Battery cluster anomaly detection method, detection apparatus and battery management system
Patent Information
- Application Number
- PCT/CN2025/109754
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-03-26
- Filing Date
- 2025-07-22
- Publication Date
- 2026-10-01
Smart Images
Figure CN2025109754_01102026_PF_FP_ABST
Abstract
Description
A method, device and battery management system for detecting abnormalities in battery clusters Technical Field
[0001] This invention belongs to the technical field of battery pack anomaly detection, and particularly relates to a battery cluster anomaly detection method, detection device and battery management system. Background Technology
[0002] With the continuous advancement of new energy technologies and their increasing market share, the usage rate of new energy batteries is also rising. Therefore, the safety monitoring of new energy batteries has become an increasingly important research topic.
[0003] In existing technologies, although real-time monitoring of battery cells is set up and abnormal cells can be found in a timely manner, for some persistent outlier anomalies, sensor detachment and damage, and thermal design defects, the parameter characteristics exhibited in the early stage of the fault are only slightly biased and cannot be distinguished from normal battery cells. Therefore, existing technologies can only achieve real-time monitoring of the above-mentioned faults and deal with them as soon as they occur.
[0004] However, the consequences of the above-mentioned faults are very serious, and the impact on the battery cell cluster system is very large after the fault occurs. The required active repair time is long. Therefore, the existing real-time monitoring method can only provide a very short effective active protection time, which may cause major energy storage safety accidents. Summary of the Invention
[0005] To address the aforementioned problems, this invention proposes a battery cluster anomaly detection method, detection device, and battery management system. By recalculating the original fluctuation parameters before using existing methods to detect anomalies through position fluctuation parameters, the original fluctuation parameters can provide early warning and detection of faults such as persistent outlier anomalies, sensor detachment, damage, and thermal design defects, thereby preventing safety accidents and providing early warning time for handling unavoidable accidents.
[0006] The present invention is achieved through the following technical solutions:
[0007] In a first aspect, the present invention proposes a method for detecting anomalies in a battery cluster, wherein the battery cluster comprises M batteries, and each battery comprises N cells, including:
[0008] The first parameter of each battery cell is periodically acquired, and a first feature parameter set containing multiple first parameters is generated.
[0009] Sort the first parameters of the first feature parameter set, obtain the sorted data of each first parameter, and generate a second feature parameter set including the sorted data of different battery cells;
[0010] Within a preset continuous period, multiple sorted data of the same battery cell are acquired, and a third feature parameter set including the sorted data of the same battery cell is generated.
[0011] The first fluctuation parameter of the battery cell is calculated based on the third set of characteristic parameters; the status of each battery cell is monitored in real time using the first fluctuation parameter, and the first fluctuation parameter is normalized by standard score to obtain the second fluctuation parameter; the status of each battery cell is given an early warning based on the second fluctuation parameter.
[0012] By periodically acquiring the first parameter of each cell in the battery cluster, a first feature parameter set containing multiple first parameters is generated. These first parameters are then sorted to obtain sorted data for each first parameter, simultaneously generating a second feature parameter set including sorted data from different cells. Within a preset continuous period, multiple sorted data points for the same cell are acquired, generating a third feature parameter set including sorted data from the same cell. A first fluctuation parameter for the cell is calculated based on the third feature parameter set. A second fluctuation parameter is obtained through standard score normalization. The state of each cell is monitored and alerted based on both the first and second fluctuation parameters. By calculating the sorting parameter before calculating the fluctuation parameter, it is possible to detect and monitor some cells with minor deviations in performance in advance, thereby preventing major accidents.
[0013] In some implementations, a warning is issued for the status of each cell based on a second fluctuation parameter, including:
[0014] Obtain the preset first warning range and first window threshold;
[0015] If the second fluctuation parameter obtained in consecutive cycles is not within the first warning range, and the number of consecutive cycles is greater than the first window threshold, then a warning signal is output.
[0016] By using the first warning range and the first window threshold, it is possible to identify some persistent outlier anomalies, sensor detachment, damage, and thermal design defects, and provide early warnings for these problems. It can also solve situations where the fluctuation range is large and the characteristic parameters are always below the safety limit.
[0017] In some implementations, a second warning range and a third warning range are also preset;
[0018] If the second fluctuation parameter obtained in consecutive cycles is not within the range of the second warning, and the number of consecutive cycles is greater than the first window threshold, then the warning strategy outputs the second warning signal.
