Battery state detection device and program

Abstract

The battery system has a battery module (11) in which a plurality of battery cells (13) are connected in series, and the state of each battery cell is detected by a state detection unit. A battery state detection device (20) is provided with: an acquisition unit that acquires a state detection value for each battery cell detected by a state detection unit; a parameter change amount calculation unit that calculates, as a state parameter of the battery cell, either the state detection value acquired by the acquisition unit or a state calculation value indicating the state of the battery cell calculated on the basis of the state detection value, and calculates, as a parameter change amount, a change amount of the state parameter in a predetermined period; and an abnormality detection unit that detects an abnormality of a target cell on the basis of whether or not a predetermined deviation has occurred in the amount of change in parameters with respect to the battery cells other than the target cell, the target cell being a target cell to be subjected to abnormality detection, each of the plurality of battery cells being a target cell to be subjected to abnormality detection.

Description

Cross-reference of related applications

[0001] This application is based on Japanese Application No. 2023-148760, filed on September 13, 2023, and Japanese Application No. 2024-097399, filed on June 17, 2024, the contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to a battery status detection device and program. Background Technology

[0003] In the past, various technologies have been proposed for determining whether there are any abnormalities in batteries installed in vehicles, etc. For example, in the technology described in Patent Document 1, in an energy storage module formed by connecting energy storage devices to each other using a bus, the instantaneous resistance value after the start of constant current charging and the diffusion resistance value after a certain period of time have elapsed since the start of constant current charging are extracted, and the ratio of the instantaneous resistance value to the diffusion resistance value is compared with a predetermined threshold to determine whether there is any abnormality in the bus connection.

[0004] Existing technical documents Patent documents Patent Document 1: Japanese Patent Application Publication No. 2015-68821 Summary of the Invention

[0005] Thermal runaway is an anomaly in batteries, and various techniques have been proposed for its detection. For example, in the initial stage of battery thermal runaway, an internal short circuit occurs within the battery cell. Therefore, a technique is considered whereby changes in internal resistance and terminal voltage within the battery cell are determined by comparing them to predetermined thresholds, and based on these results, an internal short circuit anomaly is determined to have occurred within the battery cell.

[0006] However, in cases where an internal short circuit occurs in a battery cell as a precursor to thermal runaway, the change in diffusion resistance develops slowly over time in the early stages of the internal short circuit. Therefore, in situations such as voltage fluctuations caused by charging and discharging in vehicles, and vehicle noise, there are concerns about the inability to accurately detect internal short circuits in battery cells. In particular, in anomaly detection based on individual battery cells, there are concerns about false detections due to environmental factors such as temperature.

[0007] This disclosure was made in view of the above-mentioned issues, and its purpose is to provide a battery state detection device and program that can quickly and accurately detect internal short-circuit abnormalities in battery cells that are signs of thermal runaway.

[0008] The battery state detection device disclosed herein is applied to a battery system having a battery module formed by connecting multiple battery cells in series. A state detection unit detects the state of each of the battery cells. The battery state detection device comprises: an acquisition unit that acquires the state detection values ​​of each of the battery cells detected by the state detection unit; a parameter change calculation unit that takes one of the state detection values ​​acquired by the acquisition unit and a state calculation value representing the state of the battery cell calculated based on the state detection values ​​as a state parameter of the battery cell, and calculates the change in the state parameter over a predetermined period as a parameter change; and an anomaly detection unit that identifies each of the multiple battery cells as a target cell for anomaly detection, and detects an anomaly of the target cell based on whether a predetermined deviation has occurred in the parameter change relative to other battery cells.

[0009] Based on the above structure, anomaly detection of individual battery cells in the battery module is not performed as an individual cell anomaly detection, but rather as a relative anomaly detection with respect to other battery cells. Therefore, even if voltage fluctuations or noise occur in the power supply path of the battery module, their impact can be suppressed and anomaly detection can be implemented. Furthermore, false detections caused by environmental conditions such as temperature in the battery module are suppressed. Therefore, in the event of an internal short circuit occurring as the initial stage of thermal runaway, even if the change in diffusion resistance in the battery cell occurs slowly over time, the difference in parameter changes between normal and abnormal cells can be detected early, allowing for appropriate detection of abnormal cells. As a result, internal short circuit anomalies in battery cells that are precursors to thermal runaway can be detected rapidly and with high accuracy.

[0010] Furthermore, the battery state detection device disclosed herein is applied to a battery system having a battery module formed by connecting multiple battery cells in series. A state detection unit detects the state of each battery cell. The battery state detection device comprises: a first anomaly detection unit that performs anomaly detection on each battery cell based on the state detection values ​​detected by the state detection unit using computationally lightweight processing; a second anomaly detection unit that performs anomaly detection on each battery cell based on the state detection values ​​detected by the state detection unit using computationally heavy processing; and a switching execution unit that switches between the first anomaly detection unit and the second anomaly detection unit detecting anomalies in each battery cell according to predetermined switching conditions.

[0011] Based on the above structure, two types of anomaly determinations can be performed during anomaly detection of each battery cell, and these anomaly determinations can be switched according to predetermined switching conditions. Here, under long-term use of battery cells, situations may arise where cell anomalies are easily detected and situations where cell anomalies are difficult to detect. In such cases, by switching the anomaly determination according to predetermined switching conditions, excessive computational load can be suppressed, and appropriate anomaly detection can be achieved. Attached Figure Description

[0012] The above-mentioned and other objects, features, and advantages of this disclosure will become more apparent from the accompanying drawings and from the detailed description below.

[0013] Figure 1 This is a diagram of the overall structure of the battery system.

[0014] Figure 2 It is a diagram used to illustrate the cell voltage of each battery cell in the battery module.

[0015] Figure 3 This is a diagram representing the equivalent circuit of a battery cell.

[0016] Figure 4 It is a time-series diagram showing the changes in the cell voltage of each battery cell.

[0017] Figure 5 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the first embodiment.

[0018] Figure 6 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the first embodiment.

[0019] Figure 7 It is a timing diagram used to specifically illustrate the abnormal detection and handling of battery cells.

[0020] Figure 8 This is a diagram illustrating the outline of cell anomaly detection in the second embodiment.

[0021] Figure 9 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the second embodiment.

[0022] Figure 10 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the second embodiment.

[0023] Figure 11 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the third embodiment.

[0024] Figure 12 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the third embodiment.

[0025] Figure 13 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the third embodiment.

[0026] Figure 14 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the fourth embodiment.

[0027] Figure 15 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the fourth embodiment.

[0028] Figure 16 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the fourth embodiment.

[0029] Figure 17 This is a diagram showing the structure of the battery system in the fifth embodiment.

[0030] Figure 18 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the fifth embodiment.

[0031] Figure 19 It is a time-series diagram showing the changes in cell voltage in a battery pack.

[0032] Figure 20 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the sixth embodiment.

[0033] Figure 21 This is a diagram showing the structure of the battery system in the seventh embodiment.

[0034] Figure 22 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the seventh embodiment.

[0035] Figure 23 This is a flowchart illustrating the abnormal detection and processing of the battery cell in the eighth embodiment.

[0036] Figure 24 This is a flowchart illustrating the anomaly detection and processing of battery cells in other examples.

[0037] Figure 25 It is a timing diagram showing the changes in cell voltage and cell temperature during the internal short-circuit process.

[0038] Figure 26 This is a diagram illustrating an example of the structure of a battery system.

[0039] Figure 27 This is a flowchart illustrating the anomaly detection and processing of battery cells in other examples.

[0040] Figure 28 This is a graph representing the voltage-capacitance characteristics of a battery cell.

[0041] Figure 29 This is a flowchart illustrating the anomaly detection and processing of battery cells in other examples. Detailed Implementation

[0042] The battery system of this embodiment will now be described. In this embodiment, a battery system mounted on an electric vehicle such as a hybrid electric vehicle or an electric vehicle will be described with reference to the accompanying drawings. This battery system has a high-voltage battery with a voltage of several hundred volts between the positive and negative terminals. Furthermore, in addition to electric vehicles, this battery system can also be applied to aircraft such as drones, ships, construction machinery, and agricultural machinery.

[0043] (First Implementation) First, use Figure 1 This section provides an overview of the vehicle's battery system. Figure 1 In this battery system, there are multiple battery cells 10 and a battery ECU 20 that manages these battery cells 10 uniformly. Each battery cell 10 has a battery module 11 and a monitoring device 12 that monitors the battery module 11. The battery module 11 is constructed by connecting multiple battery cells 13 in series. Each battery cell 13 is, for example, a lithium-ion secondary battery or a nickel-metal hydride secondary battery. Furthermore, lithium-ion secondary batteries are secondary batteries that use lithium as the charge carrier; in addition to conventional lithium-ion secondary batteries with a liquid electrolyte, all-solid-state batteries using a solid electrolyte may also be included.

[0044] The battery modules 11 of each battery cell 10 are connected in series to form a battery pack B. Battery pack B is a high-voltage battery that generates an inter-terminal voltage of several hundred volts. In addition, battery pack B can also connect the battery modules 11 in parallel, or connect the battery modules 11 in series and in parallel.

[0045] Each monitoring device 12 of the battery cell 10 is a slave unit of the battery ECU 20, and the battery ECU 20 and the multiple monitoring devices 12 can communicate with each other via wired or wireless communication units. The battery ECU 20 is composed of a microcomputer including a CPU, ROM, RAM, etc. The battery ECU 20 executes various processes related to battery control based on the program stored in the ROM. For example, the battery ECU 20 executes control related to the charging or discharging of the battery pack B. The monitoring devices 12 obtain information related to the battery module 11 according to instructions from the battery ECU 20 and send the information to the battery ECU 20. The monitoring devices 12 can obtain information related to the battery module 11 at a predetermined period. In addition, the monitoring devices 12 are also referred to as satellite monitoring devices. In this embodiment, the battery ECU 20 is equivalent to a "battery state detection device".

