Battery diagnostic device, battery diagnostic method, battery pack, and vehicle

Abstract

The battery diagnostic apparatus according to the present invention is used to diagnose a battery pack comprising a plurality of battery cells connected in series, and includes: a voltage sensing circuit configured to periodically generate a voltage signal indicating the cell voltage of each battery cell; and a control circuit configured to generate time-series data indicating the change of the cell voltage of each battery cell over time based on the voltage signal. The control unit is configured to: (i) determine a first average cell voltage and a second average cell voltage of each battery cell based on the time-series data, [wherein the first average cell voltage corresponds to a short-term moving average and the second average cell voltage corresponds to a long-term moving average]; and (ii) detect anomalies in the voltage of each battery cell based on the difference between the first average cell voltage and the second average cell voltage.

Description

Technical Field

[0001] This disclosure relates to a battery diagnostic device, a battery diagnostic method, a battery pack including the battery diagnostic device, and a vehicle including the battery pack, all of which are technologies for diagnosing abnormal voltages in batteries.

[0002] This application claims the benefit of Korean Patent Application No. 10-2020-0163366, filed with the Korean Intellectual Property Office on November 27, 2020, the disclosure of which is incorporated herein by reference in its entirety. Background Technology

[0003] Recently, demand for portable electronic products such as laptops, cameras, and mobile phones has grown rapidly, and with the widespread development of electric vehicles, energy storage batteries, robots, and satellites, much research is being conducted on high-performance rechargeable batteries.

[0004] Currently, commercially available batteries include nickel-cadmium batteries, nickel-metal hydride batteries, nickel-zinc batteries, and lithium batteries. Among them, lithium batteries have almost no memory effect and are gaining increasing attention than nickel-based batteries due to their advantages of being able to be charged at any time, having a very low self-discharge rate, and high energy density.

[0005] Recently, with the widespread adoption of applications requiring high voltage (such as energy storage systems and electric vehicles), there is an increasing need for accurate diagnosis of abnormal voltages in each of the multiple battery cells connected in series in a battery pack.

[0006] Abnormal voltage conditions of a battery cell refer to fault conditions caused by abnormal voltage drops and / or rises due to internal short circuits, external short circuits, defects in the voltage sensing line, or poor connection with the charging / discharging line.

[0007] An attempt has been made to diagnose abnormal voltages in each battery cell by comparing the voltage across each cell at a specific time (i.e., cell voltage) with the average cell voltage of multiple cells at the same time. However, the cell voltage of each battery cell depends on the temperature, current, and / or state of health (SOH) of the corresponding cell, making it difficult to accurately diagnose abnormal voltages in each cell by simply comparing the cell voltages of multiple cells measured at a specific time. For example, when there is a large difference in temperature or SOH between a cell without abnormal voltage and the remaining cells, the difference between the cell voltage of that cell and the average cell voltage may also be large.

[0008] To address this issue, in addition to the cell voltage of each battery cell, additional parameters for each battery cell, such as the cell voltage, charging / discharging current, temperature, and / or state of charge (SOC), can also be used for abnormal voltage diagnosis of each battery cell. However, diagnostic methods using additional parameters involve the process of detecting and comparing each parameter, thus requiring more complexity and time than diagnostic methods using cell voltage as the sole parameter. Summary of the Invention

[0009] Technical issues

[0010] This disclosure aims to solve the above-mentioned problems, and therefore relates to providing a battery diagnostic apparatus, battery diagnostic method, battery pack, and vehicle for efficient and accurate abnormal voltage diagnosis of battery cells, wherein for each of at least one moving window having a given time length, a moving average of the cell voltage of each of a plurality of battery cells is determined in each unit time, and abnormal voltage diagnosis of each battery cell is performed based on each moving average of each battery cell.

[0011] These and other objects and advantages of this disclosure will be understood from the following description and will be apparent from embodiments of this disclosure. Furthermore, it will be readily understood that the objects and advantages of this disclosure can be achieved by the means set forth in the appended claims and combinations thereof.

[0012] Technical solution

[0013] The battery diagnostic device for achieving the above objectives is a battery diagnostic device for a cell group comprising multiple battery cells connected in series, and may include: a voltage sensing circuit configured to periodically generate a voltage signal indicating the cell voltage of each battery cell; and a control circuit configured to generate time-series data indicating the change of the cell voltage of each battery cell over time based on the voltage signal.

[0014] Preferably, the control circuit can be configured to (i) determine a first average cell voltage and a second average cell voltage for each battery cell based on time series data, wherein the first average cell voltage is a short-term moving average and the second average cell voltage is a long-term moving average, and (ii) detect abnormal voltages of each battery cell based on the difference between the first average cell voltage and the second average cell voltage.

[0015] In one aspect, the control circuit can be configured to: determine for each battery cell a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage; determine for each battery cell a cell diagnostic deviation corresponding to the deviation between the average of the short-term / long-term average differences of all battery cells and the short-term / long-term average difference of that battery cell; and detect battery cells that meet the requirement that the cell diagnostic deviation exceeds a diagnostic threshold as abnormal voltage cells.

[0016] Preferably, the control circuit can be configured to generate time-series data of cell diagnostic deviation for each battery cell, and detect abnormal voltage of the battery cell based on the time period during which the cell diagnostic deviation exceeds a diagnostic threshold or the number of data points of cell diagnostic deviation exceeding the diagnostic threshold.

[0017] On the other hand, the control circuit can be configured to: determine a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage for each battery cell; determine a cell diagnostic deviation for each battery cell by calculating the deviation between the average of the short-term / long-term average differences of all battery cells and the short-term / long-term average difference of that battery cell; determine a statistical adaptive threshold that depends on the standard deviation of the cell diagnostic deviation for all battery cells; generate time-series data of filter diagnostic values ​​by filtering the time-series data of the cell diagnostic deviation for each battery cell based on the statistical adaptive threshold; and detect abnormal voltages of battery cells based on the time periods during which the filter diagnostic values ​​exceed the diagnostic threshold or the number of filter diagnostic values ​​exceeding the diagnostic threshold.

[0018] In another aspect, the control circuit can be configured to determine a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage for each battery cell; determine a normalized value of the short-term / long-term average difference as a normalized cell diagnostic deviation for each battery cell; determine a statistical adaptive threshold that depends on the standard deviation of the normalized cell diagnostic deviation for all battery cells; generate time-series data of filtered diagnostic values ​​by filtering the time-series data of the normalized cell diagnostic deviation for each battery cell based on the statistical adaptive threshold; and detect abnormal voltages of battery cells based on the time periods during which the filtered diagnostic values ​​exceed the diagnostic threshold or the number of filtered diagnostic values ​​exceeding the diagnostic threshold.

[0019] Preferably, the control circuit can normalize the short-term / long-term average difference for each battery cell by dividing the short-term / long-term average difference by the average of the short-term / long-term average differences of all battery cells.

[0020] Alternatively, the control circuit can normalize the short-term / long-term average difference for each battery cell by performing a logarithmic calculation on the short-term / long-term average difference.

[0021] On the other hand, the control circuit can be configured to generate time-series data indicating the change of the cell voltage of each battery cell over time using a voltage corresponding to the voltage difference between the average cell voltage of all battery cells measured at each unit time and the cell voltage of each battery cell.

[0022] In another aspect, the control circuit can be configured to: determine a short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage for each battery cell; determine a normalized value of the short-term / long-term average difference as a normalized cell diagnostic bias for each battery cell; and generate time-series data of the normalized cell diagnostic bias for each battery cell by recursively repeating (i) to (iv) at least once.

[0023] (i) For the time series data of normalized cell diagnostic bias for each battery cell, determine a first moving average and a second moving average, wherein the first moving average is a short-term moving average and the second moving average is a long-term moving average; (ii) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first moving average and the second moving average; (iii) For each battery cell, determine the normalized value of the short-term / long-term average difference as the normalized cell diagnostic bias; and (iv) For each battery cell, generate time series data of normalized cell diagnostic bias.

[0024] A statistical adaptive threshold is determined, which depends on the standard deviation of the normalized cell diagnostic bias for all battery cells; time series data of the normalized cell diagnostic bias for each battery cell are filtered based on the statistical adaptive threshold to generate time series data of filter diagnostic values; and abnormal voltage of battery cells is detected based on the time period during which the filter diagnostic values ​​exceed the diagnostic threshold or the number of filter diagnostic values ​​that exceed the diagnostic threshold.