[0019] If the second fluctuation parameter obtained in consecutive cycles is not within the range of the third warning, and the number of consecutive cycles is greater than the first window threshold, then the warning strategy outputs the third warning signal.
[0020] By setting multiple warning ranges, multi-level warnings can be implemented in large-scale practical applications to capture individual units with better operational and maintenance value.
[0021] In some implementations, the first warning range includes an upper warning limit and a lower warning limit;
[0022] The upper and lower warning limits are opposites of each other.
[0023] In some implementations, the absolute values of the upper and lower warning limits are:
[0024] Among them, erf -1 It is the inverse function of the standard score error function.
[0025] This method allows for the calculation of upper and lower threshold limits using an error function, thus meeting the requirements for cell safety monitoring.
[0026] In some implementations, the state of each cell is monitored in real time using a first fluctuation parameter, including:
[0027] Obtain the preset first fluctuation parameter threshold and the second window threshold;
[0028] If the first fluctuation parameter obtained in consecutive cycles is greater than the first fluctuation parameter threshold, and the number of consecutive cycles is greater than the second window threshold, then an alarm signal is output.
[0029] In some implementations, each cell calculates a first fluctuation parameter using a third feature set, including:
[0030] The fluctuation probability P(x) of the third feature set is calculated and substituted into the following formula to obtain the first fluctuation parameter E. i :
[0031] Where a and b are constants, G is the number of different values appearing in the set of the third feature parameters, and i∈(0, M·N).
[0032] The first fluctuation parameter can also be calculated by calculating the fluctuation probability and obtaining the first fluctuation parameter through the formula for the change of information entropy, so as to obtain the operating status of the battery cell.
[0033] In some implementations, each cell calculates a first fluctuation parameter using the third feature set, including:
[0034] The first fluctuation parameter is obtained by calculating the position standard deviation of the data in the third feature set.
[0035] The first fluctuation parameter can be calculated using the third feature set. The first fluctuation parameter is the original position fluctuation parameter. The first fluctuation parameter can be obtained by calculating the position standard deviation of the parameter in the third feature set, thereby understanding the parameter status of the battery cell.
[0036] Secondly, the present invention provides a detection device, comprising:
[0037] The first cell data acquisition module is used to periodically acquire the first parameter of each cell and generate a first feature parameter set containing multiple first parameters.
[0038] The cell data sorting module is used to sort the first parameters of the first feature parameter set, obtain the sorted data of each first parameter, and generate a second feature parameter set including the sorted data of different cells.
[0039] The second cell data acquisition module is used to acquire multiple sorted data of the same cell within a preset continuous period and generate a third feature parameter set including the sorted data of the same cell.
[0040] The early warning monitoring module is used to calculate the first fluctuation parameter of the battery cell based on the third set of characteristic parameters; to monitor the status of each battery cell in real time using the first fluctuation parameter, and to normalize the first fluctuation parameter by standard score to obtain the second fluctuation parameter, and to issue an early warning for the status of each battery cell based on the second fluctuation parameter.
[0041] Thirdly, the present invention also proposes a battery management system, including: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;
[0042] The memory is used to store at least one executable instruction that causes the processor to perform operations such as the battery cluster anomaly detection method of the first aspect.
[0043] The beneficial effects of the battery cluster anomaly detection method, detection device, and battery management system of the present invention are as follows:
[0044] By sorting the first parameter data before calculating the fluctuation parameters, this method can provide early warning and detection of faults such as persistent outlier anomalies, sensor detachment and damage, and thermal design defects, in order to avoid energy storage accidents. This allows for earlier focus on and tracking of abnormal cells, and provides sufficient proactive response time before cell problems occur, thus preventing major safety incidents in energy storage.
[0045] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description
[0046] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:
[0047] Figure 1 is a flowchart of a battery cluster anomaly detection method according to the present invention;
[0048] Figure 2 shows the second fluctuation parameters of each cell calculated within a certain preset continuous period;
[0049] Figure 3 shows the second fluctuation parameters of each cell calculated within a preset continuous period using the battery cluster anomaly detection method of the present invention.
