Battery short plate cell detection method and system and readable storage medium

Abstract

The invention discloses a battery short plate cell detection method and system and a readable storage medium, and the method comprises the steps: S1, calculating a cell global abnormal threshold value and an extreme and tail end abnormal threshold value based on a big data statistics method; s2, carrying out global anomaly detection and tail end anomaly detection on the battery cell; when the global anomaly of the battery cell is detected and the global detection value of the battery cell exceeds a global anomaly threshold value, judging that the battery cell is global anomaly; when the tail abnormal end of the battery cell is detected and the tail end detection value of the battery cell exceeds a tail end abnormal threshold value, determining that the tail end is abnormal; and S3, carrying out final detection on the charging and discharging data corresponding to the global abnormal cell and the charging and discharging data corresponding to the terminal abnormal cell, analyzing the data volume of the abnormal cell exceeding the abnormal threshold, and judging the position of the short-board cell according to the distribution position of the data volume. According to the invention, the short-plate battery cell in the battery can be detected more accurately, more practically and more quickly, and the inconsistency among the single batteries in the battery pack can be effectively relieved or even eliminated.

Description

Technical Field

[0001] This invention relates to the field of battery system management technology, and more specifically, to a method, system, and readable storage medium for detecting battery defect cells. Background Technology

[0002] With the rapid popularization of new energy vehicles, research on electric vehicle power batteries is receiving increasing attention, and the detection of weakest cells is a crucial research area. A weakest cell refers to the worst-performing cell in a battery pack, and it is a major cause of shortened battery pack life, unstable performance, reduced usable capacity, and safety hazards. Therefore, it is necessary to identify weakest cells through specific methods to facilitate repair or replacement and extend the battery pack's lifespan.

[0003] In existing technologies, charging identification or discharging identification methods are generally used to identify defective battery cells. These two methods are relatively simple for detecting defective battery cells, and the detection data may not be accurate. Summary of the Invention

[0004] In view of the above problems, the purpose of this invention is to provide a method, system, and readable storage medium for detecting battery defect cells.

[0005] The first aspect of the present invention provides a method for detecting battery cells with short circuits, the method comprising: S1: calculating the global anomaly threshold and the extreme end anomaly threshold of the battery cell based on big data statistical methods;

[0006] S2: Perform global anomaly detection and end-point anomaly detection on the battery cell;

[0007] When a global anomaly is detected in a battery cell, if the global detection value exceeds the global anomaly threshold, it is determined to be a global anomaly.

[0008] When detecting abnormalities at the end of a battery cell, if the detected value at the end of the cell exceeds the abnormality threshold, it is determined to be an abnormality at the end.

[0009] S3: Perform final detection on the charging and discharging data corresponding to cells with global anomalies and those with end-stage anomalies. Analyze the amount of data exceeding the anomaly threshold for the abnormal cells, and determine the location of the weakest cells by analyzing the distribution of this data. Specifically, statistically analyze the data distribution according to quantiles; cells within the abnormal distribution range are identified as weakest cells, and their locations are recorded.

[0010] Preferably, the global anomaly detection includes:

[0011] S2.1.1: Calculate the global offset of each cell in the battery charging data, including the global maximum upper offset and the global maximum lower offset;

[0012] S2.1.2: Merge all global offset data of battery data; that is, merge all global maximum upper offsets to obtain a global maximum upper offset set, and merge all global maximum lower offsets to obtain a global maximum lower offset set;

[0013] S2.1.3: Set global anomaly thresholds for battery data, including global upper offset anomaly thresholds and global lower offset anomaly thresholds;

[0014] S2.1.4: Calculate the global maximum upward offset and maximum downward offset of a single cell in the battery to be tested; and compare the global maximum upward offset of a single cell with the global upward offset anomaly threshold.

[0015] When the global offset is below the high voltage global upward offset abnormal threshold, or below the global downward offset abnormal threshold, the cell is determined to be in global abnormality, and the cell's specific location is recorded.

[0016] Preferably, S2.1.1 specifically includes:

[0017] The average of the voltage data of each cell in the battery from at least two charging data points is subtracted from the average voltage of all cells in the battery; the offset of each cell voltage relative to the average voltage of all cells is obtained; the maximum value of the voltage offset of each cell in the battery is taken as the global maximum upward offset of the battery.

[0018] The average of at least two discharge data points for each cell voltage in the battery is collected, and the average of all cell voltages in the battery is subtracted from the average of all cell voltages in the battery. The offset of each cell voltage relative to the average of all cell voltages is obtained, and the minimum cell voltage offset is taken as the global maximum downward offset of the battery.