[0046] The monitoring device 12 is equipped with a voltage sensor 31 for detecting the voltage between the terminals of each battery cell 13. Specifically, in the battery module 11, an electrical path 32 is connected to both ends of each battery cell 13, and a voltage sensor 31 is connected to this electrical path 32. The voltage sensor 31 acts as a "state detection unit," and the voltage between the terminals (cell voltage) detected by the voltage sensor 31 acts as a "state detection value." Furthermore, a current sensor 33 is provided in the cell series path of the battery module 11 to detect the current flowing through each battery cell 13.

[0047] Battery pack B is connected to a rotary motor unit as an electrical load driven by power supplied from battery pack B. The rotary motor unit includes a rotary motor 41 and an inverter 42 that acts as a power converter to allow phase current to flow through the rotary motor 41. The rotary motor 41 is the vehicle's main unit, capable of transmitting power to drive wheels (not shown). The rotary motor 41 is, for example, a three-phase AC motor. In the inverter 42, the current flowing through each phase winding is adjusted according to the on / off state of switches (semiconductor switching elements) provided for each phase winding.

[0048] Furthermore, in the event of thermal runaway in battery module 11, as part of the process leading to thermal runaway, an internal short circuit occurs in the battery cell 13 in the initial stage, followed by gas generation, deformation of the cell casing, and subsequent fire. In this situation, to quickly detect the signs of thermal runaway, it is preferable to detect the internal short circuit in the battery cell 13. In this embodiment, the internal short circuit anomaly is detected based on the change in cell voltage in each battery cell 13.

[0049] The following explains the change in cell voltage of battery cell 13 caused by an internal short circuit.

[0050] Figure 2 This is a diagram used to illustrate the cell voltage of each battery cell 13 in battery module 11. Figure 2 In the battery module 11, n battery cells 13 are connected in series. The battery cells 13 are sequentially designated as the first cell C1, the second cell C2, the third cell C3, ..., the nth cell Cn, starting from the low potential side. Furthermore, the terminal voltages of each cell C1~Cn are designated as cell voltages V1~Vn, respectively. For each cell voltage V1~Vn, the change over a specified period is designated as voltage change VA1~VAn. For every two adjacent battery cells 13 in the series path, the difference between voltage change VA1~VAn is designated as voltage difference ΔV21, ΔV32, ..., ΔVn,n-1.

[0051] The voltage differences ΔV21, ΔV32, ..., ΔVn,n-1 are values ​​calculated for each pair of adjacent battery cells 13 in the series path by subtraction operations based on the ascending or descending order of the cell series connection. Here, they are defined as the values ​​obtained by subtracting the voltage change VA of the battery cell 13 on the low-potential side from the voltage change VA of the battery cell 13 on the high-potential side. For example, the voltage difference ΔV21 is the value obtained by subtracting the voltage change VA1 of the first battery cell C1 from the voltage change VA2 of the second battery cell C2, and the voltage difference ΔV32 is the value obtained by subtracting the voltage change VA2 of the second battery cell C2 from the voltage change VA3 of the third battery cell C3. Furthermore, the voltage difference ΔV between the battery cells 13 can simply represent the difference in voltage change VA between two battery cells 13 out of the n battery cells 13. For example, the voltage difference ΔV31 is obtained by subtracting the voltage change VA1 of the first battery cell C1 from the voltage change VA3 of the third battery cell C3, and the voltage difference ΔV1n is obtained by subtracting the voltage change VAn of the nth battery cell Cn from the voltage change VA1 of the first battery cell C1.

[0052] Figure 3 This is a diagram showing the equivalent circuit of battery cell 13. Battery cell 13 is constructed by connecting the electromotive force OCV, solution resistance RΩ, charge transfer resistance Rct, and interface capacitance Cd in parallel, and the diffusion resistance Zw in series. In battery cell 13, when an internal short circuit occurs, the positive electrode path and the negative electrode path are short-circuited through the internal short-circuit resistance Rint, and a short-circuit current flows inside the cell.

[0053] Figure 4 This is a timing diagram showing the changes in cell voltage of multiple battery cells 13 in battery module 11. Here, an example is shown. Figure 2 The voltage changes of the first cell C1 to the third cell C3 among the n battery cells 13 shown are given. Assuming that the second cell C2 has an internal short circuit, explain the changes in the cell voltage of each battery cell 13 and the changes in the voltage difference between the battery cells 13.

[0054] In the event of an internal short circuit in the second cell C2, a short-circuit current flows inside the cell. Therefore, when the battery module 11 discharges, the voltage drop of each cell 13 is greater than that of the first cell C1 and the third cell C3 per unit time (i.e., the slope of the voltage change) when the cell voltage V2 of the second cell C2 drops. Here, taking the start of discharge as the reference (initial value), the voltage changes of each cell C1 to C3 from the start of discharge are denoted as VA1 to VA3. The voltage changes VA1 to VA3 gradually increase towards the negative side over time, but their magnitudes differ in the second cell C2 compared to the other cells C1 and C3. In absolute terms, the relationship between the voltage changes VA1 to VA3 is |VA2| > |VA1| and |VA2| > |VA3|.

[0055] Furthermore, the absolute difference between the voltage changes VA1 and VA2 of cells C1 and C2, i.e., the voltage difference ΔV21, gradually increases towards the positive side over time, while the absolute difference between the voltage changes VA2 and VA3 of cells C2 and C3, i.e., the voltage difference ΔV32, gradually increases towards the negative side over time. Moreover, regarding cells C1 and C3, even over time, the voltage difference ΔV13 remains approximately constant (approximately zero).

[0056] exist Figure 4 The diagram illustrates the period during which power is cut off from battery module 11 after discharge. During this period, the cell voltages V1 and V3 of cells C1 and C3 remain constant. In contrast, in the second cell C2, a short-circuit current flows, causing the cell voltage V2 to decrease. Therefore, even after discharge, the voltage difference ΔV21 increases towards the positive side, and the voltage difference ΔV32 increases towards the negative side.

[0057] This embodiment focuses on detecting internal short-circuit anomalies in each battery cell 13 by considering the voltage changes of each battery cell 13 as described above. In general, the voltage change VA of each battery cell 13 is calculated, and each of the multiple battery cells 13 is considered as a target cell for anomaly detection. The anomaly of the target cell is detected based on whether the voltage change VA in that target cell deviates from the predetermined value of the other battery cells 13.

[0058] In this embodiment, specifically for the target battery cell, the difference between the voltage change VA and that of two or more different battery cells 13 is calculated as a voltage difference value ΔV. Based on this voltage difference value ΔV, it is determined whether a predetermined deviation has occurred in the target battery cell relative to other battery cells 13. Based on this determination result, an anomaly of the target battery cell is detected. Furthermore, the cell voltage V, which is a state detection value, corresponds to a "state parameter," the voltage change VA corresponds to a "parameter change," and the voltage difference value ΔV corresponds to a "change difference." The anomaly detection processing in this embodiment will be described in detail below.

[0059] Figure 5 This is a flowchart illustrating the steps of abnormal detection and processing for battery cell 13, which is executed by battery ECU 20. Battery ECU 20 performs abnormal detection on battery cell 13 for each battery module 11.

[0060] In step S101, the cell voltage V(tm) of each battery cell 13 at time tm is obtained from the monitoring device 12. Time tm is the time when the cell voltage V of each battery cell 13 is obtained at a predetermined time period. The interval of time tm is, for example, about tens to hundreds of milliseconds. However, the interval of time tm can also be several seconds, several minutes, or several hours.

[0061] In step S102, the change in cell voltage V(tm) from the initial value (initial voltage value) is calculated for each battery cell 13, i.e., the voltage change VA(tm) (VA(tm) = V(tm) - initial value). In the next step S103, the absolute value of the voltage change VA(tm) is calculated. The voltage change VA(tm) is the change in cell voltage V(tm) from the initial value to the current time, which is a specified period. The initial value can be any value set for each battery cell 13 at any time interval, and it remains set until the next set time interval arrives. The initial value can be reset according to the period of the specified period (e.g., an interval of several days to several months), or according to the specified driving distance of the vehicle, or according to the specified number of starts of the vehicle, or according to the timing of the switching between discharging and charging of the battery module 11. Furthermore, if the equalization process is performed in the battery module 11 to equalize the cell voltage of each battery cell 13, the initial value can be set again based on the fact that the equalization process has been performed. In addition, step S101 is equivalent to the "acquisition unit", and step S102 is equivalent to the "parameter change calculation unit".

[0062] Then, in step S104, for all battery cells 13, the difference in the absolute value of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as the voltage difference ΔVi+1,i(tm). The voltage difference ΔVi+1,i(tm) can be calculated using the following equation (1).

[0063] ΔVi+1,i(tm)=|VAi+1(tm)|-|VAi(tm)|…(1) Furthermore, the designation i+1, i represents the cell number from 1 to n. The voltage difference ΔVi+1,i(tm) is calculated for all combinations of two adjacent cells in battery module 11.