[0050] Figure 4 is a line graph of the fluctuation parameters calculated without sorting the data;
[0051] Figure 5 is a line graph of the fluctuation parameters calculated using the battery cluster anomaly detection method of the present invention;
[0052] Figure 6 is a schematic diagram of the detection device of the present invention;
[0053] Figure 7 is a schematic diagram of the battery management system of the present invention. Detailed Implementation
[0054] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.
[0055] Example 1:
[0056] As shown in Figure 1, this embodiment proposes a method for detecting anomalies in battery clusters. A battery cluster comprises M batteries, and each battery contains N cells, including:
[0057] Step 101: Periodically acquire the first parameter of each cell and generate a first feature parameter set containing multiple first parameters:
[0058] Specifically, the first parameters of N battery cells are acquired periodically. The period can be 50ms, and the period interval can be 30s. The first parameters can be safety parameters, attribute parameters, such as temperature, internal air pressure, battery pack, etc., preferably temperature and internal air pressure data. After acquiring the first parameters, a first feature parameter set is formed. For example, if there are 5 battery packs, and each battery pack contains 10 battery cells, then the first feature parameter set of the battery cells collected by one battery pack in a certain period can be represented as [25,26,35,27,25,26,24,29,28,29].
[0059] Step 102: Sort the first parameters of the first feature parameter set, obtain the sorted data of each first parameter, and generate a second feature parameter set including the sorted data of different battery cells;
[0060] Specifically, after sorting, the first parameter is replaced with the sequence number as the second parameter. The first feature parameter set is then either in ascending or descending order. The sorted data forms the second feature parameter set. Referring to the example in step 101, the second feature parameter set can be in descending order [2,4,10,6,2,4,1,8,7,8] or in ascending order [8,6,1,5,8,6,10,2,4,2].
[0061] Step 103: Within a preset continuous period, acquire multiple sorting data of the same battery cell and generate a third feature parameter set including the sorting data of the same battery cell.
[0062] Specifically, within a continuous cycle, steps 101 and 102 are performed to calculate the corresponding sorting data for each cell. These sorting data are then grouped into a third feature parameter set for the same cell group. For example, if cell A is sorted as 8 in the first cycle, 7 in the second cycle, and 5 in the third cycle, then the third feature parameter set for cell A in cycles 1-3 is [8, 7, 5].
[0063] The beneficial effect of steps 102-103 is that by adding sorting and calculating the fluctuation parameters of the sorted data, it is possible to detect when there are small differences in the characteristic parameters of individual cells in the battery cluster, but some cells have slight biases. In such cases, the mean, standard deviation and other parameters cannot fully capture the relevant characteristics. By sorting and calculating the fluctuation parameters of the sorted data first, such slight biases can be captured in advance, thereby realizing the early detection and monitoring of abnormal cells.
[0064] Step 104: Calculate the first fluctuation parameter of the battery cell based on the third set of feature parameters; monitor the status of each battery cell in real time using the first fluctuation parameter, and normalize the first fluctuation parameter by standard score to obtain the second fluctuation parameter; issue an early warning for the status of each battery cell based on the second fluctuation parameter.
[0065] Specifically, after calculating the first fluctuation parameter based on the third feature parameter set, the first fluctuation parameter is calculated for the second feature set of each of the N cells in the M battery packs. The first fluctuation parameter can be understood as the original position fluctuation parameter, which can be represented as E. i This is used to determine whether the battery cell is operating normally, and the second fluctuation parameter is obtained through standard score normalization. Standard score (Z-sore) normalization is performed. According to the standard score normalization formula, it can be expressed as follows:
[0066] Among them, E i ′ The second fluctuation parameter, also known as the positional fluctuation parameter, is Ei, which is the first fluctuation parameter mentioned above, i.e., the original positional fluctuation parameter. M represents the number of battery packs, and N represents the number of cells, where i ∈ (0, M·N). Based on the second fluctuation parameter, data analysis and early warning detection can be performed to proactively monitor and track changes in the first parameter of the cells, thus preventing energy storage safety accidents. The first fluctuation parameter is used to detect the cell status in real time, and its magnitude is used to determine the fluctuation of the first parameter of the cells, thereby detecting abnormal cells.
[0067] It should be noted that after selecting a window time for detection, a certain window time can be selected again for detection when the window time ends, and steps 101 to 104 can be repeated.