[0025] The battery diagnostic method according to this disclosure for achieving the above objectives is a battery diagnostic method for a cell group comprising a plurality of battery cells connected in series, and may include: (a) periodically generating time-series data of voltage signal variation over time indicating the cell voltage of each battery cell; (b) determining a first average cell voltage and a second average cell voltage for each battery cell based on the time-series data, wherein the first average cell voltage is a short-term moving average and the second average cell voltage is a long-term moving average; and (c) detecting abnormal voltage of each battery cell based on the difference between the first average cell voltage and the second average cell voltage.

[0026] In one aspect, step (c) may include: (c1) determining for each battery cell a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage; (c2) determining for each battery cell a cell diagnostic deviation corresponding to the deviation between the average of the short-term / long-term average differences of all battery cells and the short-term / long-term average difference of that battery cell; and (c3) detecting battery cells that meet the requirement that the cell diagnostic deviation exceeds a diagnostic threshold as abnormal voltage cells.

[0027] Preferably, step (c) may include (c1) generating time-series data of cell diagnostic deviation for each battery cell; and (c2) detecting abnormal voltage of the battery cell based on the time period during which the cell diagnostic deviation exceeds the diagnostic threshold or the number of data points of cell diagnostic deviation exceeding the diagnostic threshold.

[0028] On the other hand, step (c) may include: (c1) determining a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage for each battery cell; (c2) determining a cell diagnostic deviation for each battery cell by calculating the deviation between the average of the short-term / long-term average differences of all battery cells and the short-term / long-term average difference of that battery cell; (c3) determining a statistical adaptive threshold that depends on the standard deviation of the cell diagnostic deviation for all battery cells; (c4) generating time-series data of filter diagnostic values ​​by filtering the time-series data of the cell diagnostic deviation for each battery cell based on the statistical adaptive threshold; and (c5) detecting abnormal voltages of battery cells based on the time periods during which the filter diagnostic values ​​exceed the diagnostic threshold or the number of filter diagnostic values ​​that exceed the diagnostic threshold.

[0029] In another aspect, step (c) may include: (c1) determining a short-term / long-term average difference corresponding to the difference between a first average cell voltage and a second average cell voltage for each battery cell; (c2) determining a normalized value of the short-term / long-term average difference as a normalized cell diagnostic bias; (c3) determining a statistically adaptive threshold that depends on the standard deviation of the normalized cell diagnostic bias for all battery cells; (c4) generating time-series data of filtered diagnostic values ​​by filtering the time-series data of the normalized cell diagnostic bias for each battery cell based on the statistically adaptive threshold; and (c5) detecting abnormal voltages of battery cells based on the time period during which the filtered diagnostic values ​​exceed the diagnostic threshold or the number of data points of filtered diagnostic values ​​exceeding the diagnostic threshold.

[0030] Preferably, step (c2) can be the following steps: normalizing the short-term / long-term average difference for each battery cell by dividing the short-term / long-term average difference by the average of the short-term / long-term average differences of all battery cells.

[0031] Alternatively, step (c2) may be the following steps: for each battery cell, normalize the short-term / long-term average difference by performing a logarithmic calculation on the short-term / long-term average difference.

[0032] On the other hand, step (a) may be the following steps: using the voltage difference between the average cell voltage of all battery cells measured at each unit time and the cell voltage of each battery cell, to generate time series data indicating the change of cell voltage of each battery cell over time.

[0033] In another aspect, step (c) may include: (c1) determining, for each battery cell, a short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage; (c2) determining, for each battery cell, a normalized value of the short-term / long-term average difference as a normalized cell diagnostic bias; (c3) generating time-series data of the normalized cell diagnostic bias for each battery cell; and (c4) generating time-series data of the normalized cell diagnostic bias for each battery cell by recursively repeating (i) to (iv) at least once:

[0034] (i) For the time series data of normalized cell diagnostic bias for each battery cell, determine a first moving average and a second moving average, wherein the first moving average is a short-term moving average and the second moving average is a long-term moving average; (ii) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first moving average and the second moving average; (iii) For each battery cell, determine the normalized value of the short-term / long-term average difference as the normalized cell diagnostic bias; and (iv) For each battery cell, generate time series data of normalized cell diagnostic bias.

[0035] (c5) Determine a statistical adaptive threshold that depends on the standard deviation of the normalized cell diagnostic bias for all battery cells; (c6) Filter the time series data of the normalized cell diagnostic bias for each battery cell based on the statistical adaptive threshold to generate time series data of filter diagnostic values; and (c7) Detect abnormal voltage of battery cells based on the time period during which the filter diagnostic values ​​exceed the diagnostic threshold or the number of filter diagnostic values ​​that exceed the diagnostic threshold.

[0036] The aforementioned technical objectives can also be achieved through a battery pack including a battery diagnostic device and a vehicle including the battery pack.

[0037] Technical effect

[0038] According to one aspect of this disclosure, an effective and accurate diagnosis of abnormal voltage for each battery cell can be achieved by determining the moving average of two cell voltages for each battery cell at two different time lengths at each unit time, and by performing abnormal voltage diagnosis for each battery cell based on the difference between the two moving averages for each of the plurality of battery cells.

[0039] According to another aspect of this disclosure, by applying advanced techniques such as normalization and / or statistical adaptive thresholding to analyze the difference in the changing trends of the two moving averages of each battery cell, accurate diagnosis of abnormal voltage of each battery cell can be achieved.

[0040] According to another aspect of this disclosure, the time range of abnormal voltage occurrence and / or abnormal voltage detection count for each battery cell can be accurately detected by analyzing time series data of filter diagnostic values ​​determined based on statistical adaptive thresholds.

[0041] The effects of this disclosure are not limited to those described above, and those skilled in the art will clearly understand from the appended claims the effects not mentioned herein. Attached Figure Description

[0042] The accompanying drawings illustrate preferred embodiments of the present disclosure and, together with the detailed description of the present disclosure below, are intended to provide a further understanding of the technical aspects of the present disclosure; therefore, the present disclosure should not be construed as being limited to the drawings.

[0043] Figure 1 A diagram illustrating an electric vehicle according to an embodiment of the present disclosure is provided.

[0044] Figures 2a to 2h In describing according to instructions Figure 1 The chart shown is a reference used in the process of diagnosing abnormal voltages of each of the multiple battery cells by analyzing the time series data of the cell voltage change over time.

[0045] Figure 3 This is an exemplary flowchart illustrating a battery diagnostic method according to a first embodiment of the present disclosure.

[0046] Figure 4 This is an exemplary flowchart of a battery diagnostic method according to a second embodiment of the present disclosure.

[0047] Figure 5 This is an exemplary flowchart of a battery diagnostic method according to a third embodiment of the present disclosure.

[0048] Figure 6 This is an exemplary flowchart of a battery diagnostic method according to a fourth embodiment of the present disclosure.

[0049] Figure 7 This is an exemplary flowchart of a battery diagnostic method according to a fifth embodiment of the present disclosure. Detailed Implementation

[0050] Preferred embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Before the description, it should be understood that the terms or words used in the specification and appended claims should not be construed as limited to their general and dictionary meanings, but rather as being interpreted based on their meanings and concepts corresponding to the technical aspects of the present disclosure, on the basis of allowing the inventors to appropriately define the terms for the best interpretation.

[0051] Therefore, the embodiments described herein and the illustrations shown in the accompanying drawings are merely the most preferred embodiments of this disclosure and are not intended to fully describe the technical aspects of this disclosure. It should be understood that various other equivalents and modifications have been made to them at the time of filing this application.

[0052] Ordinal terms such as “first” and “second” are used to distinguish one element from others among various elements, but are not intended to limit elements by the terms.

[0053] Unless the context clearly indicates otherwise, the term "comprising," as used herein, specifies the presence of the mentioned element, but does not exclude the presence or addition of one or more other elements. Additionally, as used herein, the term "control unit" refers to a processing element having at least one function or operation, which can be implemented by hardware and software, alone or in combination.

[0054] Furthermore, throughout the specification, it will be further understood that when an element is referred to as being “connected to” another element, it can be directly connected to the other element or there can be an intermediary element.