[0064] When each battery cell 13 of the battery module 11 is considered as a target cell for anomaly detection, the voltage difference ΔVi,i-1 (tm) between the target cell (cell Ci) and the first adjacent cell (cell Ci-1) on the negative side of the target cell is calculated, and the voltage difference ΔVi+1,i (tm) between the target cell (cell Ci) and the second adjacent cell (cell Ci+1) on the positive side of the target cell is calculated. These voltage differences ΔVi,i-1 (tm) and ΔVi+1,i (tm) are obtained by subtracting the absolute value of the voltage change VA (tm) of each battery cell 13 in descending order of cell series connection. Alternatively, the voltage difference ΔV (tm) can also be calculated by subtraction in ascending order of cell series connection. The voltage difference ΔVi,i-1(tm) between the target cell and the first adjacent cell on the negative side corresponds to the "first difference," and the voltage difference ΔVi+1,i(tm) between the target cell and the second adjacent cell on the positive side corresponds to the "second difference." Furthermore, the first and second adjacent cells are battery cells 13 that are different from the target cell, and are therefore also equivalent to the "first non-target cell" and "second non-target cell," respectively. Step S104 corresponds to the "difference calculation unit."

[0065] Next, in step S105, each battery cell 13 (cells C1 to Cn) is taken as a target cell, and for each target cell, the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm) are compared with a predetermined threshold. Specifically, it is determined whether the voltage difference ΔVi,i-1(tm) calculated in the combination of the target cell and the first adjacent cell on the negative side is above the predetermined positive side threshold TH11, and whether the voltage difference ΔVi+1,i(tm) calculated in the combination of the target cell and the second adjacent cell on the positive side is below the predetermined negative side threshold TH12.

[0066] The positive threshold TH11 is a threshold determined by a positive value (TH11 > 0), and the negative threshold TH12 is a threshold determined by a negative value (TH12 < 0). All thresholds TH11 and TH12 have the same absolute value. However, the absolute values ​​of each threshold TH11 and TH12 can also be different.

[0067] In addition, the specific determination method for step S105 can be arbitrary. In short, as long as for the cell Ci which is the target cell, it is determined whether the polarities of the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm) are opposite to each other, and whether the absolute values ​​of the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm) are respectively above the specified threshold.

[0068] Furthermore, among all the battery cells 13, if there is no battery cell that meets the condition that the voltage difference ΔVi,i-1(tm) is above the positive-side threshold TH11 and the voltage difference ΔVi+1,i(tm) is below the negative-side threshold TH12, then step S105 is negated, and in step S106, m is increased by 1, and the process returns to step S101. Alternatively, among all the battery cells 13, if a certain battery cell 13 meets the condition that the voltage difference ΔVi,i-1(tm) is above the positive-side threshold TH11 and the voltage difference ΔVi+1,i(tm) is below the negative-side threshold TH12, then step S105 is affirmed, and the process proceeds to step S107.

[0069] At this time, if no internal short circuit abnormality occurs in the battery cell 13 that is the target cell, then the condition of step S105 is not met for the battery cell 13. Conversely, if an internal short circuit abnormality occurs in the battery cell 13 that is the target cell, then the condition of step S105 is met for the battery cell 13.

[0070] In step S107, the battery cell 13 for which the condition of step S105 has been met is determined to have experienced an anomaly. That is, it is determined that a predetermined deviation has occurred in the voltage change VA(tm) relative to other battery cells 13 within one of the plurality of battery cells 13. Then, in step S108, the anomaly is notified. At this time, for example, the anomaly warning light can be illuminated to notify the vehicle user of the anomaly. Additionally, diagnostic information indicating the anomaly can be stored in a memory.

[0071] In battery module 11, the first cell C1 with the lowest potential and the nth cell Cn with the highest potential are adjacent cells only on one side of the cell series path. Therefore, for each of these cells C1 and Cn, the processing of steps S104 and S105 can be modified as follows.

[0072] When the first cell C1 is selected as the target cell, the battery ECU 20 combines the first cell C1 with the two battery cells 13 that serve as its positive electrode side, and calculates two voltage differences ΔV(tm). Specifically, the difference in the absolute value of the voltage change VA(tm) between the first cell C1 and the nth cell Cn is calculated as the voltage difference ΔV1n(tm), and the difference in the absolute value of the voltage change VA(tm) between the first cell C1 and the second cell C2 is calculated as the voltage difference ΔV21(tm). Furthermore, it is determined whether the voltage difference ΔV1n(tm) is above the positive side threshold TH11, and whether the voltage difference ΔV21(tm) is below the negative side threshold TH12. Furthermore, the voltage difference ΔV1n(tm) is the cell order in which the subtraction operation of the voltage difference ΔV21(tm) is performed, and is obtained by subtracting the voltage change VAn of the nth cell Cn on the high-potential side from the voltage change VA1 of the first cell C1 on the low-potential side in the battery module 11. Assuming the first cell C1 is an abnormal cell, the voltage difference ΔV1n(tm) is calculated as a positive value. Furthermore, if the voltage difference ΔV1n(tm) is above the positive-side threshold TH11 and the voltage difference ΔV21(tm) is below the negative-side threshold TH12, then the first cell C1 is determined to be abnormal.

[0073] Here, the second cell C2 and the nth cell Cn are equivalent to the "first non-target cell" and the "second non-target cell," respectively. Furthermore, the voltage difference ΔV(tm) can be calculated by other combinations, such as setting the second cell C2 and the third cell C3 as the "first non-target cell" and the "second non-target cell."

[0074] Furthermore, when the nth cell Cn is considered as the target cell, the battery ECU 20 combines the nth cell Cn with the two battery cells 13 that serve as its negative electrode side, and calculates two voltage differences ΔV(tm). Specifically, the difference in the absolute value of the voltage change VA(tm) between the nth cell Cn and the (n-1)th cell Cn-1 is calculated as the voltage difference ΔVn,n-1(tm), and the difference in the absolute value of the voltage change VA(tm) between the first cell C1 and the nth cell Cn is calculated as the voltage difference ΔV1n(tm). It is then determined whether the voltage difference ΔVn,n-1(tm) is above the positive side threshold TH11, and whether the voltage difference ΔV1n(tm) is below the negative side threshold TH12. Furthermore, if the voltage difference ΔVn,n-1(tm) is above the positive threshold TH11 and the voltage difference ΔV1n(tm) is below the negative threshold TH12, then the nth cell Cn is determined to be abnormal.

[0075] Here, the first cell C1 and the (n-1)th cell Cn-1 are equivalent to the "first non-target cell" and the "second non-target cell". In addition, the voltage difference ΔV(tm) can also be calculated by setting the (n-2)th cell Cn-2 and the (n-1)th cell Cn-1 as the "first non-target cell" and the "second non-target cell", respectively.

[0076] For all battery cells 13 in battery module 11, it is also possible to combine the target cell and two battery cells 13 (non-target cells) on either the positive or negative side of the target cell separately, calculate two voltage differences ΔV(tm), and perform anomaly detection on the target cell based on these two voltage differences ΔV(tm). In short, the following structure is sufficient: for the target cell, calculate the difference between the absolute values ​​of the voltage changes VA(tm) with two or more different battery cells 13 as the voltage difference ΔV(tm), and use this voltage difference ΔV(tm) as a parameter for anomaly determination.

[0077] Regarding the combination of battery cells 13 for calculating the voltage difference ΔV(tm), the target battery cell 13 can be combined with adjacent cells and other battery cells 13, or it can be combined with two other battery cells 13. For example, it can be configured such that, for the target battery cell, two battery cells 13 separated by one cell are combined, or two battery cells 13 separated by two cells are combined, to calculate the voltage difference ΔV(tm). When the target battery cell is cell C3, cell C1 and cell C3 can be combined to calculate the voltage difference ΔV(tm), and cell C3 and cell C5 can be combined to calculate the voltage difference ΔV(tm).

[0078] It can also be like Figure 6 That would change the above Figure 5 Anomaly detection and handling. Figure 6 The processing has been changed. Figure 5 As part of the processing, steps S201 to S203 are related to Figure 5 Different treatments.

[0079] exist Figure 6 As explained above, in steps S101 to S103, the cell voltage V(tm) of each battery cell 13 at time tm is obtained, the voltage change VA(tm) from the initial value is calculated, and the absolute value of the voltage change VA(tm) is calculated. Furthermore, in step S104, the difference in the absolute values ​​of the voltage changes VA(tm) of two adjacent cells in the series path is calculated as the voltage difference ΔVi+1,i(tm).

[0080] Then, in step S201, the absolute value of the voltage difference ΔVi+1,i(tm) is calculated. In the next step S202, the absolute values ​​of the voltage difference ΔVi+1,i(tm) calculated at each time tm are accumulated, and the cumulative difference value QAi+1,i(tm) is calculated. At this time, the absolute value of the voltage difference ΔVi,i-1(tm) between the target cell and the first adjacent cell (the first non-target cell) is accumulated at a predetermined period, and the cumulative difference value QAi,i-1(tm) is calculated. In addition, the absolute value of the voltage difference ΔVi+1,i(tm) between the target cell and the second adjacent cell (the second non-target cell) is accumulated at a predetermined period, and the cumulative difference value QAi+1,i(tm) is calculated. Furthermore, the cumulative difference value QAi,i-1(tm) is equivalent to the "first cumulative difference value", and the cumulative difference value QAi+1,i(tm) is equivalent to the "second cumulative difference value".

[0081] Next, in step S203, each battery cell 13 (cells C1~Cn) is treated as a target cell, and for each target cell, the cumulative difference value QAi,i-1(tm) and QAi+1,i(tm) are compared with a predetermined threshold. Specifically, it is determined whether the cumulative difference value QAi,i-1(tm) calculated in the combination of the target cell and the first adjacent cell on the negative electrode side is greater than or equal to the predetermined threshold TH2, and it is also determined whether the cumulative difference value QAi+1,i(tm) calculated in the combination of the target cell and the second adjacent cell on the positive electrode side is greater than or equal to the threshold TH2.