[0019] Preferably, the global offset anomaly threshold includes a mild anomaly threshold g. a1 and the threshold e for extreme anomalies a1

[0020] The global downshift anomaly threshold includes the mild anomaly downshift threshold e. b1 and the threshold e under extreme anomalies b1 ;

[0021] S2.1.3 specifically refers to:

[0022] Box plot anomaly detection is used to detect global offset data. Let Q be the upper quartile of the global offset data. 31 The lower quartile of the global offset data is Q. 11 The global quartile interval is ΔQ1; the threshold is set by controlling the mild offset coefficient w1 and the extreme offset coefficient w2:

[0023] Set a mild anomaly threshold g in the global maximum upper offset set. a1 and the threshold e for extreme anomalies a1 ;

[0024] Set a mild anomaly threshold g in the global maximum lower offset set. b1 and the threshold e under extreme anomalies b1 ;

[0025] Therefore, g a1 =Q 31 +w1×ΔQ1

[0026] e a1 =Q 31 +w2×ΔQ1

[0027] g b1 =Q 11 -w1×ΔQ1

[0028] e a1 =Q 11 -w2×ΔQ1.

[0029] Preferably, the end-point anomaly detection includes the following steps:

[0030] S2.2.1: Collect charging and discharging data for each battery each time, and calculate the end offset data for each battery, that is, calculate the maximum upper and lower offsets at the end of each charge / discharge cutoff time.

[0031] S2.2.2: Merge the maximum upward offset of the end of all batteries to obtain a set of maximum upward offsets at the end; merge the maximum downward offset of the end of all batteries to obtain a set of maximum downward offsets at the end.

[0032] S2.2.3: Set the threshold for end-of-battery offset data, including the threshold for end-of-battery upper offset and the threshold for end-of-battery lower offset;

[0033] S2.2.4: Calibrate the threshold for abnormal downward offset at the end;

[0034] S2.2.5: Calculate the maximum upward and downward offset of the end of a single cell in the battery to be tested; and compare the maximum upward offset of the end of a single cell with the global upward offset anomaly threshold.

[0035] When the terminal offset reaches the high voltage global upward offset abnormal threshold, or is lower than the terminal downward offset abnormal threshold, the cell is determined to be terminal abnormal, and the specific location of the cell is recorded.

[0036] Preferably, the maximum upward offset at the end is calculated by subtracting the second highest cell voltage from the highest cell voltage at the time of each battery charge cutoff.

[0037] The maximum downward offset at the end is calculated by subtracting the second lowest cell voltage from the lowest cell voltage at the time of each battery discharge cutoff.

[0038] Preferably, the end-off offset anomaly threshold includes a mild anomaly threshold g. a2 and the threshold e for extreme anomalies a2

[0039] The end-point downward offset anomaly threshold includes the mild anomaly downward threshold e. b2 and the threshold e under extreme anomalies b2 ;

[0040] S2.2.3 specifically refers to:

[0041] Box plot anomaly detection is used to detect end-point offset data. Let Q be the upper quartile of the end-point offset data. 32 The lower quartile of the end offset data is Q. 12 The final quartile interval is ΔQ2; the threshold is set by controlling the mild offset coefficient w3 and the extreme offset coefficient w4:

[0042] Set a mild anomaly threshold g in the set of maximum upper offsets at the end. a3 and the threshold e for extreme anomalies a3 ;

[0043] Set a mild anomaly threshold e in the set of maximum downward offsets at the end. b3 and the threshold e under extreme anomalies b4 ;

[0044] Therefore, g a2 =Q 32 +w3×ΔQ2

[0045] e a2 =Q 32 +w4×ΔQ2

[0046] g b2 =Q 12 -w3×ΔQ2

[0047] e a2 =Q 12 -w4×ΔQ2.

[0048] Preferably, S2.2.4 specifically involves setting a threshold g higher than the threshold for mild anomalies. a2 and below the mild abnormality threshold e b2The data is used as the calibration target, and the data within the calibration target is recalibrated according to the method for setting the end abnormality threshold described in S2.2.3 to recalibrate the end abnormality threshold.

[0049] A second aspect of the present invention provides a battery short cell detection system, including a memory and a processor. The memory includes a battery short cell detection method program, which, when executed by the processor, implements a battery short cell detection method step.

[0050] A third aspect of the present invention provides a computer-readable storage medium including a battery short cell detection method program, wherein when the battery short cell detection method program is executed by a processor, it implements the steps of the battery short cell detection method.