[0055] Figure 1 A diagram illustrating an electric vehicle according to an embodiment of the present disclosure is provided.

[0056] Reference Figure 1 The electric vehicle 1 includes a battery pack 2, an inverter 3, an electric motor 4, and a vehicle controller 5.

[0057] Battery pack 2 includes cell group CG, switch 6 and battery management system 100.

[0058] The unit group CG can be connected to the inverter 3 via a pair of power terminals located in the battery pack 2. The unit group CG includes multiple battery units BC1 to BC2 connected in series. N (N is a natural number of 2 or greater). Each battery cell BC i It is not limited to a specific type and can include any rechargeable battery cell, such as a lithium-ion battery cell. i is the index for cell identification. i is a natural number between 1 and N.

[0059] Switch 6 is connected in series to the cell group CG. Switch 6 is mounted on the current path for charging / discharging the cell group CG. Switch 6 controls the switching between an on and off state in response to a switching signal from the battery management system 100. Switch 6 can be a mechanical relay that is turned on / off by the electromagnetic force of a coil, or a semiconductor switch such as a metal-oxide-semiconductor field-effect transistor (MOSFET).

[0060] Inverter 3 is configured to convert direct current (DC) power from cell bank CG into alternating current (AC) power in response to commands from battery management system 100. Motor 4 may be, for example, a three-phase AC motor. Motor 4 operates using AC power from inverter 3.

[0061] The battery management system 100 is configured to perform overall control related to the charging / discharging of the cell group CG.

[0062] The battery management system 100 includes a battery diagnostic device 200. The battery management system 100 may also include at least one of a current sensor 310, a temperature sensor 320, and an interface unit 330.

[0063] Battery diagnostic device 200 is configured for multiple battery cells BC1 to BC2. N Abnormal voltage diagnosis for each of the components. The battery diagnostic device 200 includes a voltage sensing circuit 210 and a control circuit 220.

[0064] Voltage sensing circuit 210 is connected to multiple battery cells BC1 to BC2 via multiple voltage sensing lines. N Each of the cells has a positive and a negative terminal. The voltage sensing circuit 210 is configured to measure the cell voltage across each cell BC and generate a voltage signal indicating the measured cell voltage.

[0065] Current sensor 310 is connected in series to cell group CG via a current path. Current sensor 310 is configured to detect the battery current flowing through cell group CG and generate a current signal indicating the detected battery current.

[0066] Temperature sensor 320 is configured to detect the temperature of unit group CG and generate a temperature signal indicating the detected temperature.

[0067] The control circuit 220 can be implemented in hardware using at least one of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a microprocessor, and an electrical unit for performing other functions.

[0068] The control circuit 220 may have a storage unit. The storage unit may include at least one type of storage medium selected from flash memory, hard disk, solid-state drive (SSD), silicon disk drive (SDD), micro multimedia card, random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and programmable read-only memory (PROM). The storage unit may store data and programs required for calculations performed by the control circuit 220. The storage unit may also store data indicating the results of calculations performed by the control circuit 220. Specifically, the control circuit 220 may record in the storage unit at least one of a plurality of parameters calculated per unit time as described below.

[0069] Control circuit 220 can be operatively connected to voltage sensing circuit 210, temperature sensor 320, current sensor 310, interface unit 330, and / or switch 6. Control circuit 220 can collect sensing signals from voltage sensing circuit 210, current sensor 310, and temperature sensor 320. Sensing signals refer to voltage signals, current signals, and / or temperature signals detected synchronously.

[0070] Interface unit 330 may include communication circuitry configured to support wired or wireless communication between control circuitry 220 and vehicle controller 5 (e.g., electronic control unit (ECU)). Wired communication may be, for example, Controller Area Network (CAN) communication, while wireless communication may be, for example, Zigbee or Bluetooth communication. The communication protocol is not limited to a specific type and may include any communication protocol that supports wired / wireless communication between control circuitry 220 and vehicle controller 5.

[0071] Interface unit 330 can be connected to an output device (e.g., a display, a speaker) that provides information received from vehicle controller 5 and / or control circuitry 220 in a recognizable format. Vehicle controller 5 can control inverter 3 based on battery information (e.g., voltage, current, temperature, SOC) collected via communication with battery management system 100.

[0072] Figures 2a to 2h This is an example showing the instructions. Figure 1 The graph shows the process of diagnosing abnormal voltages for each of the multiple battery cells by analyzing the time-series data of the cell voltage changes over time.

[0073] Figure 2a Multiple battery cells BC1 to BC2 are shown. N The voltage curve for each of the cells is shown. The number of battery cells is 14. The control circuit 220 collects voltage signals from the voltage sensing circuit 210 every unit time and sets the voltage of each battery cell BC... i The voltage value of the unit voltage is recorded in the storage unit. The unit time can be an integer multiple of the voltage measurement cycle of the voltage sensing circuit 210.

[0074] Control circuit 220 can be based on each battery cell BC recorded in the storage unit i The cell voltage values ​​are used to generate cell voltage time series data indicating the cell voltage history over time for each cell. The number of cell voltage time series data points is increased by 1 each time a cell voltage is measured.

[0075] Figure 2a The multiple voltage curves shown correspond to multiple battery cells BC1 to BC2.N They are linked in a one-to-one relationship. Therefore, each voltage curve indicates the cell voltage change history of any battery cell BC associated with it.

[0076] Control circuit 220 can use one or two moving windows to determine multiple battery cells BC1 to BC2 in each unit of time. N The moving average of each of the moving windows. When using two moving windows, the duration of either moving window differs from the duration of the other moving window.

[0077] Here, the duration of each moving window is an integer multiple of a unit of time, and the end point of each moving window is the current time, while the start point of each moving window is a time point that is a given duration earlier than the current time.

[0078] In the following text, for ease of description, the moving window associated with the shorter time duration will be referred to as the first moving window, and the moving window associated with the longer time duration will be referred to as the second moving window.

[0079] The control circuit 220 can use the first moving window alone or both the first and second moving windows to execute the control for each battery cell BC. i Diagnosis of abnormal voltage.

[0080] The control circuit 220 can be based on the i-th battery cell BC collected in each unit time. i The cell voltage is compared with that of the i-th cell BC in each unit time. i The short-term and long-term trends of the unit voltage.

[0081] Control circuit 220 can determine the i-th battery cell BC in each unit time using either Equation 1 or Equation 2 via the first moving window. i The first average unit voltage (i.e., the moving average).

[0082] Equation 1 is the formula for calculating the moving average using the arithmetic mean method, while Equation 2 is the formula for calculating the moving average using the weighted average method.

[0083] <Formula 1>

[0084]

[0085] <Formula 2>

[0086]

[0087] In Equations 1 and 2, k is the time index indicating the current time, SMA i [k] is the i-th battery cell BC iThe first average cell voltage at the current time, S, is a value obtained by dividing the length of the first moving window by the unit time, while V... i [k] represents the i-th battery cell BC i The unit voltage at the current time. For example, when the unit time is 1 second and the duration of the first moving window is 10 seconds, S is 10. When x is a natural number k or smaller, V i [kx] and SMA i [kx] represents the i-th battery cell BC when the time index is kx. i The unit voltage and the first average unit voltage. For reference, the control circuit 220 can be configured to increment the time index by 1 in each unit of time.

[0088] Control circuit 220 can determine the i-th battery cell BC each unit time using the following formula 3 or 4 via the second moving window. i The second average unit voltage is used as a moving average.

[0089] Equation 3 is the formula for calculating the moving average using the arithmetic mean method, while Equation 4 is the formula for calculating the moving average using the weighted average method.

[0090] <Formula 3>

[0091]

[0092] <Formula 4>

[0093]

[0094] In equations 3 and 4, k is the time index representing the current time, LMA i [k] is the i-th battery cell BC i The second average unit voltage at the current time, L, is the value obtained by dividing the length of the second moving window by the unit time, while V... i [k] is the i-th battery cell BC i The unit voltage at the current time. For example, when the unit time is 1 second and the second moving window duration is 100 seconds, L is 100. When x is k or a smaller natural number, LMA i [kx] represents the second average cell voltage when the time index is kx.

