in this charge and discharge cycle. Based on the data to be detected and the detection sensitivity requirements, it is set that when A1 exceeds threshold threshold1 for m consecutive charge cycles, and when A2 exceeds threshold threshold2 for n consecutive charge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

4. The method for detecting and locating inconsistent abnormal cells in a battery cluster according to claim 1, wherein: The remaining chargeable capacity and the remaining dischargeable capacity in step 3 are calculated as follows: Where t_charge0 indicates that A1, the first battery, reaches the high-end charging voltage V charge time, and t_charge i1 Indicates that the i1th battery reaches the high-end charging voltage V charge Time, QR i1 Indicates the remaining chargeable capacity of the i1th battery; similarly, t_discharge0 indicates that A2 is the first to reach the low-end discharge voltage V discharge time, and t_discharge i2 Indicates that the i2th reaches the discharge low end voltage V discharge Time, QD i2 Indicates the remaining discharge capacity of the i2th battery.

5. The method for detecting and locating inconsistent abnormal cells in a battery cluster according to claim 1, wherein: In step 3, the determination method for whether there are other abnormal cells that affect the available capacity of the battery in addition to battery A1 and battery A2 is as follows: Among them, Alarm3 i1 Represents the i1th reaching the high-end charging voltage V charge Battery cell A_charge i1 In a charging cycle, whether the threshold value threshold3 is exceeded, the battery cell A_charge is set according to the data to be detected and the detection sensitivity. i1 When the number of consecutive charging cycles exceeds the threshold value threshold3, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster. Among them, Alarm4 i2 Represents the i2th reaching the high-end discharge voltage V discharge Battery cell A_discharge i2 Whether a discharge cycle exceeds the threshold threshold4, according to the data to be detected and the detection sensitivity requirements, it is set that when A_dischargei2 exceeds the threshold threshold4 for q consecutive discharge cycles, it is determined that there is a large inconsistency, which significantly affects the available capacity of the battery cluster.

6. The method for detecting and locating inconsistent abnormal cells in a battery cluster according to claim 1, wherein: The linear regression slope in step 4 is calculated as follows: First, define the window and variables: assume that the data sequence is A = [A1, A2, ..., An], there are n cycles in total, the time window length is L, and for each starting position j from 1 to n-L+1, calculate the linear regression slope of Aj, Aj+1, ...Aj+L-1 in the window.

Images