[0051] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention detects anomalies from two dimensions—global anomalies and end-of-pipe anomalies—based on multiple charge-discharge segment data from multiple batteries. It extracts the vertical offset features of the battery cells and calculates the quantiles of the battery cell data using box plots from big data statistical analysis methods. Based on this, it calculates the temperature threshold and anomaly threshold for the short-pole cells. Then, based on these thresholds, it performs short-pole cell anomaly detection from both global and end-of-pipe dimensions, identifying the anomaly points, i.e., the short-pole cells, and locating their positions. This invention can detect short-pole cells in batteries more accurately, practically, and quickly. It can effectively alleviate or even eliminate inconsistencies between individual cells in a battery pack, improving the battery pack's lifespan, performance, usable capacity, and safety. Attached Figure Description

[0052] Figure 1 This is a flowchart of a battery short-circuit cell detection method as described in Example 1.

[0053] Figure 2 This is a flowchart of the global anomaly detection process.

[0054] Figure 3 This is a schematic diagram of the box plot principle.

[0055] Figure 4 This is a flowchart for end-point anomaly detection.

[0056] Figure 5 This is a distribution chart of the global maximum upward offset monitoring data in the global anomaly detection of lithium iron phosphate battery short-board cells.

[0057] Figure 6 This is a distribution chart of the global maximum downward offset monitoring data in the global anomaly detection of lithium iron phosphate battery short-board cells.

[0058] Figure 7 This is a graph showing the monitoring and detection results of the maximum upward offset in the global anomaly detection of lithium iron phosphate battery cells with short circuits.

[0059] Figure 8 This is a graph showing the monitoring and detection results of the maximum downward offset in the global anomaly detection of lithium iron phosphate battery cells with short circuits.

[0060] Figure 9 It is a charging curve of the short-board battery cell at the beginning of operation in the global anomaly detection.

[0061] Figure 10 It is a charging curve of the short-circuit battery cell during a long period of operation in the global anomaly detection.

[0062] Figure 11 This is a graph showing the monitoring and detection results of the maximum upward offset at the end of the fault detection terminal of a lithium iron phosphate battery short-circuit cell.

[0063] Figure 12 This is a graph showing the monitoring and detection results of the maximum downward offset at the end of the fault detection terminal of a lithium iron phosphate battery short-circuit cell.

[0064] Figure 13 This is a charging curve of the battery at the beginning of operation during end-of-life anomaly detection.

[0065] Figure 14 It is the charging curve of the battery after a relatively long period of time during end-of-life anomaly detection.

[0066] Figure 15 It is the charging curve of the battery at the beginning of operation in global anomaly detection and end-of-life anomaly detection.

[0067] Figure 16 It is the charging curve of the battery after a relatively long time in global anomaly detection and end-of-pipe anomaly detection. Detailed Implementation

[0068] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0069] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0070] Example 1

[0071] like Figure 1As shown in the figure, this embodiment discloses a method for detecting battery cell defects, the method comprising:

[0072] S1: Based on big data statistical methods, calculate the global anomaly threshold and extreme end anomaly threshold of the battery cell;

[0073] S2: Perform global anomaly detection and end-point anomaly detection on the battery cell;

[0074] When a global anomaly is detected in a battery cell, if the global detection value exceeds the global anomaly threshold, it is determined to be a global anomaly.

[0075] When detecting abnormalities at the end of a battery cell, if the detected value at the end of the cell exceeds the abnormality threshold, it is determined to be an abnormality at the end.

[0076] S3: Perform final detection on the charging and discharging data corresponding to cells with global anomalies and those with end-stage anomalies. Analyze the amount of data exceeding the anomaly threshold for the abnormal cells, and determine the location of the weakest cells by analyzing the distribution of this data. Specifically, statistically analyze the data distribution according to quantiles; cells within the abnormal distribution range are identified as weakest cells, and their locations are recorded.

[0077] It should be noted that the final detection is based on the results of global anomaly detection and end-point anomaly detection. Specifically, the number of times the abnormal cell exceeds the anomaly threshold is counted to obtain the amount of data of the abnormal cell exceeding the anomaly threshold. Then, the data amount is distributed statistically according to the quantile. The cells that are in the distribution anomaly interval are the short-board cells. The location of the short-board cells is the location of the cells with global anomalies or end-point anomalies that have been obtained by local detection or end-point detection.