[0082] Furthermore, among all the battery cells 13, if there is no battery cell that meets the condition that the cumulative difference value QAi,i-1(tm) is above the threshold TH2 and the cumulative difference value QAi+1,i(tm) is above the threshold TH2, then step S203 is negated, and after incrementing m by 1 in step S106, the process returns to step S101. Additionally, among all the battery cells 13, if a certain battery cell 13 meets the condition that the cumulative difference value QAi,i-1(tm) is above the threshold TH2 and the cumulative difference value QAi+1,i(tm) is above the threshold TH2, then step S203 is affirmed, and the process proceeds to step S107.

[0083] In step S107, the battery cell 13 for which the condition of step S203 has been met is determined to have experienced an abnormality. Subsequently, in step S108, the abnormality is notified.

[0084] Figure 7 This is a timing diagram used to specifically illustrate the abnormal detection and handling process of battery cell 13. Figure 7The diagram shows the changes of parameters (a) to (e) with respect to time. Here, it is assumed that repeated discharging and charging occur within battery module 11. Furthermore, the diagram illustrates three battery cells 13 (cells C1 to C3), with one (here, the second cell C2) designated as an abnormal cell experiencing an internal short circuit. Changes in cell voltage and other parameters are shown in time series for both normal cells (other than the second cell) and the abnormal cell. Figure 7 In the diagram, a dashed line represents the cell voltage V and voltage change VA of a normal cell, while a solid line represents the cell voltage V and voltage change VA of an abnormal cell.

[0085] exist Figure 7 During discharge, the cell voltage V of each battery cell 13 decreases, and during charging, the cell voltage V of each battery cell 13 increases. At this time, the voltage change VA differs between normal and abnormal cells. Therefore, regarding the abnormal cell (cell C2), the difference in voltage change VA with its adjacent cell on the negative side (voltage difference ΔV21) and the difference in voltage change VA with its adjacent cell on the positive side (voltage difference ΔV32) become values ​​of different polarities, and these differences are larger than the differences calculated between normal cells excluding the abnormal cell. Furthermore, as the internal short circuit in the abnormal cell gradually progresses, the voltage differences ΔV21 and ΔV32 between normal and abnormal cells gradually increase.

[0086] exist Figure 7 At time t1, the voltage difference ΔV21 is above the positive threshold TH11 and the voltage difference ΔV32 is below the negative threshold TH12. Therefore, the voltage change VA in the second cell C2 is considered to have deviated from the specified value relative to the other battery cells 13, and the second cell C2 is determined to be abnormal. In this case, the change in diffusion resistance occurs slowly when the battery cell 13 has an internal short circuit, and the voltage change in the abnormal cell is small. However, by comparing the cells relative to each other, the difference between the abnormal cell and the normal cell can be identified early.

[0087] In addition, Figure 7 In the example shown, the voltage difference ΔV43 is presented as a comparative example. The voltage difference ΔV43 is the difference between the voltage changes VA3 and VA4 of cells C3 and C4, both of which are normal cells, and is a value that does not exceed the thresholds TH11 and TH12. Although not illustrated, the changes in cell voltage V and voltage change VA of cell C4 are the same as those of cells C1 and C3.

[0088] Additionally, the cumulative voltage difference value QAi+1,i is calculated based on the cumulative voltage difference ΔVi+1,i for each battery cell. For example, in... Figure 6As explained, when the cumulative differential value QAi+1,i is used as a parameter for anomaly determination, at time t2, both the cumulative differential values ​​QA21 and QA32 exceed the threshold TH2, thus determining that the second cell C2 is abnormal. The cumulative differential value QAi+1,i is a value continuously calculated during the charging and discharging and non-powered periods of the battery module 11. By performing anomaly detection based on the cumulative differential value QAi+1,i, the accuracy of anomaly detection is improved. Even when equalization processing is performed in the battery module 11 to equalize the cell voltages of each battery cell 13, the difference in voltage change VA between the battery cells 13 can be accumulated, thus enabling earlier and more accurate detection of signs of anomalies.

[0089] In addition, Figure 7 In the diagram, the differential cumulative value QA43 is illustrated as a comparative example. The differential cumulative value QA43 is the cumulative value obtained by accumulating the voltage difference ΔV43 between cells C3 and C4, both of which are normal cells, and becomes a value that does not exceed the threshold TH2.

[0090] Based on the embodiment described above, the following excellent effects can be obtained.

[0091] In the battery module 11, anomaly detection of each battery cell 13 is performed relative to other battery cells 13, rather than as an individual cell anomaly detection. Therefore, even if voltage fluctuations or noise occur in the power supply path of the battery module 11, their effects can be suppressed and anomaly detection can be implemented. Furthermore, false detections caused by environmental conditions such as temperature are suppressed in the battery module 11. Therefore, in the event of an internal short circuit occurring as the initial stage of thermal runaway, even if the change in diffusion resistance in the battery cell 13 occurs slowly over time, the difference in voltage change VA(tm) between normal and abnormal cells can be detected early, thus appropriately detecting the abnormal cell. As a result, internal short circuit anomalies in battery cells 13 that are precursors to thermal runaway can be detected quickly and accurately.

[0092] For each target cell in battery module 11, the difference in voltage change VA(tm) between the target cell and two or more different battery cells 13 is calculated as a voltage difference value ΔV(tm). Based on this voltage difference value ΔV(tm), it is determined that a predetermined deviation has occurred in the voltage change VA(tm) relative to other battery cells 13 within the target cell. In this case, if each battery cell 13 in battery module 11 contains an abnormal cell, the voltage difference value ΔV(tm) calculated based on the voltage change VA(tm) of two normal cells differs from the voltage difference value ΔV(tm) calculated based on the voltage change VA(tm) of the two cells containing the abnormal cell. Therefore, abnormalities in the target cell can be appropriately detected.

[0093] For cell Ci, which is the target cell, anomaly detection is performed based on the following conditions: the polarities of the voltage differences ΔVi,i-1(tm) and ΔVi+1,i(tm) are opposite, and the absolute values ​​of the voltage differences ΔVi,i-1(tm) and ΔVi+1,i(tm) are both above a threshold. Therefore, abnormal cells can be appropriately detected while considering the differences in voltage changes between normal and abnormal cells.

[0094] In an abnormal cell experiencing an internal short circuit, the voltage change VA(tm) is greater than that in a normal cell, and this state persists through the charging and discharging of the battery module 11. Therefore, if the voltage difference ΔV(tm) is continuously monitored in each battery cell 13, the accuracy of internal short circuit detection is improved. In this regard, the accuracy of anomaly detection can be improved by using the cumulative value of the voltage difference ΔV(tm) calculated in each battery cell 13, i.e., the cumulative difference value QAi+1,i(tm), as a parameter for anomaly detection.

[0095] Hereinafter, other embodiments that differ from the first embodiment will be described, focusing on the differences from the first embodiment.

[0096] (Second Implementation) Here, firstly, using Figure 8 This section provides an overview of the cell anomaly detection method in this embodiment. Figure 8 In the battery module 11, five battery cells 13 (cells Ci-2, Ci-1, Ci, Ci+1, Ci+2) are shown. The voltage change in each cell Ci-2 to Ci+2 from the initial value during charging and discharging is VAi-2(tm) to VAi+2(tm). Furthermore, for every two adjacent battery cells 13 in the series path, the difference in the absolute value of the voltage change VA(tm), i.e., the voltage difference ΔV(tm), is sequentially ΔVi-1,i-2(tm), ΔVi,i-1(tm), ΔVi+1,i(tm), and ΔVi+2,i+1(tm) from the low potential side. Also, the difference in voltage difference ΔV(tm), i.e., the secondary difference ΔΔV(tm), is sequentially ΔΔVi,i-1,i-2(tm), ΔΔVi+1,i,i-1(tm), and ΔΔVi+2,i+1,i(tm) from the low potential side. The secondary difference ΔΔVi+1,i,i-1(tm) is calculated in the series-connected three-continuous battery cells 13 based on the voltage changes VAi--1(tm), VAi(tm), and VAi+1(tm) of each cell.

[0097] In this scenario, assuming that cell Ci among the five cells Ci-2 to Ci+2 is an abnormal cell, the voltage change VAi(tm) of cell Ci becomes a larger value on the negative side compared to the voltage changes VA(tm) of the other cells. Therefore, in the absolute difference of the voltage changes VA(tm) between any two cells, i.e., the voltage difference ΔV(tm), the voltage difference ΔVi,i-1(tm) between the abnormal cell (cell Ci) and its adjacent cell on the negative side is positive, the voltage difference ΔVi+1,i(tm) between the abnormal cell (cell Ci) and its adjacent cell on the positive side is negative, and the other voltage differences ΔVi-1,i-2(tm) and ΔVi+2,i+1(tm) are approximately zero.

[0098] Furthermore, ΔΔVi+1,i,i-1(tm) in each quadratic difference ΔΔV(tm) is obtained by subtracting the positive value ΔVi,i-1(tm) from the negative value ΔVi+1,i(tm). Therefore, the quadratic difference ΔΔVi+1,i,i-1(tm) becomes extremely negative, and its difference from other quadratic differences ΔΔV(tm) becomes significant. That is, the quadratic difference ΔΔVi+1,i,i-1(tm) becomes an indicator that clearly represents the deviation of the voltage change VAi(tm) in the abnormal cell. In this embodiment, the abnormal cell is appropriately detected by using the quadratic difference ΔΔV(tm) as a parameter for anomaly determination.