[0068] The system acquires a first parameter for each cell in a battery cluster periodically, generating a first set of characteristic parameters. These parameters are then sorted to generate a second set of characteristic parameters, including sorted data from different cells. Within a preset continuous period, multiple sorted data points for the same cell are acquired, generating a third set of characteristic parameters for the same cell's sorted data. A first fluctuation parameter for the cell is calculated based on the third set of characteristic parameters. A second fluctuation parameter is obtained through standard score normalization. The status of each cell is then monitored and alerted based on both the first and second fluctuation parameters. By calculating the sorting parameter before calculating the fluctuation parameter, cells with minor deviations can be detected and monitored in advance, thus preventing major accidents.
[0069] Furthermore, the advantages of the battery cluster anomaly detection method of the present invention can be illustrated by comparing Figures 2 and 3. Figures 2 and 3 show the second fluctuation parameters calculated within a preset continuous period. Figure 2 shows the second fluctuation parameters of each cell calculated without using the battery cluster anomaly detection method of the present invention, while Figure 3 shows the second fluctuation parameters of each cell calculated using the battery cluster anomaly detection method of the present invention. The horizontal axis represents the cell number, and the vertical axis represents the second fluctuation parameter value. As can be seen from Figure 2, the second fluctuation parameter of cell number 10 (anomaly) is not significantly different from that of other cells when calculated without using the battery cluster anomaly detection method of the present invention, but the second fluctuation parameter calculated using the battery cluster anomaly detection method of the present invention differs significantly from the calculated values of other cells.
[0070] Example 2:
[0071] This embodiment further explains and optimizes the battery cluster anomaly detection method proposed in Embodiment 1:
[0072] In some embodiments, a warning is issued for the state of each cell based on a second fluctuation parameter, including:
[0073] Obtain the preset first warning range and first window threshold;
[0074] If the second fluctuation parameter obtained in consecutive cycles is not within the first warning range, and the number of consecutive cycles is greater than the first window threshold, then a warning signal is output.
[0075] Specifically, a preset first warning range is used to determine whether the magnitude of the second fluctuation parameter falls within the first warning range, thereby enabling the monitoring of the second fluctuation parameter. Under normal circumstances, if the second fluctuation parameter exceeds the first warning range, it can be considered that the fluctuation of the cell is relatively large, which can be further understood as the possible existence of abnormal or unstable factors. In this case, the cell needs to be monitored in advance to focus on the first parameter of the cell. This method allows for earlier attention and tracking, preventing the occurrence of major safety incidents in energy storage.
[0076] By using the first warning range and the first window threshold, it is possible to identify some persistent outlier anomalies, sensor detachment, damage, and thermal design defects, and provide early warnings for these problems. It can also solve situations where the fluctuation range is large and the characteristic parameters are always below the safety limit.
[0077] In some embodiments, a second warning range and a third warning range are also preset;
[0078] If the second fluctuation parameter obtained in consecutive cycles is not within the range of the second warning, and the number of consecutive cycles is greater than the first window threshold, then the warning strategy outputs the second warning signal.
[0079] If the second fluctuation parameter obtained in consecutive cycles is not within the range of the third warning, and the number of consecutive cycles is greater than the first window threshold, then the warning strategy outputs the third warning signal.
[0080] Specifically, in large-scale practical applications, multiple warning ranges for fluctuation parameters can be set, and multi-level warnings can be implemented to capture individual cells with better operational and maintenance value. For example, when the number of clusters is 100, and the total number of individual cells is 24,000, the ranges can be set according to the actual situation as follows: First warning range [-3.2783, 3.2783], Second warning range [-3.68824, 3.68824], Third warning range [-4.09804, 4.09804]. When the second fluctuation parameter is outside the first warning range but within the second warning range, a yellow first warning signal can be issued; when the second fluctuation parameter is outside the second warning range but within the third warning range, an orange second warning signal can be issued; when the second fluctuation parameter is outside the third warning range, a red third warning signal can be issued. It is understood that this embodiment only adds two ranges. In some larger-scale practical applications, a fourth warning range, a fifth warning range, etc., can also be added. The number of warning ranges depends on the actual situation and is not limited to only the first, second, and third warning ranges.
[0081] By setting multiple warning ranges, multi-level warnings can be implemented in large-scale practical applications to capture individual units with better operational and maintenance value.