[0095] In this implementation, the control circuit 220 can be input with respect to the reference cell voltage of the current time unit group CG and the battery cell BC. i The difference between the unit voltages is used as V in equations 1 to 4. i [k], instead of each battery cell BC at the current time. i The unit voltage.

[0096] The reference cell voltage for the current time cell group CG is from multiple battery cells BC1 to BC2. N The average value of multiple cell voltages at the current time. In a variant, the average value of multiple cell voltages can be replaced by its median.

[0097] Specifically, control circuit 220 can convert VD in equation 5 below. i [k] is set as V in equations 1 to 4. i [k].

[0098] <Formula 5>

[0099] VD i [k]=V av [k]-V i [k]

[0100] In Equation 5, V av [k] is the reference cell voltage of the current time cell group CG, and is the average value of multiple cell voltages.

[0101] When the duration of the first moving window is less than the duration of the second moving window, the first average cell voltage can be called the "short-term moving average" of the cell voltage, while the second average cell voltage can be called the "long-term moving average" of the cell voltage.

[0102] Figure 2b It shows that according to Figure 2a The i-th battery cell BC is determined by the multiple voltage curves shown. i The short-term moving average and long-term moving average of the unit voltage. Figure 2b In the diagram, the horizontal axis represents time, while the vertical axis represents the short-term moving average and long-term moving average of the unit voltage.

[0103] Reference Figure 2b In the middle, the dashed lines represent multiple moving average lines S. i With multiple battery cells BC1 to BC N They are associated in a one-to-one relationship, and represent the first average cell voltage SMA of each battery cell BC. i [k] Historical changes over time. Additionally, the solid lines represent multiple moving average lines L. i With multiple battery cells BC1 to BC N They are associated in a one-to-one relationship, and represent the second average cell voltage LMA of each battery cell BC. i The history of [k] over time.

[0104] The dashed and solid curves were obtained using Equations 2 and 4, respectively. Additionally, VD in Equation 5...i [k] is used as V in equations 2 and 4. i [k], and V av [k] is set to the average value of multiple cell voltages. The first moving window has a duration of 10 seconds, while the second moving window has a duration of 100 seconds.

[0105] Figure 2c It shows the relationship with Figure 2b The first average cell voltage (SMA) of each battery cell shown i [k] and second average unit voltage LMA i The difference between [k] corresponds to the change over time in the short-term / long-term average difference (absolute value). Figure 2c In the diagram, the horizontal axis represents time, while the vertical axis represents each battery cell BC. i The short-term / long-term average difference.

[0106] Each battery cell BC i The short-term / long-term average difference is per unit time per cell BC i First average cell voltage SMA i Second average unit voltage LMA i The difference between them. For example, the i-th battery cell BC i The short-term / long-term average difference can be equal to the difference from the SMA. i [k] and LMA i The value of one (e.g., the larger one) minus the value of another (e.g., the smaller one) in [k].

[0107] The i-th battery cell BC i The short-term / long-term average difference depends on the i-th cell BC. i The short-term and long-term variation history of the unit voltage.

[0108] The i-th battery cell BC i Temperature or SOH has a short-term and long-term stable effect on the i-th cell BC i The cell voltage. Therefore, in the i-th cell BC i In the absence of abnormal voltage, the i-th battery cell BC i There was no significant difference between the short-term / long-term average difference of the [cell name] and the short-term / long-term average difference of the other cell cells.

[0109] Conversely, due to internal short circuits and / or external short circuits in the i-th battery cell BC i The sudden appearance of abnormal voltage affects the first average cell voltage SMA i The influence of [k] on the second average unit voltage LMA i The influence of [k] is greater. As a result, the i-th battery cell BCi The short-term / long-term average difference of the battery cells deviates significantly from the short-term / long-term average difference of the other battery cells without abnormal voltage.

[0110] Control circuit 220 can determine the time for each battery cell BC per unit time. i Short-term / long-term average difference | SMA i [k]-LMA i [k]|. Additionally, control circuit 220 can determine the average of the short-term / long-term mean difference |SMA. i [k]-LMA i [k]|. In the following text, the average value is expressed as |SMA|. i [k]-LMA i [k]| av Additionally, control circuit 220 can control the short-term / long-term average difference |SMA i [k]-LMA i The deviation of [k]| relative to the average of the short-term / long-term average differences|SMAi[k]-LMAi[k]|av is determined as the unit diagnostic deviation D. diag,i [k]. Additionally, the control circuit 220 can be based on the unit diagnostic deviation D. diag,i [k] is used to execute BC for each battery cell. i Diagnosis of abnormal voltage.

[0111] In the implementation method, when the i-th battery cell BC i Unit diagnostic deviation D diag,i When [k] exceeds a preset diagnostic threshold (e.g., 0.015), the control circuit 220 can diagnose that in the corresponding i-th battery cell BC i An abnormal voltage exists.

[0112] Preferably, the control circuit 220 can use a normalized reference value for abnormal voltage diagnosis for each battery cell BC. i Short-term / long-term average difference | SMA i [k]-LMA i [k]| Normalization is performed. Preferably, the normalization reference value is the average of the short-term / long-term mean differences |SMA i [k]-LMA i [k]| av .

[0113] Specifically, control circuit 220 can control the first battery cell to the Nth battery cell (BC). i To BC N The average of the short-term / long-term mean differences | SMA i [k]-LMA i [k]|av Set to a normalized reference value. Additionally, control circuit 220 controls each battery cell BC... i Divide the short-term / long-term mean difference |SMAi[k]-LMAi[k]| by the normalized reference value to approximate the short-term / long-term mean difference |SMAi[k]-LMAi[k]|. i [k]-LMA i [k]|Normalize.

[0114] Equation 6 below is for each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i [k]| The formula for normalization. In the implementation, the product of Equation 6 can be referred to as the normalized unit diagnostic bias D. * diag,i [k].

[0115] <Formula 6>

[0116] D * diag,i [k]=(|SMA i [k]-LMA i [k]|)÷(|SMA i [k]-LMA i [k]| av )

[0117] In Equation 6, |SMA i [k]-LMA i [k]| is the i-th battery cell BC i The difference between the short-term and long-term averages at the current time, |SMA i [k]-LMA i [k]| av It is the average of the short-term / long-term average differences of all battery cells (normalized reference value), while D * diag,i [k] represents the i-th battery cell BC i Normalized cell diagnostic bias at the current time. The symbol "*" indicates that the parameter has been normalized.

[0118] Each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i [k] can be normalized using the logarithm of Equation 7. In the implementation, the product of Equation 7 can also be called the normalized unit diagnostic bias D. * diag,i [k].

[0119] <Formula 7>

[0120] D * diag,i [k] = Log|SMA i [k]-LMA i [k]|

[0121] Figure 2d Each battery cell BC is shown i Normalized unit diagnostic bias D * diag,i [k] varies with time. Equation 6 is used to calculate the unit diagnostic deviation D. * diag,i [k]. In Figure 2d In the diagram, the horizontal axis represents time, while the vertical axis represents the time per battery cell (BC). i Unit diagnostic deviation D * diag,i [k].

[0122] Reference Figure 2d It can be seen that, through each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i Based on the normalized average value of [k]|, the value of each battery cell BC is amplified. i The variation in the short-term / long-term average difference allows for more accurate diagnosis of abnormal voltages in battery cells.

[0123] Preferably, the control circuit 220 can control each battery cell BC i Normalized unit diagnostic bias D * diag,i [k] and statistical adaptive threshold D threshold [k] compares to implement each battery cell BC i Diagnosis of abnormal voltage.

[0124] Preferably, the control circuit 220 can set the statistical adaptive threshold D using Equation 8 at each unit time. threshold [k].

[0125] <Formula 8>

[0126] D threshold [k] = β*Sigma(D) * diag,i [k])

[0127] In Equation 8, Sigma is the normalized cell diagnostic deviation D of all battery cells BC at time index k. * diag,iThe standard deviation of [k] is a function. Additionally, β is an experimentally determined constant. β is a factor determining diagnostic sensitivity. When this disclosure is applied to a group of cells including battery cells in which abnormal voltages actually occur, β can be appropriately determined by trial and error to detect the corresponding battery cell as an abnormal voltage cell. In the example, β can be set to at least 5, or at least 6, or at least 7, or at least 8, or at least 9. The D generated by Equation 8... threshold [k] consists of multiple [k] values ​​and is used to construct time series data.