[0099] Figure 9 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20.

[0100] exist Figure 9 In step S301, the cell voltage V(tm) of each battery cell 13 at time tm is obtained. In step S302, the change in cell voltage from the initial value, i.e., the voltage change VA(tm), is calculated for each battery cell 13. In the following step S303, the absolute value of the voltage change VA(tm) is calculated. Furthermore, in step S304, the difference between the absolute values ​​of the voltage changes VA(tm) of two adjacent battery cells 13 in the series path is calculated as the voltage difference ΔVi+1,i(tm). Steps S301 to S304 are related to... Figure 5 The same process applies to steps S101 to S104.

[0101] Then, in step S305, the difference between two adjacent voltage differences ΔVi+1,i(tm) and ΔVi,i-1(tm) in the series arrangement of the battery cells 13 is calculated as a secondary difference ΔVi+1,i,i-1(tm). The secondary difference ΔVi+1,i,i-1(tm) can be calculated using the following equation (2).

[0102] ΔΔVi+1,i,i-1(tm)=ΔVi+1,i(tm)-ΔVi,i-1(tm)…(2) The second difference ΔΔVi+1,i,i-1(tm) is calculated for all combinations of battery cells 13 connected in series in battery module 11.

[0103] Next, in step S306, it is determined whether the quadratic difference ΔΔVi+1,i,i-1(tm) is below the specified negative-side threshold TH3. Furthermore, if the quadratic difference ΔΔVi+1,i,i-1(tm) is not below the negative-side threshold TH3 in all battery cells 13, then step S306 is negated, and after incrementing m by 1 in step S307, the process returns to step S301. Alternatively, if the quadratic difference ΔΔVi+1,i,i-1(tm) is below the negative-side threshold TH3 in any one of the battery cells 13, then step S306 is affirmed, and the process proceeds to step S308.

[0104] In step S308, the battery cell 13 for which the condition of step S306 has been met is determined to have an anomaly. That is, it is determined that a specified deviation has occurred in the voltage change VA(tm) relative to other battery cells 13 in the target cell. Then, in step S309, the anomaly is notified.

[0105] It can also be like Figure 10 That would change the above Figure 9 Anomaly detection and handling. Figure 10 The processing has been changed. Figure 9 As part of the processing, steps S401 to S402 are related to Figure 9 Different treatments.

[0106] exist Figure 10 As explained above, in steps S301 to S304, the voltage difference ΔVi+1,i(tm) is calculated based on time tm, and in step S305, the secondary difference ΔΔVi+1,i,i-1(tm) is calculated.

[0107] Then, in step S401, the quadratic difference ΔΔVi+1,i,i-1(tm) calculated over time tm is accumulated, and the cumulative quadratic difference value QBi+1,i,i-1(tm) is calculated. Additionally, in step S402, for all combinations of series-connected triple-continuous battery cells 13 in battery module 11, it is determined whether the cumulative quadratic difference value QBi+1,i,i-1(tm) is above a predetermined threshold TH4. The threshold TH4 is a positive threshold.

[0108] Furthermore, if the quadratic difference ΔΔVi+1,i,i-1(tm) in all battery cells 13 does not exceed the threshold TH4, then step S402 is negated, and after increasing m by 1 in step S307, the process returns to step S301. Alternatively, if the quadratic difference ΔΔVi+1,i,i-1(tm) in any one of the battery cells 13 exceeds the threshold TH4, then step S402 is affirmed, and the process proceeds to step S308. In step S308, an anomaly is determined in battery cell 13, and in step S309, the anomaly is notified.

[0109] In a series-connected triple-cell battery cell 13, when the central cell 13 (cell Ci) becomes an abnormal cell due to an internal short circuit, the voltage change VAi(tm) of the abnormal cell is larger than that of a normal cell. The voltage difference ΔVi,i-1(tm) calculated between the abnormal cell and its adjacent negative cell is opposite in polarity to the voltage difference ΔVi+1,i(tm) calculated between the abnormal cell and its adjacent positive cell. Therefore, the quadratic difference ΔΔVi+1,i,i-1(tm), calculated as the difference between the voltage difference ΔVi,i-1(tm) and the voltage difference ΔVi+1,i(tm) in the series-connected triple-cell battery cell 13, clearly indicates the deviation of the voltage change VAi(tm) in the abnormal cell. In this regard, abnormal cells can be properly detected by using the quadratic difference ΔΔVi+1,i,i-1(tm) calculated based on the voltage change VA(tm) of the series-connected three continuous battery cells 13 as a parameter for anomaly determination.

[0110] When an internal short circuit occurs in an abnormal cell, the voltage difference ΔVi,i-1(tm) calculated between the abnormal cell and its adjacent negative cell continues to have opposite polarities to the voltage difference ΔVi+1,i(tm) calculated between the abnormal cell and its adjacent positive cell. Therefore, by continuously monitoring the difference between these voltage differences ΔVi,i-1(tm) and ΔVi+1,i(tm), i.e., the secondary difference ΔΔVi+1,i,i-1(tm), the accuracy of internal short circuit detection is improved. In this regard, the accuracy of anomaly detection can be improved by using the cumulative value of the secondary difference ΔΔVi+1,i,i-1(tm), i.e., the cumulative value QBi+1,i(tm), as a parameter for anomaly determination.

[0111] (Third Implementation) In this embodiment, the degree of deviation of the voltage change VA(tm) of the target cell relative to the average value of the voltage change VA(tm) of each battery cell 13 is calculated as the degree of deviation in the battery module 11. Based on the degree of deviation, it is determined whether a predetermined deviation has occurred in the voltage change VA(tm) of the target cell relative to the battery cells 13 other than the target cell.

[0112] Furthermore, if an internal short circuit occurs in a battery cell 13 within the battery module 11, the average voltage change VA(tm) of each battery cell 13 will change due to this effect. In this case, when performing anomaly detection on the target cell based on the degree of deviation of the target cell's voltage change VA(tm) from the average voltage change VA(tm) of each battery cell 13, there is a concern about a decrease in detection accuracy. Therefore, in this embodiment, the degree of deviation of the target cell's voltage change VA(tm) is calculated based on the average value of the voltage change VA(tm) of the remaining battery cells 13 (excluding the target cell) in the battery module 11, and anomaly detection is performed based on this degree of deviation.

[0113] Figure 11 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20.

[0114] exist Figure 11 In step S501, the cell voltage V(tm) of each battery cell 13 at time tm is obtained. In step S502, for each battery cell 13, the change in cell voltage from the initial value, i.e., the voltage change VA(tm), is calculated. Furthermore, steps S501 to S502 are related to... Figure 5 The same process applies to steps S101 to S102.

[0115] Next, in step S503, the average value μi(tm) of the voltage change VA(tm) is calculated for each battery cell 13 of the battery module 11. In this embodiment, when each battery cell 13 of the battery module 11 is considered as an object cell for anomaly detection, the voltage change VA(tm) of the battery cell 13 that is the object cell is not used; in other words, the average value μi(tm) of the voltage change VA(tm) is calculated except for the voltage change VA(tm) of the battery cell 13 that is the object cell. At this time, the average value μi(tm) is calculated separately for each voltage change VA(tm) of the combination of cell numbers.

[0116] For example, if the target cell is cell C1, the voltage change VA1(tm) of cell C1 is excluded, and the average value μ1(tm) is calculated based on the remaining voltage change VA(tm). Similarly, if the target cell is cell C2, the voltage change VA2(tm) of cell C2 is excluded, and the average value μ2(tm) is calculated based on the remaining voltage change VA(tm).

[0117] In step S504, the standard deviation σi(tm) of the voltage change VA(tm) is calculated for each battery cell 13 of the battery module 11. The standard deviation σi(tm) can be calculated by taking the square root of the variance of the average of the squares of the differences between the voltage change VAi(tm) of each battery cell 13 and the average value μi(tm). In this embodiment, the voltage change VA(tm) of the battery cell 13 that is the target cell is not used. Instead, the standard deviation σi(tm) of the voltage change VA(tm) of the remaining battery cells 13 other than the target cell is used and the average value μi(tm) calculated excluding the target cell is used to calculate the standard deviation σi(tm) of the voltage change VA(tm). At this time, the standard deviation σi(tm) is calculated based on the voltage change VA(tm) of each combination of cell numbers.

[0118] For example, when the target cell is cell C1, the standard deviation σ1(tm) is calculated using the voltage change VA(tm) excluding cell C1 and the average value μ1(tm) calculated excluding the voltage change VA1(tm) of cell C1. Similarly, when the target cell is cell C2, the standard deviation σ2(tm) is calculated using the voltage change VA(tm) excluding cell C2 and the average value μ2(tm) calculated excluding the voltage change VA2(tm) of cell C2.

[0119] In step S505, in each battery cell 13 that becomes the target cell, the absolute value of the difference between the voltage change VAi(tm) and the average value μi(tm) and the standard deviation σi(tm) are calculated as the deviation degree Ri(tm). The deviation degree Ri(tm) is calculated using the following equation (3).

[0120] Ri(tm)=|VAi(tm)-μi(tm)| / σi(tm)…(3) The degree of deviation Ri(tm) represents the extent to which the voltage change VA(tm) of the target cell deviates from the average voltage change VA(tm) of all battery cells 13. Furthermore, the degree of deviation Ri(tm) represents the proportion of the deviation (VAi(tm) - μi(tm)) in each battery cell 13 relative to the standard deviation σi(tm).