[0078] like Figure 2 As shown, the global anomaly detection includes:

[0079] S2.1.1: Calculate the global offset of each cell in the battery charging data, including the global maximum upper offset and the global maximum lower offset, specifically:

[0080] The average of the voltage data of each cell in the battery from at least two charging data points is subtracted from the average voltage of all cells in the battery; the offset of each cell voltage relative to the average voltage of all cells is obtained; the maximum value of the voltage offset of each cell in the battery is taken as the global maximum upward offset of the battery.

[0081] The average of at least two discharge data points for each cell voltage in the battery is collected, and the average of all cell voltages in the battery is subtracted from the average of all cell voltages in the battery. The offset of each cell voltage relative to the average of all cell voltages is obtained, and the minimum cell voltage offset is taken as the global maximum downward offset of the battery.

[0082] In other words, this application uses the average of multiple charge and discharge data for each cell voltage to subtract the average of all cell voltages to obtain the offset of each cell relative to the overall average cell voltage. The largest upper offset (i.e., the maximum offset value) is retained as the global maximum upper offset, and the largest lower offset (i.e., the minimum offset value) is retained as the global maximum lower offset.

[0083] S2.1.2: Merge all global offset data of battery data; that is, merge all global maximum upper offsets to obtain a global maximum upper offset set, and merge all global maximum lower offsets to obtain a global maximum lower offset set.

[0084] It should be noted that in this application, for all battery data of each type (e.g., lithium iron phosphate, ternary lithium battery), the global maximum upper offset and global maximum lower offset are calculated according to S2.1.1. Then, all the calculated results of this type of battery data are merged together. For example, if there are 100 battery data of lithium iron phosphate, and the global maximum upper offset and global maximum lower offset are retained for each battery, then there will be a total of 200 global offset data after merging, of which there are 100 global maximum upper offsets and 100 global maximum lower offsets.

[0085] S2.1.3: Set global anomaly thresholds for battery data, including global upper offset anomaly thresholds and global lower offset anomaly thresholds, specifically:

[0086] Box plot anomaly detection is used to detect global offset data. Let Q be the upper quartile of the global offset data. 31 The lower quartile of the global offset data is Q. 11 The global quartile interval is ΔQ1; the threshold is set by controlling the mild offset coefficient w1 and the extreme offset coefficient w2:

[0087] Set a mild anomaly threshold g in the global maximum upper offset set. a1 and the threshold e for extreme anomalies a1 ;

[0088] Set a mild anomaly threshold g in the global maximum lower offset set. b1 and the threshold e under extreme anomalies b1 ;

[0089] Therefore, g a1 =Q 31 +w1×ΔQ1

[0090] e a1 =Q 31 +w2×ΔQ1

[0091] g b1 =Q 11-w1×ΔQ1

[0092] e a1 =Q 11 -w2×ΔQ1.

[0093] As a specific implementation, an anomaly threshold is set for all global offset data of a certain type of battery data obtained by merging in S2.1.2, and the box plot method is used to calculate the upper threshold g for mild anomalies. a1 and the threshold e for extreme anomalies a1 Threshold g under mild anomalies b1 and the threshold e under extreme anomalies b1 .

[0094] Meanwhile, the box plot anomaly detection method is used to detect the global offset data. Thresholds are set by controlling the mild offset coefficient w1 and the extreme offset coefficient w2. In this embodiment, w1 = 1.5 and w2 = 3.

[0095] Therefore, as Figure 3 As shown, the specific threshold setting formula in this embodiment is as follows:

[0096] The upper quartile of the global offset data is Q 31 The lower quartile of the global offset data is Q. 11 The global quartile interval is ΔQ1;

[0097] Mild abnormal upper threshold g a1 =Q 31 +1.5×ΔQ1

[0098] Extreme anomaly threshold e a1 =Q 31 +3×ΔQ1

[0099] Threshold g under mild anomaly b1 =Q 11 -1.5×ΔQ1

[0100] Threshold e under extreme anomalies a1 =Q 11 -3×ΔQ1

[0101] It should be noted that, in this embodiment, after obtaining the global abnormal threshold of the battery cell, it can also be saved in a file in a specified location, and the threshold can be directly called in the subsequent short-board battery cell abnormality detection algorithm.

[0102] S2.1.4: Calculate the global maximum upward offset and maximum downward offset of a single cell in the battery to be tested; and compare the global maximum upward offset of a single cell with the global upward offset anomaly threshold.

[0103] When the global offset is below the high voltage global upward offset abnormal threshold, or below the global downward offset abnormal threshold, the cell is determined to be in global abnormality, and the cell's specific location is recorded.