[0128] Meanwhile, the normalized cell diagnostic bias D of battery cells in abnormal voltage conditions * diag,i [k] is greater than the normalized cell diagnostic bias of a normal battery cell. Therefore, to improve the accuracy and reliability of the diagnosis, Sigma(D) is calculated at time index k. * diag,i When [k]), we want to exclude max(D) corresponding to the maximum value. * diag,i [k]). Here, max is a function that returns the maximum value for multiple input parameters, and the input parameters are the normalized cell diagnostic bias D of all battery cells. * diag,i [k].

[0129] exist Figure 2d In the middle, D represents the statistical adaptive threshold. threshold [k] The time series data that changes over time corresponds to the darkest color among all the distributions (profiles).

[0130] The statistical adaptive threshold D at time index k was determined. threshold [k] After that, the control circuit 220 can use the following formula 9 to control each battery cell BC i Normalized unit diagnostic bias D * diag,i [k] Perform filtering to determine the filter diagnostic value D. filter,i [k].

[0131] It can be done for each battery cell BC i Filter diagnostic value D filter,i [k] is assigned two values. That is, in the unit diagnostic deviation D * diag,i [k] is greater than the statistical adaptive threshold D threshold In the case of [k], the unit diagnostic deviation D * diag,i [k] and statistical adaptive threshold D Threshold The difference between [k] is assigned to the filter diagnostic value D. filter,i[k]. Conversely, in the unit diagnostic deviation D * diag,i [k] is equal to or less than the statistical adaptive threshold D threshold In the case of [k], assign 0 to the filter diagnostic value D. filter,i [k].

[0132] <Form 9>

[0133] D filter,i [k]=D * diag,i [k]-D threshold [k](if D) * diag,i [k]>D threshold [k])

[0134] D filter,i [k] = 0 (if D) * diag,i [k]≤D threshold [k])

[0135] Figure 2e This demonstrates how to diagnose the deviation D of the cell at time index k. * diag,i [k] is the filter diagnostic value D obtained by filtering. filter,i A graph of the time series data of [k].

[0136] Reference Figure 2e The irregular pattern indicates the filter diagnostic value D for a specific battery cell. filter,i [k] has a positive value at approximately 3000 seconds. For reference, a specific battery cell with an irregular pattern is one that has... Figure 2d A in the figure indicates the cell of the time series data.

[0137] In the example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is a time step greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement of accumulating time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0138] Preferably, the control circuit 220 can accumulate and successively satisfy the filter diagnostic value D filter,i [k] is the time step greater than the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0139] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is the number of data included in a time step that is greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement that the cumulative number of data is greater than a preset reference count are diagnosed as abnormal voltage cells.

[0140] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the number of data points contained in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0141] Meanwhile, the control circuit 220 can be used Figure 2d Each battery cell BC shown i Normalized unit diagnostic bias D * diag,i [k] Replace V in equations 1 to 5 i [k]. Additionally, the control circuit 220 can recursively perform the following at time index k: calculate the unit diagnostic deviation D. * diag,i [k] short-term / long-term mean difference | SMA i [k]-LMA i [k]|;Calculate the diagnostic deviation D of the calculation unit * diag,i [k] short-term / long-term mean difference | SMA i [k]-LMA i The average value of [k]; calculate the difference between the short-term / long-term average |SMA|. i [k]-LMA i [k]|The unit diagnostic bias D corresponding to the difference compared to the mean. diag,i [k]; Use Equation 6 to calculate the short-term / long-term mean difference |SMA i [k]-LMA i Normalized unit diagnostic bias D of [k]| * diag,i [k]; Use Equation 8 to determine the normalized unit diagnostic bias D. * diag,i The statistical adaptive threshold D of [k] threshold [k]; Using Equation 9, the unit diagnostic deviation D is analyzed. * diag,i [k] Perform filtering to determine the filter diagnostic value D filter,i [k]; and using the filter diagnostic value Dfilter,i [k] time series data is used to diagnose abnormal voltages in battery cells.

[0142] Figure 2f This shows the diagnostic bias D for the normalized unit. * diag,i [k] time series data ( Figure 2d The short-term / long-term average difference | SMA i [k]-LMA i A chart showing the time variation of [k]|. Used to calculate the short-term / long-term mean difference |SMA. i [k]-LMA i In equations 2, 4, and 5 of [k]|, D can be used. * diag,i [k] replaces V i [k], and can be used with D * diag,i The average value of [k] replaces V av [k].

[0143] Figure 2g This shows the normalized unit diagnostic bias D calculated using Equation 6. * diag,i A graph of the time series data for [k]. Figure 2g In the context of statistical adaptive threshold D threshold The time series data of [k] corresponds to the distribution represented by the darkest color.

[0144] Figure 2h This demonstrates the use of Equation 9 to diagnose the deviation D of the unit. * diag,i The filter diagnostic value D obtained by filtering the time series data of [k] is... filter,i Distribution of time series data for [k].

[0145] In the example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k] includes the filter diagnostic value (D). filter,i [k]) is greater than the diagnostic threshold (e.g., 0) in time steps, and battery cells that meet the requirement of accumulating time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0146] Preferably, the control circuit 220 can accumulate and successively satisfy the filter diagnostic value D filter,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0147] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is the number of data included in a time step that is greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement that the cumulative number of data is greater than a preset reference count are diagnosed as abnormal voltage cells.

[0148] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the number of data points included in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0149] Control circuit 220 can additionally repeat the above recursive calculation process a certain number of times. That is, control circuit 220 can use normalized unit diagnostic bias D. * diag,i [k] time series data (e.g., Figure 2g (data) instead Figure 2a The voltage time series data is shown. Additionally, control circuit 220 can recursively perform the following at time index k: calculate the short-term / long-term average difference |SMA i [k]-LMA i [k]|;Calculate the short-term / long-term mean difference |SMA i [k]-LMA i The average value of [k]; calculate the difference between the short-term / long-term average |SMA|. i [k]-LMA i [k]| The unit diagnostic bias D corresponding to the difference compared to the mean. diag,i [k]; Use Equation 6 to calculate the short-term / long-term mean difference |SMA i [k]-LMA i Normalized unit diagnostic bias D of [k]| * diag,i [k]; Use Equation 8 to determine the unit diagnostic deviation D. * diag,i The statistical adaptive threshold D of [k] threshold [k]; Using Equation 9, the unit diagnostic deviation D is analyzed. * diag,i [k] Perform filtering to determine the filter diagnostic value D filter,i [k]; and using the filter diagnostic value D filter,i [k] time series data is used for abnormal voltage diagnosis of battery cells.

[0150] By repeating the above recursive calculation process, abnormal voltage diagnosis of battery cells can be performed more accurately. That is, referring to... Figure 2e The filter diagnostic value D of the battery cell under abnormal voltage conditions filter,i In the time series data of [k], a positive profile pattern was observed only at two time steps. However, referring to... Figure 2h The filter diagnostic value D of the battery cell under abnormal voltage conditions filter,i In the time series data of [k], in comparison Figure 2e A positive distribution pattern was observed over more time steps. Therefore, when the recursive calculation process is performed iteratively, the timing of abnormal voltage occurrences in battery cells can be detected more accurately.

[0151] The battery diagnostic method using the battery diagnostic device 200 described above will be described in detail below. The operation of the control circuit 220 will be described in more detail in various embodiments of the battery diagnostic method.

[0152] Figure 3 This is an exemplary flowchart of a battery diagnostic method according to a first embodiment of the present disclosure. It can be periodically executed by the control circuit 220 at each unit time. Figure 3 The method.

[0153] Reference Figures 1 to 3 In step S310, the control circuit 220 collects data from the voltage sensing circuit 210 representing multiple battery cells BC1 to BC2. N The voltage signal of each cell voltage is used to generate the voltage signal of each battery cell BC (see [link]). Figure 2a The time series data of the unit voltage. At each unit time, the number of time series data points for the unit voltage increases by 1.

[0154] Preferably, V in Formula 5 i [k] or VD i [k] can be used as a unit voltage.