[0121] Next, in step S506, for all battery cells 13, it is determined whether the deviation degree Ri(tm) is above the predetermined threshold TH5. The threshold TH5 is the positive threshold. Furthermore, for all battery cells 13, if the deviation degree Ri(tm) is not above the threshold TH5, step S506 is negated, m is incremented by 1 in step S507, and the process returns to step S501. Conversely, if the deviation degree Ri(tm) in a certain battery cell 13 is above the threshold TH5, step S506 is affirmed, and the process proceeds to step S508. In step S508, battery cells 13 for which the conditions of step S506 are met are determined to have experienced an anomaly. Then, in step S509, an anomaly is notified.

[0122] It can also be like Figure 12 That would change the above Figure 11 Anomaly detection and handling. Figure 12 The processing has been changed. Figure 11 As part of the processing, steps S601 to S602 are related to Figure 11 Different treatments.

[0123] exist Figure 12 As explained above, in steps S501 to S502, the voltage change VA(tm) is calculated over time tm. Furthermore, in step S503, for each battery cell 13 of the battery module 11, the average value μi(tm) of the voltage change VA(tm) is calculated. In step S504, for each battery cell 13, the standard deviation σi(tm) of the voltage change VA(tm) is calculated. In step S505, for each battery cell 13 that is the target cell, the ratio of the absolute value of the difference between the voltage change VAi(tm) and the average value μi(tm) to the standard deviation σi(tm) is calculated as the deviation degree Ri(tm). In this case, as described above, the average value μi(tm) and standard deviation σi(tm) are calculated only for the battery cell 13 that is the target cell.

[0124] Next, in step S601, the cumulative deviation degree Ri(tm) is accumulated for each battery cell 13 to calculate the cumulative deviation value QCi(tm). In step S602, it is determined whether the cumulative deviation value QCi(tm) is above the specified threshold TH6. The threshold TH6 is the positive threshold.

[0125] Furthermore, if the cumulative deviation value QCi(tm) of all battery cells 13 does not exceed the threshold TH6, then step S602 is negated, and m is incremented by 1 in step S507 before returning to step S501. Conversely, if the cumulative deviation value QCi(tm) of any battery cell 13 exceeds the threshold TH6, then step S602 is affirmed, and the process proceeds to step S508. In step S508, an anomaly is determined for battery cell 13, and in step S509, the anomaly is notified.

[0126] Using the average voltage change VA(tm) of the remaining battery cells 13 (excluding the target cell) in the battery module 11 as a benchmark, the degree of deviation of the voltage change VA(tm) of the target cell is calculated, and this degree of deviation is used as an anomaly judgment parameter. This improves the accuracy of anomaly detection.

[0127] In an abnormal battery cell that has experienced an internal short circuit, the voltage change VA(tm) exhibits a significant deviation over time. Therefore, by using the cumulative value of the deviation Ri(tm), i.e., the cumulative deviation value QCi(tm), as a parameter for anomaly detection, the accuracy of anomaly detection can be improved. Furthermore, since the cumulative deviation value QCi(tm) is calculated based on the average value of the voltage change VA of battery cells 13 other than the target cell, the accuracy of anomaly detection can be further improved.

[0128] It can also be like Figure 13 That would change the above Figure 11 Anomaly detection and handling. Figure 13 The processing has been changed. Figure 11 As part of the processing, steps S611 to S613 are related to Figure 11 Different treatments.

[0129] exist Figure 13 In steps S501 to S504, by interacting with... Figure 11 Similarly, for each battery cell 13, the average value μi(tm) and standard deviation σi(tm) of the voltage change VA(tm) are calculated. In this case, as described above, the average value μi(tm) and standard deviation σi(tm) are calculated only for the battery cell 13 that will be the target cell.

[0130] Next, in step S611, the difference between the voltage change VAi(tm) and the average value μi(tm) in each battery cell 13 is calculated, i.e., the deviation Di(tm). At this time, for each battery cell 13 that is the target cell, the deviation Di(tm) is calculated using the average value μi(tm) of the voltage change VA(tm) of the battery cells 13 other than the target cell. In addition, in step S612, the absolute value of the deviation Di(tm) is accumulated to calculate the cumulative deviation value QDi(tm). Furthermore, step S611 corresponds to the "deviation calculation unit", and step S612 corresponds to the "cumulative value calculation unit".

[0131] In step S613, for each battery cell 13, it is determined whether the cumulative deviation value QDi(tm) is above the threshold (K×σi(tm)) formed by the product of the standard deviation σi(tm) and the specified coefficient K.

[0132] Furthermore, if the cumulative deviation value QDi(tm) of all battery cells 13 does not exceed the threshold, then step S613 is negated, and after incrementing m by 1 in step S507, the process returns to step S501. Conversely, if the cumulative deviation value QDi(tm) of any battery cell 13 exceeds the threshold, then step S613 is affirmed, and the process proceeds to step S508. In step S508, an anomaly is determined for battery cell 13, and in step S509, the anomaly is notified.

[0133] When an internal short circuit occurs in a battery cell 13 within the battery module 11, the average value μ(tm) and standard deviation σ(tm) of the voltage change VA(tm) of each battery cell 13 change due to this effect. In this case, when performing anomaly detection of the target cell by comparing the cumulative deviation QD(tm) of the voltage change VA(tm) of each battery cell 13 with a threshold set based on the standard deviation σ(tm), there is a concern about a decrease in detection accuracy. To address this, the deviation D(tm) is calculated using the average value μ(tm) of the voltage change VA(tm) of the remaining battery cells 13 (excluding the target cell), and the standard deviation σ(tm) is calculated using the voltage change VA(tm) of the remaining battery cells 13 (excluding the target cell). This improves the accuracy of anomaly detection.

[0134] (Fourth Implementation) In this embodiment, based on the voltage changes VAi+1(tm), VAi(tm), and VAi-1(tm) of each of the three consecutively connected battery cells 13 in the series, a secondary difference operation is performed to calculate the secondary difference ΔΔVi+1,i,i-1(tm). Furthermore, the degree of deviation of the secondary difference ΔΔVi+1,i,i-1(tm) from the average value is calculated, and the abnormality of the target battery cell is detected based on the degree of deviation.

[0135] In this situation, when an internal short circuit occurs in a battery cell 13 within the battery module 11, the voltage difference ΔV(tm) among the combinations of battery cells 13 containing the abnormal cell becomes different from the voltage difference ΔV(tm) among combinations of normal cells. Consequently, the average value of the secondary difference ΔV(tm) changes. In this case, the accuracy of the degree of deviation of the target cell's secondary difference ΔV(tm) from its average value decreases, raising concerns about reduced accuracy in anomaly detection based on the degree of deviation.

[0136] Therefore, in this embodiment, when calculating the secondary difference ΔΔV(tm) by treating each battery cell 13 of the battery module 11 as a designated cell, the secondary difference ΔΔV(tm) calculated using the target cell as the designated cell is taken as the specific secondary difference. Furthermore, the degree to which the specific secondary difference deviates from the average of the remaining secondary differences ΔΔV(tm) calculated by treating each battery cell 13 of the battery module 11 as a designated cell is calculated as the degree of deviation.

[0137] Figure 14 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20.

[0138] exist Figure 14 In step S701, the cell voltage V(tm) of each battery cell 13 at time tm is obtained. In step S702, the change in cell voltage from the initial value, i.e., the voltage change VA(tm), is calculated for each battery cell 13. In the following step S703, the absolute value of the voltage change VA(tm) is calculated. Furthermore, in step S704, the voltage difference ΔVi+1,i(tm) is calculated, and in step S705, the quadratic difference ΔΔVi+1,i,i-1(tm) is calculated. Furthermore, steps S701 to S705 are based on... Figure 9 The processing is based on steps S301 to S305.

[0139] In this embodiment, particularly in step S704, each battery cell 13 of the battery module 11 is designated as a specific cell, and the voltage difference ΔV(tm) between the designated cell and the first adjacent cell on the negative electrode side and the voltage difference ΔV(tm) between the designated cell and the second adjacent cell on the positive electrode side are calculated by subtraction operations in ascending or descending order of the cell series connection. Furthermore, in step S705, for all combinations of battery cells 13 where the designated cell is set as the central cell in a series three-continuous configuration, a secondary difference ΔΔVi+1,i,i-1(tm) is calculated.

[0140] Next, in step S706, the average value μAi(tm) of the quadratic difference ΔΔV(tm) calculated by treating each battery cell 13 of the battery module 11 as a designated cell is calculated. Specifically, in the quadratic difference ΔΔV(tm) calculated by treating each battery cell 13 as a designated cell, the quadratic difference ΔΔVi+1,i,i-1(tm) calculated by treating the target cell (cell Ci) as a designated cell, i.e., the specific quadratic difference, is removed, and the average value μAi(tm) of the quadratic difference ΔΔV(tm) is calculated. In step S706, multiple average values ​​μAi(tm) are calculated for all combinations of battery modules 11.

[0141] In step S707, the standard deviation σAi(tm) of the quadratic difference ΔΔV(tm) calculated by treating each battery cell 13 of the battery module 11 as a designated cell is calculated. The standard deviation σAi(tm) can be calculated by taking the square root of the variance of the difference between the quadratic difference ΔΔV(tm) calculated by treating each battery cell 13 as a designated cell and the average value μAi(tm). Specifically, when the quadratic difference ΔΔVi+1,i,i-1(tm) calculated by treating the target cell (cell Ci) as a designated cell is used as the specific quadratic difference, the standard deviation σAi(tm) of the quadratic difference ΔΔV(tm) is calculated using the quadratic difference ΔΔV(tm) other than the specific quadratic difference among the quadratic differences ΔΔV(tm) calculated by treating each battery cell 13 as a designated cell, and the average value μAi(tm) calculated excluding the specific quadratic difference. In step S707, multiple standard deviations σAi(tm) are calculated using all combinations of battery modules 11.