[0104] It should be noted that the method for calculating the global maximum upward offset and maximum downward offset of the individual cells in the battery under test in this step is the same as the calculation method in step S2.1.1. The average voltage of each cell is subtracted from the average voltage of all cells to obtain the offset of each cell relative to the overall average cell voltage, and all offsets are retained. This includes the global maximum upward offset and the global maximum downward offset.

[0105] In this embodiment, the global anomaly threshold can be retrieved from a file located at a specified position, and then the global maximum upper offset and global maximum lower offset can be compared with the corresponding threshold.

[0106] When the global maximum upper offset is greater than the upper threshold of mild anomalies, g a1 Or extreme anomalies above the threshold e a1 When this occurs, it is determined to be a global anomaly, and the specific location of the battery cell is recorded.

[0107] In this embodiment, the data and location results of the globally abnormal cells can also be saved to a designated file in a specified directory for easy viewing and retrieval later.

[0108] like Figure 4 As shown, the terminal anomaly detection includes the following steps:

[0109] S2.2.1: Collect charging and discharging data for each battery each time, and calculate the end offset data for each battery, that is, calculate the maximum upper and lower offsets at the end of each charge / discharge cutoff time.

[0110] The maximum upward offset at the end is calculated by subtracting the second highest cell voltage from the highest cell voltage at the time of each battery charge cutoff.

[0111] The maximum downward offset at the end is calculated by subtracting the second lowest cell voltage from the lowest cell voltage at the time of each battery discharge cutoff.

[0112] In other words, the calculation method for the maximum upper offset and the maximum lower offset at the end in this embodiment is as follows: the maximum upper offset at the end is obtained by subtracting the second highest cell voltage from the highest cell voltage at each charging cutoff time, and the maximum lower offset at the end is obtained by subtracting the second lowest cell voltage from the lowest cell voltage at each discharging cutoff time (the charging start time in this embodiment).

[0113] S2.2.2: Merge the maximum upward offset of the end of all batteries to obtain a set of maximum upward offsets at the end; merge the maximum downward offset of the end of all batteries to obtain a set of maximum downward offsets at the end.

[0114] Specifically, for all battery data in each category, the maximum upward offset and maximum downward offset at the end are calculated according to S2.2.1. Then, all the calculated results for this type of battery data are merged together. For example, if there are 100 battery data for lithium iron phosphate, each battery has 10 charging data and 10 discharging data. The maximum upward offset at the end is calculated for each charging cycle and the maximum downward offset at the end is calculated for each discharging cycle, resulting in a total of 10 maximum downward offsets at the end and 10 maximum downward offsets at the end. That is, each battery has a total of 20 offsets at the end. After merging the 100 batteries, there are a total of 2000 offset data, of which there are 1000 maximum upward offsets at the end and 1000 maximum downward offsets at the end.

[0115] S2.2.3: Set the threshold for end-of-battery offset data, including the threshold for end-of-battery upper offset and the threshold for end-of-battery lower offset.

[0116] The threshold for abnormal offset at the end includes the mild abnormal offset threshold g. a2 and the threshold e for extreme anomalies a2

[0117] The end-point downward offset anomaly threshold includes the mild anomaly downward threshold e. b2 and the threshold e under extreme anomalies b2 ;

[0118] Box plot anomaly detection is used to detect end-point offset data. Let Q be the upper quartile of the end-point offset data. 32 The lower quartile of the end offset data is Q. 12 The final quartile interval is ΔQ2; the threshold is set by controlling the mild offset coefficient w3 and the extreme offset coefficient w4:

[0119] Set a mild anomaly threshold g in the set of maximum upper offsets at the end. a3 and the threshold e for extreme anomalies a3 ;

[0120] Set a mild anomaly threshold e in the set of maximum downward offsets at the end. b3 and the threshold e under extreme anomalies b4 ;

[0121] Therefore, g a2 =Q 32 +w3×ΔQ2

[0122] e a2 =Q 32+w4×ΔQ2

[0123] g b2 =Q 12 -w3×ΔQ2

[0124] e a2 =Q 12 -w4×ΔQ2.

[0125] In this embodiment, w3 = 1.5 and w4 = 3.

[0126] S2.2.4: Calibrate the threshold for end-point downshift anomalies. Specifically: set the threshold to be higher than the upper threshold for mild anomalies, g. a2 and below the mild abnormality threshold e b2 The data is used as the calibration target, and the data within the calibration target is recalibrated according to the method for setting the end abnormality threshold described in S2.2.3 to recalibrate the end abnormality threshold.