[0155] In step S320, the control circuit 220 is based on each battery cell BC i The time series data of the cell voltage determines the BC of each battery cell. i First average cell voltage SMA i [k] (see Equations 1 and 2) and the second average unit voltage LMA i [k](See Equations 3 and 4)(See also) Figure 2b First average cell voltage SMA i [k] is the value of each battery cell BC iThe cell voltage is a short-term moving average over a first moving window with a first time length. The second average cell voltage is LMA. i [k] is the value of each battery cell BC i The long-term moving average of the cell voltage over a second moving window with a second time length. V can be used. i [k] or VD i [k] is used to calculate the first average unit voltage SMA. i [k] and second average unit voltage LMA i [k].

[0156] In step S330, the control circuit 220 determines each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i [k]|(see also) Figure 2c ).

[0157] In step S340, the control circuit 220 determines each battery cell BC i Unit diagnostic deviation D diag,i [k]. Unit diagnostic deviation D diag,i [k] is the average of the short-term / long-term average differences of all battery cells |SMA i [k]-LMA i [k]| av With the i-th battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i The deviation between [k] and [k].

[0158] In step S350, the control circuit 220 determines whether the diagnostic time has elapsed. The diagnostic time is preset. If step S350 is determined to be "yes", step S360 is executed; if step S350 is determined to be "no", steps S310 to S340 are repeated.

[0159] In step S360, control circuit 220 generates data for each battery cell BC collected during the diagnostic time. i Unit diagnostic deviation D diag,i [k] is a time series data.

[0160] In step S370, the control circuit 220 diagnoses the deviation D through the analysis unit. diag,i [k] time series data, for each battery cell BC i Perform abnormal voltage diagnosis.

[0161] In the example, control circuitry 220 can be accumulated in each battery cell BC.i Unit diagnostic deviation D diag,i In the time series data of [k], the unit diagnostic bias D diag,i A time step [k] greater than the diagnostic threshold (e.g., 0.015) will diagnose battery cells that meet the requirement of having a cumulative time greater than a preset reference time as abnormal voltage cells.

[0162] Preferably, the control circuit 220 can only accumulate the successively satisfied unit diagnostic deviation D diag,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0163] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Unit diagnostic deviation D diag,i In the time series data of [k], the unit diagnostic bias D diag,i [k] The number of data points greater than the diagnostic threshold (e.g., 0.015) will be used to diagnose battery cells that meet the requirement that the cumulative number of data points is greater than a preset reference count as abnormal voltage cells.

[0164] Preferably, the control circuit 220 can only accumulate the successively satisfied unit diagnostic deviation D diag,i [k] is the number of data points included in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0165] Figure 4 This is an exemplary flowchart of a battery diagnostic method according to a second embodiment of the present disclosure. It can be periodically executed by the control circuit 220 at each unit time. Figure 4 The method.

[0166] In the battery diagnostic method of the second embodiment, steps S310 to S360 are substantially the same as in the first embodiment, and their descriptions are omitted. After step S360 is completed, step S380 is executed.

[0167] In step S380, control circuit 220 uses Equation 8 to generate a statistical adaptive threshold D. threshold [k] is the time series data. The input to the Sigma function in Equation 8 is the cell diagnostic deviation D of all battery cells generated in step S360. diag,i [k] is a time series data point. Preferably, the unit diagnostic bias D can be excluded from the input values ​​of the Sigma function. diag,i The maximum value of [k]. Unit diagnostic deviation D diag,i [k] is the short-term / long-term average difference | SMAi [k]-LMA i [k]| Deviation from the mean.

[0168] In step S390, the control circuit 220 uses Equation 9 to control each battery cell BC. i Unit diagnostic deviation D diag,i [k] performs filtering to generate filter diagnostic value D. filter,i [k] is a time series data.

[0169] When using Equation 9, D can be used. diag,i [k] replaces D * diag,i [k].

[0170] In step S400, the control circuit 220 analyzes the filter diagnostic value D. filter,i [k] time series data, for each battery cell BC i Perform abnormal voltage diagnosis.

[0171] In the example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is a time step greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement of having a cumulative time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0172] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0173] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is the number of data included in a time step that is greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement that the cumulative number of data is greater than a preset reference count are diagnosed as abnormal voltage cells.

[0174] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the number of data points included in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0175] Figure 5 This is an exemplary flowchart of a battery diagnostic method according to a third embodiment of the present disclosure. Figure 5 The method can be executed periodically by the control circuit 220 at each unit of time.

[0176] Except that steps S340, S360, and S370 are changed to steps S340', 360', and S370', the battery diagnostic method according to the third embodiment is substantially the same as that of the first embodiment. Therefore, the third embodiment will be described with respect to the differences.

[0177] In step S340', control circuit 220 uses Equation 6 to determine each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i Normalized unit diagnostic bias D of [k]| * diag,i [k]. The normalized reference value is the difference between the short-term and long-term averages |SMA i [k]-LMA i The average value of [k]|. Equation 7 can be used instead of Equation 6.

[0178] In step S360', the control circuit 220 targets each battery cell BC collected during the diagnostic time. i Normalized unit diagnostic bias D * diag,i [k] Generate time series data (see [k]) Figure 2d ).

[0179] In step S370', the control circuit 220 analyzes the normalized unit diagnostic deviation D. * diag,i [k] time series data, for each battery cell BC i Perform abnormal voltage diagnosis.

[0180] In the example, control circuit 220 can accumulate each battery cell BC. i Normalized unit diagnostic bias D * diag,i In the time series data of [k], the unit diagnostic bias D * diag,i [k] is a time step greater than the diagnostic threshold (e.g., 4), and battery cells that meet the requirement of having a cumulative time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0181] Preferably, the control circuit 220 can only accumulate the diagnostic deviations D of the units that successively satisfy the normalization. *diag,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0182] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Normalized unit diagnostic bias D * diag,i The number of cells in the time series data of [k] whose cell diagnostic deviation is greater than the diagnostic threshold (e.g., 4) is identified, and the battery cells that meet the requirement that the cumulative number of data is greater than the preset reference count are diagnosed as abnormal voltage cells.

[0183] Preferably, the control circuit 220 can only accumulate the diagnostic deviations D of the units that successively satisfy the normalization. * diag,i [k] is the number of data points included in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0184] Figure 6 This is an exemplary flowchart of a battery diagnostic method according to a fourth embodiment of the present disclosure. Figure 6 The method can be executed periodically by the control circuit 220 at each unit of time.

[0185] Except for steps S340, S360, S380, S390, and S400 being changed to steps S340′, S360′, S380′, S390′, and S400′ respectively, the battery diagnostic method according to the fourth embodiment is substantially the same as that of the second embodiment. Therefore, the fourth embodiment will be described with respect to the differences from the second embodiment.

[0186] In step S340', control circuit 220 uses Equation 6 to determine each battery cell BC i Short-term / long-term average difference | SMA i [k]-LMA i Normalized unit diagnostic bias D of [k]| * diag,i [k]. The normalized reference value is the difference between the short-term and long-term averages |SMA i [k]-LMA i The average value of [k]|. Equation 7 can be used instead of Equation 6.

[0187] In step S360', the control circuit 220 targets each battery cell BC collected during the diagnostic time. i Normalized unit diagnostic bias D *diag,i [k], generates time series data (see [k]). Figure 2d ).

[0188] In step S380', control circuit 220 uses Equation 8 to generate a statistical adaptive threshold D. threshold [k] is the time series data. The input to the Sigma function in Equation 8 is the normalized cell diagnostic bias D of all battery cells generated in step S360′. * diag,i [k] is a time series data point. Preferably, at each time index, the cell diagnostic bias D of the cell can be excluded from the input value of the Sigma function. * diag,i The maximum value of [k].

[0189] In step S390', the control circuit 220 uses Equation 9 based on the statistical adaptive threshold D threshold [k] for each battery cell BC i Unit diagnostic deviation D * diag,i [k] performs filtering to generate filter diagnostic value D. filter,i [k] is a time series data.

[0190] In step S400', the control circuit 220 analyzes the filter diagnostic value D. filter,i [k] time series data, for each battery cell BC i Perform abnormal voltage diagnosis.