[0142] In step S708, the absolute value of the difference between the second difference ΔΔVi+1,i,i-1(tm) and the average value μAi(tm) in each battery cell 13 and the standard deviation σAi(tm) are calculated as the deviation degree RAi(tm). The deviation degree RAi(tm) is calculated using the following equation (4).

[0143] RAi(tm)=|ΔΔVi+1,i,i-1(tm)-μAi(tm)| / σAi(tm)…(4) In steps S706 to S708, when the quadratic difference ΔΔVi+1,i,i-1(tm) calculated by taking the target cell (cell Ci) as the designated cell is used as the specific quadratic difference, the degree of deviation of the specific quadratic difference relative to the average of the remaining quadratic differences ΔΔV(tm) other than the specific quadratic difference is calculated as the deviation degree RAI(tm).

[0144] Next, in step S709, for all battery cells 13, it is determined whether the deviation degree RAi(tm) is above the specified threshold TH7. The threshold TH7 is the positive side threshold.

[0145] Furthermore, if the deviation level RAi(tm) for all battery cells 13 does not exceed the threshold TH7, then step S709 is negated, and after incrementing m by 1 in step S710, the process returns to step S701. Conversely, if the deviation level RAi(tm) for a particular battery cell 13 exceeds the threshold TH7, then step S709 is affirmed, and the process proceeds to step S711. In step S711, battery cells 13 for which the conditions of step S709 are met are determined to have experienced an anomaly. Then, in step S712, an anomaly is notified.

[0146] Using the average of the remaining quadratic differences ΔΔV(tm) calculated by treating each battery cell individually as a specified cell, excluding the specific quadratic difference (the quadratic difference calculated with the target cell as the specified cell), the degree of deviation of the specific quadratic difference is calculated as a benchmark, and this degree of deviation is used as an anomaly judgment parameter. Here, it is assumed that when the target cell is an abnormal cell, the specific quadratic difference is the quadratic difference ΔΔV(tm) affected by the voltage change VA(tm) of the abnormal cell. By excluding this specific quadratic difference, the average value μi(tm) and standard deviation σi(tm) are calculated, thereby improving the accuracy of anomaly detection.

[0147] It can also be like Figure 15 That would change the above Figure 14 Anomaly detection and handling. Figure 15 The processing has been changed. Figure 14 As part of the processing, Figure 14 Steps S706 to S707 are referred to as steps S706a to S707a.

[0148] In steps S706a to S707a, the degree of deviation of a specific quadratic difference relative to the average of the remaining quadratic differences ΔΔV(tm) calculated by treating each battery cell 13 as a designated cell, excluding the specific quadratic difference and the quadratic differences calculated by treating the first adjacent cell and the second adjacent cell as designated cells, is calculated and defined as the degree of deviation RAi(tm).

[0149] In addition, for convenience, the specific quadratic difference (the quadratic difference ΔΔVi+1,i,i-1(tm) calculated with the target cell as the specified cell) and the quadratic differences ΔΔVi,i-1,i-2(tm) and ΔΔVi+2,i+1,i(tm) calculated with the first adjacent cell and the second adjacent cell as specified cells respectively are collectively referred to as "specific quadratic difference group X".

[0150] In step S706a, except for the specific set of quadratic differences X (ΔΔVi,i-1,i-2(tm), ΔΔVi+1,i,i-1(tm), ΔΔVi+2,i+1,i(tm)) calculated for each battery cell 13 as a designated cell, the average value μAi(tm) of the quadratic difference ΔΔV(tm) is calculated.

[0151] In step S707a, the standard deviation σAi(tm) of the quadratic difference ΔΔV(tm) is calculated using the quadratic difference ΔΔV(tm) calculated by treating each battery cell 13 as a designated cell, excluding the quadratic difference set X (ΔΔVi,i-1,i-2(tm), ΔΔVi+1,i,i-1(tm), ΔΔVi+2,i+1,i(tm)), and the average value μAi(tm) calculated excluding the specific quadratic difference set X.

[0152] Then, similarly as above, the deviation degree RAi(tm) is calculated, and based on the deviation degree RAi(tm), anomaly determination is performed for each battery cell 13 (steps S708, S709).

[0153] Assuming the target cell is an abnormal cell, a specific set of quadratic differences X (the specific quadratic difference and the quadratic difference calculated by taking the first adjacent cell and the second adjacent cell as the specified cells) is a quadratic difference ΔΔV(tm) affected by the change in voltage VA(tm) of the abnormal cell. By excluding this specific set of quadratic differences X, the average value μi(tm) and standard deviation σi(tm) can be calculated, thereby improving the accuracy of anomaly detection.

[0154] It can also be like Figure 16 That would change the above Figure 14Anomaly detection and handling. Figure 16 The processing has been changed. Figure 14 As part of the processing, steps S801 to S802 are related to Figure 14 Different treatments.

[0155] exist Figure 16 In step S801, after calculating the deviation degree RAi(tm), the deviation degree RAi(tm) calculated over time tm is accumulated, and the cumulative deviation degree value QEi(tm) is calculated. In step S802, it is determined whether the cumulative deviation degree value QEi(tm) is above the specified threshold TH8. The threshold TH8 is the positive threshold.

[0156] Furthermore, if the cumulative deviation value QEi(tm) of all battery cells 13 does not exceed the threshold TH8, then step S802 is negated, m is incremented by 1 in step S710, and the process returns to step S701. Alternatively, if the cumulative deviation value QEi(tm) of any battery cell 13 exceeds the threshold TH8, then step S802 is affirmed, and the process proceeds to step S711. In step S711, an anomaly is determined to have occurred in battery cell 13, and in step S712, an anomaly is notified.

[0157] (Fifth Implementation) In this embodiment, in the structure where each battery cell 13 of the battery pack B is divided into multiple battery cell groups, as the first stage of anomaly detection, a temporary anomaly determination is made for all battery cell groups, and as the second stage, a formal anomaly determination is made for the battery cell groups that have been temporarily determined to have anomalies.

[0158] Figure 17 This is a diagram illustrating the structure of the battery system in this embodiment. In the battery cell 10, multiple battery cells 13 connected in series form battery cell groups G1, G2, ..., Gn. Each battery cell group G1~Gn consists of three or more battery cells 13 connected in series. For example... Figure 1 Each battery module 11 shown can also be a battery cell group G1~Gn. Figure 17 In the battery cell 10 shown, a monitoring device 12 is provided for each battery cell group G1 to Gn. However, it is also possible to provide one monitoring device 12 for two or more battery cell groups G1 to Gn.

[0159] In this embodiment, the battery ECU 20 includes an acquisition unit 51, a parameter change calculation unit 52, an anomaly temporary determination unit 53, an anomaly detection unit 54, and a voltage adjustment unit 55. The acquisition unit 51 acquires the cell voltage V (state detection value) of each battery cell 13 detected by the voltage sensor 31 (state detection unit) from each monitoring device 12 for each battery cell group G1-Gn. The parameter change calculation unit 52 calculates the change in the cell voltage V acquired by the acquisition unit 51 over a predetermined period, i.e., the voltage change VA (parameter change), for each battery cell group G1-Gn. The anomaly temporary determination unit 53 performs a temporary anomaly determination for each battery cell 13 based on the voltage change VA for each battery cell group G1-Gn. The anomaly detection unit 54, for each battery cell group G1-Gn including the battery cells 13 temporarily determined to be abnormal by the anomaly temporary determination unit 53, treats each battery cell 13 as a target cell for anomaly detection and performs anomaly detection on that target cell.

[0160] Furthermore, in this embodiment, as an execution condition for anomaly detection, a state is specified where the terminal voltages of each battery cell 13 in each battery cell group G1 to Gn are adjusted to the same voltage. When this execution condition is met, anomaly detection is performed on each battery cell 13 (the target cell). Specifically, the battery ECU 20 has an equalization function that adjusts the terminal voltages of each battery cell 13 to the same voltage. When the equalization process (cell balancing process) for each battery cell 13 is performed, anomaly detection is performed using the cell voltages of each battery cell 13. The voltage adjustment unit 55 performs voltage equalization by discharging or charging each battery cell 13. For example, if a deviation in the terminal voltages of each battery cell 13 occurs while the vehicle is stationary (parked), the voltage adjustment unit 55 performs equalization processing. Alternatively, the voltage adjustment unit 55 performs equalization processing periodically, i.e., at a predetermined cycle.

[0161] Figure 18 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20. Figure 18 Regarding what has already been stated Figure 5 A portion of the changes were made, regarding the Figure 5 For the same process, label the steps with the same numbers and simplify the description. Figure 18 The equalization process is performed on a per-cell basis. Furthermore, it is envisioned that equalization processing can be performed while the vehicle is stationary, ideally during the period when the vehicle's power switch is turned off. Figure 18 The processing.

[0162] exist Figure 18In step S901, it is determined whether equalization processing of each battery cell 13 has just been performed. For example, if equalization processing was performed while the vehicle is stationary, step S901 is confirmed if it is before the next vehicle start-up and before a predetermined time has elapsed since the execution of the equalization processing. Furthermore, if step S901 is confirmed, the process proceeds to the subsequent step S101.

[0163] In steps S101 and S102, the cell voltage V(tm) of each battery cell 13 is obtained, and for each battery cell 13, the change in cell voltage V(tm) from the initial value (initial voltage value) is calculated, i.e., the voltage change VA(tm) (VA(tm) = V(tm) - initial value).