[0127] The main reason is that in many cases, the distribution of the downshifted data at the end is extremely unbalanced, with a very high proportion of extremely small data. If the initial calculated threshold is still used, the probability of false alarms will increase significantly. Therefore, the downshifted data that meets this condition is calibrated again, with data other than mild anomalies as the calibration target. The data within the target are then recalibrated according to the method for setting anomaly thresholds.

[0128] It should be noted that, in this embodiment, after obtaining the abnormal threshold at the end of the battery cell, it needs to be saved in a file in a specified location. In subsequent abnormal detection of the end of the short battery cell, the abnormal threshold at the end of the battery cell can be directly called for use.

[0129] S2.2.5: Calculate the maximum upward and downward offset of the end of a single cell in the battery to be tested; and compare the maximum upward offset of the end of a single cell with the global upward offset anomaly threshold.

[0130] When the terminal offset reaches the high voltage global upward offset abnormal threshold, or is lower than the terminal downward offset abnormal threshold, the cell is determined to be terminal abnormal, and the specific location of the cell is recorded.

[0131] In other words, the method for calculating the offset of a single battery cell in the battery under test is to subtract the second highest cell from the highest cell at each charging cutoff point to obtain the maximum upper offset at the end, and to subtract the second lowest cell from the lowest cell at each discharging cutoff point (i.e., the start of charging) to obtain the lower offset; the maximum and minimum cell positions are calculated because if there is an end-of-cell anomaly, it is necessary to locate the specific position.

[0132] In this embodiment, the end-point anomaly threshold can also be obtained from a specified file that stores end-point anomaly thresholds, and then used for end-point offset detection. The existence of an end-point anomaly and the specific cell location of the anomaly are calculated. The end-point detection results are then saved to a specified file in a specified directory for easy viewing and retrieval later.

[0133] As a specific example, this embodiment uses real charge and discharge data of a company's electric vehicle battery for testing. Shortcoming cell detection experiments were conducted on both lithium iron phosphate batteries and ternary lithium batteries.

[0134] (I) Global Anomaly Detection

[0135] Perform global anomaly detection on lithium iron phosphate battery cells. Figure 5 This paper presents a distribution map of the global maximum upward offset monitoring data in the global anomaly detection of lithium iron phosphate battery short-board cells. Figure 6 This paper presents a distribution chart of monitoring data on the global maximum downward offset in the global anomaly detection of lithium iron phosphate battery cells with short circuits. Figure 5 and Figure 6 The horizontal axis represents the magnitude of the deviation, in mv; the vertical axis is the probability density.

[0136] Figure 7 This image shows the monitoring and detection results of the maximum upward offset in the global anomaly detection of lithium iron phosphate battery cells with shortcomings. Figure 8 This is a graph showing the monitoring and detection results of the maximum downward offset in the global anomaly detection of lithium iron phosphate battery cells with short circuits.

[0137] Figure 7 and Figure 8 The horizontal axis represents the degree of deviation; the vertical axis has no practical meaning. The solid black vertical lines represent the upper and lower thresholds for mild anomalies, and the dashed black vertical lines represent the upper and lower thresholds for extreme anomalies. Each point represents a battery data point: points shaped like pentagrams are within the range of the mild anomaly thresholds and are therefore normal batteries; points shaped like triangles are between the mild and extreme anomaly thresholds and are therefore mildly abnormal batteries; and points shaped like squares are outside the extreme anomaly thresholds and are therefore extremely abnormal batteries.

[0138] The charging and discharging curves of batteries with problems detected by the cell defect test show that there is indeed a problem with cell consistency. Figure 9-10 The comparison of charging curves of the weakest cells in the global anomaly detection is shown. Figure 9-10 This is a comparison chart of the charging curves of battery cells with global anomalies and shortcomings. Figure 9 This is the charging curve of the battery when it first starts running. Figure 10 This is the charging curve of this battery after a longer period of time. From Figure 9 As can be seen from this, the charging process is relatively normal; it can be seen from... Figure 10It can be seen that there is a large deviation in the curve throughout the entire charging process.

[0139] (II) Terminal anomaly detection

[0140] End-point anomaly detection is performed on lithium iron phosphate battery cells.

[0141] The deviation between the cell voltage and the average cell voltage at the charging cutoff time and the discharging cutoff time is calculated. For charging, the maximum upward offset at the end is calculated, and for discharging, the maximum downward offset at the end is calculated.

[0142] Figure 11 This is a graph showing the monitoring and detection results of the maximum upward offset at the end of the fault detection terminal of a lithium iron phosphate battery short-circuit cell. Figure 12 This is a graph showing the monitoring and detection results of the maximum downward offset at the end of the fault detection terminal of a lithium iron phosphate battery short-circuit cell. Figure 11-12 Points in region A represent normal data, points in region B represent mild anomalies, and points in region C represent extreme anomalies.