[0191] In the example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The filter diagnostic value D in the time series data of [k] filter,i [k] is a time step greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement of having a cumulative time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0192] Preferably, the control circuit 220 can accumulate and successively satisfy the filter diagnostic value D filter,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0193] In another example, control circuit 220 can accumulate each battery cell BC. i Filter diagnostic value D filter,i The filter diagnostic value D in the time series data of [k] filter,i[k] is the number of data included in a time step that is greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement that the cumulative number of data is greater than a preset reference count are diagnosed as abnormal voltage cells.

[0194] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the number of data points in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points in each time step.

[0195] Figure 7 This is an exemplary flowchart of a battery diagnostic method according to a fifth embodiment of the present disclosure.

[0196] In the fifth embodiment, steps S310 to S360' are substantially the same as in the fourth embodiment. Therefore, the fifth embodiment will be described in terms of the differences compared to the fourth embodiment.

[0197] In step S410, the control circuit 220 uses each battery cell BC i Normalized unit diagnostic bias D * diag,i Time series data of [k] are used to generate a unit diagnostic bias D. * diag,i The first moving average (SMA) of [k] i [k] Time series data and second moving average (LMA) i [k] Time series data (see [k]) Figure 2f ).

[0198] In step S420, the control circuit 220 utilizes Equation 6, using each battery cell BC i First moving average (SMA) i [k] Time series data and second moving average (LMA) i [k] Time series data, generating normalized unit diagnostic bias D * diag,i [k] Time series data (see [k]) Figure 2g ).

[0199] In step S430, control circuit 220 uses Equation 8 to generate a statistical adaptive threshold D. threshold [k] time series data (see Figure 2g ).

[0200] In step S440, the control circuit 220 uses Equation 9 based on the statistical adaptive threshold D threshold [k], generate each battery cell BC iFilter diagnostic value D filter,i [k] time series data (see Figure 2h ).

[0201] In step S450, the control circuit 220 analyzes each battery cell BC i Filter diagnostic value D filter,i [k] time series data, for each battery cell BC i Perform abnormal voltage diagnosis.

[0202] In the example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is a time step greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement of having a cumulative time greater than a preset reference time are diagnosed as abnormal voltage cells.

[0203] Preferably, the control circuit 220 can accumulate and successively satisfy the filter diagnostic value D. filter,i [k] is the time step required to exceed the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently calculate the cumulative time for each time step.

[0204] In another example, control circuitry 220 can be accumulated in each battery cell BC. i Filter diagnostic value D filter,i The time series data of [k], including the filter diagnostic value D. filter,i [k] is the number of data included in a time step that is greater than the diagnostic threshold (e.g., 0), and battery cells that meet the requirement that the cumulative number of data is greater than a preset reference count are diagnosed as abnormal voltage cells.

[0205] Preferably, the control circuit 220 can only accumulate the filter diagnostic values ​​D that are successively satisfied. filter,i [k] is the number of data points included in the time step that exceeds the diagnostic threshold. When there are multiple corresponding time steps, the control circuit 220 can independently accumulate the number of data points for each time step.

[0206] In the fifth embodiment, the control circuit 220 may recursively execute steps S410 and S420 at least twice. That is, the control circuit 220 may use the normalized cell diagnostic deviation D generated in step S420. * diag,i [k] Time series data, in step S410, the unit diagnostic bias D is generated again. * diag,i The first moving average (SMA) of [k]i [k] Time series data and second moving average (LMA) i [k] Time series data. Subsequently, the control circuit 220 can reuse each battery cell BC in step S420. i First moving average (SMA) i [k] Time series data and second moving average (LMA) i [k] Based on Equation 6, the time series data generates a normalized unit diagnostic bias D. * diag,i [k] Time series data. The recursive algorithm can be repeated a preset number of times.

[0207] When steps S410 and S420 are executed according to the recursive algorithm, the unit diagnostic deviation D, which is finally calculated by the recursive algorithm, can be used. * diag,i [k] Use time series data to perform steps S430 to S450.

[0208] In embodiments of this disclosure, when an abnormal voltage is diagnosed in a specific battery cell after abnormal voltage diagnosis of all battery cells, the control circuit 220 can output diagnostic result information via a display unit (not shown). Additionally, the control circuit 220 can record in a storage unit the identification information (ID) of the battery cell whose abnormal voltage was diagnosed, the time of the abnormal voltage diagnosis, and a diagnostic flag.

[0209] Preferably, the diagnostic result information may include a message indicating that there are cells in the battery pack that are in an abnormal voltage condition. Optionally, the diagnostic result information may include a warning message indicating that the battery cells need to be precisely inspected.

[0210] In one example, the display unit may be included in a load device that is supplied with power from the unit group CG. When the load device is an electric vehicle, a hybrid vehicle, or a plug-in hybrid vehicle, diagnostic result information can be output via a cluster information display. In another example, when the battery diagnostic device 200 according to this disclosure is included in a diagnostic system, diagnostic results can be output via a display provided in the diagnostic system.

[0211] Preferably, the battery diagnostic device 200 according to an embodiment of the present disclosure may be included in the battery management system 100 or the control system (not shown) of the load device.

[0212] According to the above implementation method, the abnormal voltage of each battery cell can be effectively and accurately diagnosed by determining the two moving average values ​​of the cell voltages of each battery cell at two different time lengths in each unit time, and by performing abnormal voltage diagnosis of each battery cell based on the difference between the two moving average values ​​of each of the multiple battery cells.

[0213] On the other hand, by applying advanced techniques such as normalization and / or statistical adaptive thresholding to analyze the difference in the changing trends of the two moving averages of each battery cell, accurate diagnosis of abnormal voltages of each battery cell can be achieved.

[0214] According to another aspect, the time step of abnormal voltage occurrence and / or abnormal voltage detection count of each battery cell can be accurately detected by analyzing the time series data of filter diagnostic values ​​determined based on statistical adaptive thresholds.

[0215] The embodiments of this disclosure described above are not implemented solely by devices and methods, but can be implemented by a program that performs functions corresponding to the configuration of the embodiments of this disclosure or by a recording medium on which the program is recorded, and those skilled in the art can easily implement such implementations from the disclosure of the above embodiments.

[0216] Although this disclosure has been described above with respect to a limited number of embodiments and accompanying drawings, this disclosure is not limited thereto, and it will be apparent to those skilled in the art that various modifications and variations can be made thereto within the technical aspects of this disclosure and within the equivalent scope of the appended claims.

[0217] Additionally, since those skilled in the art can make many substitutions, modifications and changes to the present disclosure described above without departing from the technical aspects of the present disclosure, the present disclosure is not limited to the above embodiments and drawings, and some or all of the embodiments can be selectively combined to allow for various modifications.

Claims

1. A battery diagnostic device for a cell group, the cell group comprising a plurality of battery cells connected in series, the battery diagnostic device comprising: A voltage sensing circuit configured to periodically generate a voltage signal indicating the cell voltage of each battery cell; as well as A control circuit configured to generate time-series data indicating the change of cell voltage over time for each battery cell based on the voltage signal. The control circuit is configured as follows: (i) Determine a first average cell voltage and a second average cell voltage for each battery cell based on the time series data, wherein the first average cell voltage is a short-term moving average and the second average cell voltage is a long-term moving average; and (ii) Detect abnormal voltage of each battery cell based on the difference between the first average cell voltage and the second average cell voltage.

2. The battery diagnostic device according to claim 1, wherein, The control circuit is configured as follows: For each battery cell, a short-term / long-term average difference is determined corresponding to the difference between the first average cell voltage and the second average cell voltage; For each battery cell, a cell diagnostic deviation is determined corresponding to the deviation between the average of the short-term / long-term average differences of all said battery cells and the deviation of said short-term / long-term average differences of that battery cell; as well as A battery cell that meets the requirement that the diagnostic deviation of the cell exceeds the diagnostic threshold will be detected as an abnormal voltage cell.

3. The battery diagnostic device according to claim 2, wherein, The control circuit is configured to: generate time-series data of the cell diagnostic deviation for each battery cell, and detect abnormal voltage of the battery cell based on the time period during which the cell diagnostic deviation exceeds the diagnostic threshold or the number of data points of the cell diagnostic deviation exceeding the diagnostic threshold.