[0082] It should be noted that consecutive cycles can overlap. For example, the first calculation may use cycle 1-10, while the second calculation may use cycle 2-11, and so on. This allows for the calculation, judgment, early warning, and monitoring of multiple first and second fluctuation parameters.
[0083] In some embodiments, the first warning range includes a warning upper limit and a warning lower limit;
[0084] Normally, the upper and lower warning limits are opposites of each other. Optionally, the upper and lower warning limits can be set to different values. Furthermore, to reduce computational complexity, the second and third warning ranges can also be set with upper and lower warning limits, respectively. The upper and lower warning limits for the second warning range can be opposites of each other, and the upper and lower warning limits for the third warning range can also be opposites of each other.
[0085] In some embodiments, when the upper and lower warning limits are opposites, the absolute values of the upper and lower warning limits can be expressed as:
[0086] Among them, erf-1 N is the inverse function of the standard score error function. threshold If the upper and lower warning limits are the absolute values, then the first warning range can be expressed as (-N) threshold N threshold The second and third warning ranges can be within (-N) threshold N threshold Based on the actual situation of the tested battery cells, the warning range can be increased proportionally. For example, in the above example, the calculated first warning range is [-3.2783, 3.2783]. Based on the actual situation, it can be increased by 0.12%, so the second warning range is approximately [-3.68824, 3.68824] and the third warning range is approximately [-4.09804, 4.09804]. It should be noted that the selection of the second and third warning ranges is relatively flexible and can be determined according to different situations.
[0087] In some embodiments, the state of each cell is monitored in real time using the first fluctuation parameter, including:
[0088] Obtain the preset threshold values for the first fluctuation parameter and the second window threshold;
[0089] If the first fluctuation parameter obtained in consecutive cycles is greater than the first fluctuation parameter threshold, and the number of consecutive cycles is greater than the second window threshold, then an alarm signal is output.
[0090] The first fluctuation parameter is used for routine real-time monitoring of the battery cell clusters to promptly alert when problems occur. Typically, the real-time monitoring method involves setting a danger threshold, also known as the first fluctuation threshold. When the first fluctuation parameter exceeds the first fluctuation threshold, it is generally understood that the instantaneous fluctuation of the first parameter is large and exceeds the normal range. However, to avoid instantaneous misjudgment, the battery cell is generally confirmed as abnormal only after monitoring for several consecutive cycles and all values exceed the range, requiring replacement or repair.
[0091] In some preferred embodiments, each cell calculates a first fluctuation parameter using a third feature set, including:
[0092] The fluctuation probability P(x) of the third feature set is calculated and substituted into the following formula to obtain the first fluctuation parameter E. i :
[0093] Where a and b are constants, G is the number of different values appearing in the set of the third feature parameters, and i∈(0, M·N).
[0094] The first fluctuation parameter can also be calculated by calculating the fluctuation probability P(x) and obtaining the first fluctuation parameter through the formula for the change of information entropy, so as to obtain the operating state of the battery cell.
[0095] Specifically, the fluctuation probability is calculated using the third feature set obtained from the battery cell. The fluctuation probability P(x) is the probability of a certain value appearing among all possible values in the total possible values. For example, if there are 100 possible values from 1 to 100 in the third feature set, then G = 100, and the total number of fluctuations is 1000. If the value 3 appears 3 times, then the probability of fluctuation P(3) is 3 / 1000. And the first fluctuation parameter E can be calculated using the above formula (3). i The calculation is performed where i∈(0, M·N), meaning that each cell corresponds to a first fluctuation parameter, where a and b are constants, preferably a=0.5 and b=2.
[0096] In some embodiments, each cell calculates a first fluctuation parameter using the third feature set. The method may differ from the aforementioned calculation method using the first fluctuation parameter, except that it may not require calculation using the first fluctuation parameter. The calculation method includes:
[0097] The first fluctuation parameter is obtained by calculating the position standard deviation of the data in the third feature set.
[0098] The first fluctuation parameter can be calculated using the third feature set. The first fluctuation parameter is the original position fluctuation parameter. The first fluctuation parameter can be obtained by calculating the position standard deviation of the parameter in the third feature set, thereby understanding the parameter status of the battery cell.
[0099] Specifically, standard score normalization can be achieved by calculating the standard deviation of rank. The standard deviation of rank can be calculated by first calculating the mean of rank, for example, by calculating the mean of rank based on the obtained third feature set.