[0143] Figure 13-14 This is a comparison chart of the charging curves of cells with abnormal short circuits at the end. Figure 13 This is the charging curve of the battery when it first starts running. Figure 14 This is the charging curve of the battery after a relatively long period of time.

[0144] from Figure 13 As can be seen, the charging process is relatively normal; from Figure 14 As can be seen, both large and small offsets exist at the end of the charging process.

[0145] Furthermore, global anomaly detection and end-point anomaly detection are not mutually exclusive; some cells may exhibit anomalies in both types of detection. For example... Figure 15-16 The example shown is illustrated.

[0146] Figure 15-16 This is a comparison chart of the charging curves of battery cells that are weak points in both global and end-point anomaly detection. Figure 15 This is the charging curve of the battery when it first starts running. Figure 16 This is the charging curve of the battery after a relatively long period of time.

[0147] from Figure 15 It can be seen that the charging process is normal; from Figure 16 It is evident that there is a curve with a larger deviation during the charging process and a curve with a smaller deviation at the end of the charging process. This indicates that the battery cell is fully charged the fastest during the charging phase and fully discharged the fastest during the discharging phase.

[0148] In summary, the electric vehicle kinetic battery short-spot cell detection algorithm proposed in this embodiment can effectively detect abnormal cells with high accuracy; it can identify which cell is abnormal, thus having high practical value; by calculating and storing the abnormal cell threshold offline, it can load and calculate in real time when calculating the short-spot cell of the battery under test, with fast response time and high real-time performance.

[0149] Example 2

[0150] This embodiment discloses a battery short-circuit cell detection system, including a memory and a processor. The memory includes a battery short-circuit cell detection method program. When the battery short-circuit cell detection method program is executed by the processor, it implements the method steps described in Embodiment 1.

[0151] Example 3

[0152] This embodiment discloses a computer-readable storage medium, which includes a battery short cell detection method program. When the battery short cell detection method program is executed by a processor, it implements the steps of the battery short cell detection method described in Embodiment 1.

[0153] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0154] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0155] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0156] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0157] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A method for detecting battery cell defects, characterized in that, The method includes: S1: Based on big data statistical methods, calculate the global anomaly threshold and extreme end anomaly threshold of the battery cell; S2: Perform global anomaly detection and end-point anomaly detection on the battery cell; When a global anomaly is detected in a battery cell, if the global detection value exceeds the global anomaly threshold, it is determined to be a global anomaly. When detecting abnormalities at the end of a battery cell, if the detected value at the end of the cell exceeds the abnormality threshold, it is determined to be an abnormality at the end. S3: Perform final detection on the charging and discharging data corresponding to the globally abnormal cells and the charging and discharging data corresponding to the terminal abnormal cells, analyze the amount of data of abnormal cells exceeding the abnormal threshold, and determine the location of the short-board cells by the distribution of the data.

2. The method for detecting battery short-circuit cells according to claim 1, characterized in that, The global anomaly detection includes: S2.1.1: Calculate the global offset of each cell in the battery charging data, including the global maximum upper offset and the global maximum lower offset; S2.1.2: Merge all global offset data of battery data; that is, merge all global maximum upper offsets to obtain a global maximum upper offset set, and merge all global maximum lower offsets to obtain a global maximum lower offset set; S2.1.3: Set global anomaly thresholds for battery data, including global upper offset anomaly thresholds and global lower offset anomaly thresholds; S2.1.4: Calculate the global maximum upward offset and maximum downward offset of a single cell in the battery to be tested; and compare the global maximum upward offset of a single cell with the global upward offset anomaly threshold. When the global offset is below the high voltage global upward offset abnormal threshold, or below the global downward offset abnormal threshold, the cell is determined to be in global abnormality, and the cell's specific location is recorded.

3. The method for detecting battery short-circuit cells according to claim 2, characterized in that, S2.1.1 specifically refers to: The average of the voltage data of each cell in the battery from at least two charging data points is subtracted from the average voltage of all cells in the battery; the offset of each cell voltage relative to the average voltage of all cells is obtained; the maximum value of the voltage offset of each cell in the battery is taken as the global maximum upward offset of the battery. The average of at least two discharge data points for each cell voltage in the battery is collected, and the average of all cell voltages in the battery is subtracted from the average of all cell voltages in the battery. The offset of each cell voltage relative to the average of all cell voltages is obtained, and the minimum cell voltage offset is taken as the global maximum downward offset of the battery.