4. The battery diagnostic device according to claim 1, wherein, The control circuit is configured as follows: For each battery cell, a short-term / long-term average difference is determined corresponding to the difference between the first average cell voltage and the second average cell voltage; For each battery cell, the cell diagnostic deviation is determined by calculating the deviation between the average of the short-term / long-term average differences of all the battery cells and the short-term / long-term average difference of the battery cell. A statistical adaptive threshold is determined, which depends on the standard deviation of the cell diagnostic deviation for all the battery cells; Time series data of filter diagnostic values ​​are generated by filtering the time series data of the cell diagnostic deviation for each battery cell based on the statistical adaptive threshold. as well as Abnormal voltage of the battery cell is detected based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

5. The battery diagnostic device according to claim 1, wherein, The control circuit is configured as follows: For each battery cell, a short-term / long-term average difference is determined corresponding to the difference between the first average cell voltage and the second average cell voltage; For each battery cell, the normalized value of the short-term / long-term average difference is determined as the normalized cell diagnostic bias, and a statistical adaptive threshold is determined, which depends on the standard deviation of the normalized cell diagnostic bias for all battery cells. Time series data of filter diagnostic values ​​are generated by filtering the normalized cell diagnostic deviation time series data for each battery cell based on the statistical adaptive threshold. as well as Abnormal voltage of the battery cell is detected based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

6. The battery diagnostic device according to claim 5, wherein, The control circuit is configured to normalize the short-term / long-term average difference for each battery cell by dividing the short-term / long-term average difference by the average of the short-term / long-term average differences of all the battery cells.

7. The battery diagnostic device according to claim 5, wherein, The control circuit is configured to normalize the short-term / long-term average difference for each battery cell by performing a logarithmic calculation on the short-term / long-term average difference.

8. The battery diagnostic device according to claim 1, wherein, The control circuit is configured to generate time-series data indicating the change of the cell voltage of each battery cell over time, using the voltage difference between the average cell voltage of all the battery cells measured at each unit time and the cell voltage of each battery cell.

9. The battery diagnostic device according to claim 1, wherein, The control circuit is configured as follows: For each battery cell, a short-term / long-term average difference is determined corresponding to the difference between the first average cell voltage and the second average cell voltage; For each battery cell, the normalized value of the short-term / long-term average difference is determined as the normalized cell diagnostic bias, and time series data of the normalized cell diagnostic bias is generated for each battery cell. The normalized time-series data of cell diagnostic bias is generated for each cell by recursively repeating (i) to (iv) at least once: (i) For the time series data of the normalized cell diagnostic deviation for each battery cell, determine a first moving average and a second moving average, wherein the first moving average is a short-term moving average and the second moving average is a long-term moving average; (ii) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first moving average and the second moving average; (iii) For each battery cell, determine the normalized value of the short-term / long-term average difference as the normalized cell diagnostic deviation; and (iv) For each battery cell, generate the time series data of the normalized cell diagnostic deviation. A statistical adaptive threshold is determined, which depends on the standard deviation of the normalized cell diagnostic bias for all the battery cells; By filtering the time-series data of the normalized cell diagnostic deviation for each battery cell based on the statistical adaptive threshold, time-series data of filtered diagnostic values ​​are generated; and Abnormal voltage of the battery cell is detected based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

10. The battery diagnostic device according to claim 1, wherein, The first average cell voltage is a short-term moving average of the cell voltage of each battery cell over a first moving window with a first time length, and the second average cell voltage is a long-term moving average of the cell voltage of each battery cell over a second moving window with a second time length.

11. The battery diagnostic device according to claim 10, wherein, The first time length is shorter than the second time length.

12. A battery pack comprising a battery diagnostic device according to any one of claims 1 to 11.

13. A vehicle comprising a battery pack according to claim 12.

14. A battery diagnostic method for a battery pack, the battery pack comprising a plurality of battery cells connected in series, the battery diagnostic method comprising the following steps: (a) Periodically generate time-series data indicating the change of cell voltage over time for each battery cell; (b) Determine a first average cell voltage and a second average cell voltage for each battery cell based on the time series data, wherein the first average cell voltage is a short-term moving average and the second average cell voltage is a long-term moving average; as well as (c) Detect abnormal voltage of each battery cell based on the difference between the first average cell voltage and the second average cell voltage.

15. The battery diagnostic method according to claim 14, wherein, Step (c) includes the following steps: (c1) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage; (c2) For each battery cell, determine the cell diagnostic deviation corresponding to the deviation between the average of the short-term / long-term average differences of all said battery cells and the deviation between the short-term / long-term average differences of that battery cell; and (c3) The battery cell that meets the requirement that the unit diagnostic deviation exceeds the diagnostic threshold is detected as an abnormal voltage cell.

16. The battery diagnostic method according to claim 15, wherein, Step (c3) includes the following steps: For each battery cell, time-series data of the cell diagnostic deviation is generated; and Abnormal voltage of the battery cell is detected based on the time period during which the cell diagnostic deviation exceeds the diagnostic threshold or the number of data points showing the cell diagnostic deviation exceeding the diagnostic threshold.

17. The battery diagnostic method according to claim 14, wherein, Step (c) includes the following steps: (c1) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage; (c2) For each battery cell, a cell diagnostic deviation is determined by calculating the deviation between the average of the short-term / long-term average differences of all said battery cells and the short-term / long-term average difference of the battery cell. (c3) Determine a statistical adaptive threshold, which depends on the standard deviation of the cell diagnostic deviation for all the battery cells; (c4) By filtering the time-series data of the cell diagnostic deviation for each battery cell based on the statistical adaptive threshold, time-series data of filtered diagnostic values ​​are generated for each battery cell; and (c5) Detect abnormal voltage of the battery cell based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

18. The battery diagnostic method according to claim 14, wherein, Step (c) includes the following steps: (c1) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage; (c2) For each battery cell, the normalized value of the short-term / long-term average difference is determined as the normalized cell diagnostic bias; (c3) Determine a statistical adaptive threshold, which depends on the standard deviation of the normalized cell diagnostic bias for all the battery cells; (c4) Filtering the normalized cell diagnostic deviation time-series data for each battery cell based on the statistical adaptive threshold to generate time-series data of filtered diagnostic values; and (c5) Detect abnormal voltage of the battery cell based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

19. The battery diagnostic method according to claim 18, wherein, Step (c2) includes the following steps: for each battery cell, normalizing the short-term / long-term average difference by dividing the short-term / long-term average difference by the average of the short-term / long-term average differences of all the battery cells.

20. The battery diagnostic method according to claim 18, wherein, Step (c2) includes the following steps: for each battery cell, normalizing the short-term / long-term average difference by performing a logarithmic calculation on the short-term / long-term average difference.

21. The battery diagnostic method according to claim 14, wherein, Step (a) includes the following steps: using the voltage difference between the average cell voltage of all said battery cells measured at each unit time and the cell voltage of each battery cell, to generate time series data indicating the change of cell voltage of each battery cell over time.

22. The battery diagnostic method according to claim 14, wherein, Step (c) includes the following steps: (c1) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first average cell voltage and the second average cell voltage; (c2) For each battery cell, the normalized value of the short-term / long-term average difference is determined as the normalized cell diagnostic bias; (c3) Generate the normalized time-series data of cell diagnostic bias for each battery cell; (c4) Generate the normalized cell diagnostic bias time series data for each cell by recursively repeating (i) to (iv) at least once: (i) For the normalized cell diagnostic deviation time series data for each battery cell, determine a first moving average and a second moving average, wherein the first moving average is a short-term moving average and the second moving average is a long-term moving average; (ii) For each battery cell, determine the short-term / long-term average difference corresponding to the difference between the first moving average and the second moving average; (iii) For each battery cell, determine the normalized value of the short-term / long-term average difference as the normalized cell diagnostic deviation; and (iv) For each battery cell, generate the normalized cell diagnostic deviation time series data. (c5) Determine a statistical adaptive threshold, which depends on the standard deviation of the normalized cell diagnostic bias for all the battery cells; (c6) By filtering the time-series data of the normalized cell diagnostic deviation for each battery cell based on the statistical adaptive threshold, time-series data of the filter diagnostic values ​​are generated; and (c7) Detect abnormal voltage of the battery cell based on the time period during which the filter diagnostic value exceeds the diagnostic threshold or the number of data points of the filter diagnostic value that exceed the diagnostic threshold.

Images