[0100] Where, x t Let be the t-th data point in the third feature set, and n be the total number of data points in the third feature set. The average value is calculated based on the rank. Calculate the variance σ 2 , that is:
[0101] By calculating the variance, the standard deviation of the ranking can be calculated as follows: That is, to obtain the first fluctuation parameter.
[0102] As shown in Figures 4 and 5, in practical applications, temperature is selected as the first parameter for monitoring data. Figures 4 and 5 are line graphs showing the fluctuation of the second fluctuation parameter calculated for a certain battery cell temperature over multiple preset continuous periods. The vertical axis represents the numerical change of the second fluctuation parameter, and the horizontal axis represents the change over time. Figure 4 shows the line graph of the second fluctuation parameter obtained without using the method proposed in Example 1, i.e., without sorting for calculation. Figure 5 shows the line graph calculated by the battery cluster anomaly detection method of the present invention.
[0103] Figure 4 shows a line graph of the second fluctuation parameter calculated using a method other than the one described in this invention, which is compared with Figure 5. Point A represents the time point when the second fluctuation parameter became abnormal, corresponding to July 15, 2024. Figure 5 shows the line graph of the second fluctuation parameter changing over time. Point B is the initial abnormal point (July 7, 2024, is still outside the set first warning range), meaning the second fluctuation parameter exceeded the set first warning range at the initial abnormal point. From this point onward, the focus can be on detecting and tracking anomalies in this cell. If subsequent consecutive deviations from the preset first warning range occur, a warning signal can be issued, allowing for proactive protective repair or replacement of the cell. Point B corresponds to July 5, 2024. It can be easily concluded that the second fluctuation parameter calculated using the battery cluster anomaly detection method of this invention can detect anomalies 10 days earlier than the cell temperature monitored using the second fluctuation parameter calculated by the method of this invention. In other words, the battery cluster anomaly detection method of the present invention can provide early warning and detection of some faults to avoid the occurrence of energy storage accidents.
[0104] Example 3:
[0105] As shown in Figure 6, this embodiment proposes a detection device 200, including:
[0106] The first cell data acquisition module 201 is used to periodically acquire the first parameters of each cell and generate a first feature parameter set containing multiple first parameters; the first cell data acquisition module 201 executes step 101 to acquire the first parameter set formed by the first parameters.
[0107] The cell data sorting module 202 is used to sort the first parameters of the first feature parameter set, obtain the sorted data of each first parameter, and generate a second feature parameter set including sorted data of different cells; by executing step 102 through the cell data sorting module 202, after obtaining the first parameter set, the sorted data of the first parameter set is obtained, and a second feature parameter set of the sorted data is generated as intermediate data for calculation.
[0108] The second cell data acquisition module 203 is used to acquire multiple sorted data of the same cell within a preset continuous period and generate a third feature parameter set including the sorted data of the same cell; the second cell data acquisition module 203 executes step 103 to form the third feature parameter set through the sorted data.
[0109] The early warning monitoring module 204 is used to calculate the first fluctuation parameter of the battery cell based on the third set of characteristic parameters; to monitor the status of each battery cell in real time using the first fluctuation parameter, and to normalize the first fluctuation parameter by standard score to obtain the second fluctuation parameter; and to issue an early warning for the status of each battery cell based on the second fluctuation parameter. The early warning monitoring module 204 executes step 104 to monitor and issue early warnings for the battery cell status by calculating the first and second fluctuation parameters.
[0110] Example 4:
[0111] As shown in Figure 7, this embodiment illustrates the structural schematic diagram of the battery management system of the present invention. The specific implementation of the battery management system is not limited by the specific embodiment of the present invention.
[0112] The battery management system may include: a processor 301, a communications interface 303, a memory 302, and a communications bus 305.
[0113] The processor 301, communication interface 303, and memory 302 communicate with each other via communication bus 305. Communication interface 303 is used to communicate with other network elements, such as clients or other servers. The processor 301 executes program 304, specifically performing the relevant steps described in the embodiment of the battery cluster anomaly detection method.
[0114] Specifically, program 304 may include program code, which includes computer-executable instructions.
[0115] The processor 301 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The battery management system includes one or more processors 301, which may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.
[0116] Memory 301 is used to store program 310. Memory 301 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.