4. The method for detecting battery short-circuit cells according to claim 3, characterized in that, The global upper offset anomaly threshold comprises a mild anomaly upper threshold g a1 and an extreme anomaly upper threshold e a1 The global lower offset anomaly threshold comprises a mild anomaly lower threshold e b1 and an extreme anomaly lower threshold e b1 ; S2.1.3 specifically refers to: The global offset data detection is performed using the box plot anomaly detection method, and the upper quartile of the global offset data is Q 31 The lower quartile of the global offset data is Q 11 The global quartile interval is ΔQ1; threshold setting is performed by controlling the mild offset coefficient w1 and the extreme offset coefficient w2: Setting a mild anomaly upper threshold g in the set of global maximum upper offsets a1 and an extreme anomaly upper threshold e a1 ; Set a mild anomaly threshold g in the global maximum lower offset set. b1 and the threshold e under extreme anomalies b1； Therefore, g a1 =Q 31 +w1×ΔQ1 e a1 =Q 31 +w2×ΔQ1 g b1 =Q 11 -w1×ΔQ1 e a1 =Q 11 -w2×ΔQ1。 5. A method for detecting battery short-circuit cells according to claim 1 or 4, characterized in that, The terminal anomaly detection includes the following steps: S2.2.1: Collect charging and discharging data for each battery each time, and calculate the end offset data for each battery, that is, calculate the maximum upper and lower offsets at the end of each charge / discharge cutoff time. S2.2.2: Merge the maximum upward offset of the end of all batteries to obtain a set of maximum upward offsets at the end; merge the maximum downward offset of the end of all batteries to obtain a set of maximum downward offsets at the end. S2.2.3: Set the threshold for end-of-battery offset data, including the threshold for end-of-battery upper offset and the threshold for end-of-battery lower offset; S2.2.4: Calibrate the threshold for abnormal downward offset at the end; S2.1.5: Calculate the maximum upward and downward offset of the end of a single cell in the battery to be tested; and compare the maximum upward offset of the end of a single cell with the global upward offset anomaly threshold. When the terminal offset reaches the high voltage global upward offset abnormal threshold, or is lower than the terminal downward offset abnormal threshold, the cell is determined to be terminal abnormal, and the specific location of the cell is recorded.

6. The method for detecting battery short-circuit cells according to claim 5, characterized in that, The maximum upward offset at the end is calculated by subtracting the second highest cell voltage from the highest cell voltage at the time of each battery charge cutoff. The maximum downward offset at the end is calculated by subtracting the second lowest cell voltage from the lowest cell voltage at the time of each battery discharge cutoff.

7. The method for detecting battery short-circuit cells according to claim 6, characterized in that, The threshold for abnormal offset at the end includes the mild abnormal offset threshold g. a2 and extreme anomalies above the threshold e a2 The end-point downward offset anomaly threshold includes the mild anomaly downward threshold e. b2 and the threshold e under extreme anomalies b2 ; S2.2.3 specifically refers to: Box plot anomaly detection is used to detect end-point offset data. Let Q be the upper quartile of the end-point offset data. 32 The lower quartile of the end offset data is Q. 12 The final quartile interval is ΔQ2; the threshold is set by controlling the mild offset coefficient w3 and the extreme offset coefficient w4: Set a mild anomaly threshold g in the set of maximum upper offsets at the end. a3 and extreme anomalies above the threshold e a3 ; Set a mild anomaly threshold e in the set of maximum downward offsets at the end. b3 and the threshold e under extreme anomalies b4 ; Therefore, g a2 =Q 32 +w3×ΔQ2 e a2 =Q 32 +w4×ΔQ2 g b2 =Q 12 -w3×ΔQ2 e a2 =Q 12 -w4×ΔQ2。 8. The method for detecting battery short-circuit cells according to claim 7, characterized in that, S2.2.4 specifically refers to: using a threshold g higher than the upper threshold of mild anomalies. a2 and below the mild abnormality threshold e b2 The data is used as the calibration target, and the data within the calibration target is recalibrated according to the method for setting the end abnormality threshold described in S2.2.3 to recalibrate the end abnormality threshold.

9. A battery short-circuit cell detection system, characterized in that, The device includes a memory and a processor. The memory includes a battery short cell detection method program. When the battery short cell detection method program is executed by the processor, it implements the method steps of any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a battery short cell detection method program, which, when executed by a processor, implements the steps of a battery short cell detection method as described in any one of claims 1 to 8.

Images