[0117] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Furthermore, the embodiments of this invention are not directed to any particular programming language.
[0118] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. Similarly, for the sake of brevity and to aid in understanding one or more aspects of the invention, in the description of exemplary embodiments of the invention above, various features of the embodiments are sometimes grouped together in a single embodiment, figure, or description thereof. The claims, which follow the detailed description, are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.
[0119] Those skilled in the art will understand that the modules in the device of the embodiment can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiment can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.
[0120] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.
Claims
1. A battery cluster abnormality detection method characterized by comprising: Applied to a battery cluster having M batteries, each battery comprising N cells, including: The first parameter of each battery cell is periodically acquired, and a first feature parameter set containing multiple first parameters is generated. The first parameters in the first feature parameter set are sorted to obtain the sorted data of each first parameter, and a second feature parameter set including the sorted data of different battery cells is generated. Within a preset continuous period, multiple sorted data of the same battery cell are acquired, and a third feature parameter set including the sorted data of the same battery cell is generated. The first fluctuation parameter of the battery cell is calculated based on the third set of characteristic parameters; the status of each battery cell is monitored in real time using the first fluctuation parameter, and the first fluctuation parameter is normalized by standard score to obtain the second fluctuation parameter; the status of each battery cell is given an early warning based on the second fluctuation parameter.
2. The battery cluster abnormality detection method according to claim 1, characterized by, The method of providing early warnings about the status of each cell based on the second fluctuation parameter includes: Obtain the preset first warning range and first window threshold; If the second fluctuation parameter obtained in consecutive cycles is not within the first warning range, and the number of consecutive cycles is greater than the first window threshold, then the first warning signal is output.
3. The battery cluster anomaly detection method according to claim 2, characterized in that, It also has a second warning range and a third warning range preset; If the second fluctuation parameter obtained in consecutive periods is not within the range of the second warning, and the number of consecutive periods is greater than the first window threshold, then the second warning signal is output. If the second fluctuation parameter obtained in consecutive cycles is not within the range of the third warning, and the number of consecutive cycles is greater than the first window threshold, then the third warning signal is output.
4. The battery cluster abnormality detection method according to claim 2, characterized by, The first warning range includes an upper warning limit and a lower warning limit; The upper and lower warning limits are opposites of each other.
5. The battery cluster anomaly detection method according to claim 4, characterized in that, The absolute value of the pre-warning upper limit and the pre-warning lower limit is: where erf is the inverse of the standard fractional error function. -1 erf-1 6. The battery cluster abnormality detection method according to claim 1, characterized by, Real-time monitoring of the status of each cell using the first fluctuation parameter includes: Obtain the preset first fluctuation parameter threshold and the second window threshold; If the first fluctuation parameter obtained in consecutive periods is greater than the first fluctuation parameter threshold, and the number of consecutive periods is greater than the second window threshold, then an alarm signal is output.
7. The battery cluster abnormality detection method according to claim 1, characterized by, Each cell calculates a first fluctuation parameter using the third feature set, including: The third feature set is subjected to wave probability P(x) calculation, and is brought into the following formula to obtain the first wave parameter E i : Where a and b are constants, G is the number of different values appearing in the set of the third feature parameters, and i∈(0, M·N).
8. The battery cluster abnormality detection method according to claim 1, characterized by, Each battery cell calculates a first fluctuation parameter using the third feature set, and further includes: The first fluctuation parameter is obtained by calculating the position standard deviation of the data in the third feature set.
9. A detection device, characterized in that, include: The first cell data acquisition module is used to periodically acquire the first parameter of each cell and generate a first feature parameter set containing multiple first parameters. The cell data sorting module is used to sort the first parameters in the first feature parameter set, obtain the sorted data of each first parameter, and generate a second feature parameter set including the sorted data of different cells. The second cell data acquisition module is used to acquire multiple sorted data of the same cell within a preset continuous period and generate a third feature parameter set including the sorted data of the same cell. The early warning monitoring module is used to calculate the first fluctuation parameter of the battery cell based on the third set of feature parameters; to monitor the status of each battery cell in real time using the first fluctuation parameter, and to normalize the first fluctuation parameter by standard score to obtain the second fluctuation parameter, and to issue an early warning for the status of each battery cell based on the second fluctuation parameter.
10. A battery management system, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the battery cluster anomaly detection method as described in any one of claims 1-8.