[0164] Next, in step S902, based on the voltage change VA(tm) of each battery cell 13, a temporary determination of whether each battery cell 13 is abnormal is made. In the following step S903, it is determined whether the battery cell group being tested contains any battery cells 13 that have been temporarily determined to be abnormal. The processing in step S902 is equivalent to the first stage of abnormality detection. At this time, the voltage change VA(tm) of each battery cell 13 in the battery cell group is compared with each other. If there is a battery cell 13 with a large deviation, it is determined that the battery cell group containing that battery cell 13 is likely to be abnormal. Specifically, it can be that the average value of the voltage change VA(tm) of each battery cell 13 in the battery cell group is calculated, and the battery cell 13 whose voltage change VA(tm) deviates from the average value by more than a certain amount is temporarily determined to be a battery cell 13 that has generated an abnormality (abnormal cell).

[0165] Next, in steps S103 and S104, the absolute value of the voltage change VA(tm) is calculated, and for all battery cells 13, the difference in the absolute value of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as the voltage difference ΔVi+1,i(tm). In step S105, each battery cell 13 (cells C1 to Cn) is treated as a target cell, and for each target cell, the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm) are compared with a predetermined threshold. If a certain battery cell 13 among all battery cells 13 meets the condition that the voltage difference ΔVi,i-1(tm) is above the positive threshold TH11 and the voltage difference ΔVi+1,i(tm) is below the negative threshold TH12, then step S105 is confirmed, and the process proceeds to step S107. In step S107, battery cell 13 that meets the conditions of step S105 is determined to have generated an anomaly. The processing of steps S103 to S107 is equivalent to the second stage of anomaly detection.

[0166] exist Figure 18 In the anomaly detection process, as the second stage of anomaly detection (formal judgment), the anomaly detection of battery cell 13 is performed based on the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm). However, the formal judgment process can also be modified as follows.

[0167] (1) Alternatively, in each battery cell group, the cumulative difference values ​​QAi,i-1(tm) and QAi+1,i(tm) are calculated by accumulating the absolute values ​​of the voltage differences ΔVi+1,i(tm) calculated over time tm, and based on these cumulative difference values ​​QAi,i-1(tm) and QAi+1,i(tm), anomaly detection (formal judgment) of the battery cell 13 is performed (refer to...). Figure 6 Steps S202 and S203).

[0168] (2) Alternatively, in each battery cell group, the absolute difference of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as the voltage difference ΔVi+1,i(tm), and the difference between two adjacent voltage differences ΔVi+1,i(tm) and ΔVi,i-1(tm) in the series arrangement of battery cells 13 is calculated as the secondary difference ΔΔVi+1,i,i-1(tm). Based on the secondary difference ΔΔVi+1,i,i-1(tm), anomaly detection (formal judgment) of battery cell 13 is performed (refer to...). Figure 9 Steps S304~S306).

[0169] (3) Alternatively, in each battery cell group, a cumulative value of the secondary difference QBi+1,i,i-1(tm) can be calculated by accumulating the secondary difference ΔΔVi+1,i,i-1(tm) calculated over time tm, and anomaly detection (formal judgment) of battery cell 13 can be performed based on this cumulative value of the secondary difference QBi+1,i,i-1(tm) (refer to...). Figure 10 Steps S401 and S402).

[0170] Figure 19 This is a time-series diagram showing the voltage changes of the cells in battery pack B. Figure 19 For ease of explanation, the changes in cell voltage of the two battery cells 13 are shown.

[0171] exist Figure 19In this diagram, timings t1 and t2 are the timings for equalization processing of each battery cell 13. After these timings t1 and t2, the cell voltage of each battery cell 13 is equalized. For example, when equalization processing is performed at timing t1, the period TA is the period immediately after equalization and before discharge to the vehicle's rotating motor, etc., is performed while the vehicle is running. Abnormal detection processing of battery cell 13 can begin with the period TA as a base point.

[0172] In the above embodiment, the voltage adjustment unit 55 performs voltage equalization processing on each battery cell 13 as a voltage adjustment process, but this can be modified. For example, as a voltage adjustment process, the voltage adjustment unit 55 can also charge the battery pack B (each battery module 11) until it is fully charged via the charging device. In this case, the battery ECU 20 may perform anomaly detection (formal determination) on the target cell based on the condition that a predetermined time has elapsed after the battery pack B (each battery module 11) has been fully charged. The charging device may be an external charging device installed outside the vehicle.

[0173] Furthermore, the structure that uses the condition of adjusting the terminal voltages of each battery cell 13 to the same voltage as the execution condition for anomaly detection can also be applied to anomaly detection processing other than this embodiment. For example, it can also be configured such that, in each of the anomaly detection processes in the first to fourth embodiments, anomaly detection of each battery cell 13 is performed based on voltage difference ΔV, secondary difference ΔΔV, etc., as a condition that the terminal voltages of each battery cell 13 are adjusted to the same voltage.

[0174] In the embodiment described above, during the anomaly detection of each battery cell 13, by combining temporary determination and formal determination as described above, it is possible to suppress the application of excessive computational load and achieve appropriate anomaly detection.

[0175] Furthermore, the configuration is such that anomaly detection is performed on the target cell under the condition that the terminal voltages of each battery cell 13 are adjusted to the same voltage through equalization processing or the like. In this case, false anomaly detections can be suppressed for each battery cell 13.

[0176] (Sixth Implementation Method) In this embodiment, the above-described Figure 11 The anomaly detection process is structured to perform anomaly detection in two stages. Furthermore, in this embodiment, instead of performing temporary anomaly determination for each battery cell group in battery pack B, temporary anomaly determination is performed for each individual battery cell 13.

[0177] Figure 20This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20. Figure 20 Regarding what has already been stated Figure 11 Part of it has been changed, regarding the Figure 11 The same process is labeled with the same step numbers and the description is simplified.

[0178] exist Figure 20 In steps S501 and S502, the cell voltage V(tm) of each battery cell 13 is obtained, and the change in cell voltage from the initial value, i.e., the voltage change VA(tm), is calculated for each battery cell 13.

[0179] Next, in step S1001, a temporary determination of whether each battery cell 13 is abnormal is made based on the voltage change VA(tm) of each battery cell 13. At this time, if there is a battery cell 13 whose voltage change VA(tm) difference relative to the average voltage change VA(tm) of each battery cell 13 is greater than or equal to a predetermined value, then that battery cell 13 is temporarily determined to be an abnormal cell. In this embodiment, step S1001 corresponds to an "abnormal temporary determination unit".

[0180] Subsequently, in steps S503-S506, each of the battery cells 13 temporarily identified as abnormal in step S1001 is used as the target cell for anomaly detection, and it is determined whether a specified deviation has occurred in that target cell. In this case, in steps S503-S506, not all the battery cells 13 of the battery module 11 are used as the target cells for anomaly detection, but only the battery cells 13 temporarily identified as abnormal are used as the target cells for anomaly detection.

[0181] Specifically, in step S503, the average value μi(tm) of the voltage change VA(tm) is calculated for all battery cells 13 in battery module 11. At this time, when each battery cell 13 temporarily identified as an abnormal cell in step S1001 is treated as a target cell, the voltage change VA(tm) of the target cell 13 is not used; in other words, the average value μi(tm) of the voltage change VA(tm) is calculated excluding the target cell. In step S504, the voltage change VA(tm) of the remaining battery cells 13 (excluding the target cell) is not used; instead, the standard deviation σi(tm) of the voltage change VA(tm) is calculated using the voltage change VA(tm) of each remaining battery cell 13 (excluding the target cell) and the average value μi(tm) calculated excluding the target cell.

[0182] In step S505, for each battery cell 13 that is the target cell, the ratio of the absolute value of the difference between the voltage change VAi(tm) and the average value μi(tm) to the standard deviation σi(tm) is calculated as the deviation degree Ri(tm). Then, in step S506, anomaly detection is performed for each battery cell 13 that is the target of anomaly detection based on the deviation degree Ri(tm).

[0183] Although the illustration is omitted, it is about Figure 12 , Figure 13 Anomaly detection and handling can also be done as described above. Figure 20 Therefore, based on the voltage change VA(tm) of each battery cell 13, a temporary determination is made as to whether there is an abnormality in each battery cell 13 (same as step S1001). That is, in these... Figure 12 , Figure 13 In the abnormal detection process, the processing of step S1001 can be performed between step S502 and step S503.

[0184] In the embodiment described above, among the plurality of battery cells 13, a temporary determination of abnormal cells is made based on the difference between the value of the voltage change VA (parameter change) and the average value of the voltage change VA. Each of the temporarily determined abnormal cells is then designated as a target cell, and a formal determination of abnormality is performed on each target cell. This suppresses the application of excessive computational load and enables appropriate anomaly detection.

[0185] (Seventh Implementation) In this embodiment, in the structure where each battery cell 13 of the battery pack B is divided into multiple battery cell groups, as the first stage of anomaly detection, a temporary anomaly determination is made for all battery cell groups, and as the second stage, a formal anomaly determination is made for the battery cell groups that are temporarily determined to have anomalies.

[0186] Figure 21 This diagram illustrates the structure of the battery system in this embodiment. In the battery cell 10, multiple battery cells 13 connected in series form battery cell groups G1, G2, ..., Gn. Each battery cell group G1 to Gn consists of three or more battery cells 13 connected in series. For example... Figure 1 Each battery module 11 shown can be a battery cell group G1~Gn. Figure 21 In the battery cell 10 shown, a monitoring device 12 is provided for each battery cell group G1 to Gn. However, it is also possible to provide one monitoring device 12 for two or more battery cell groups G1 to Gn.

[0187] In this embodiment, the battery ECU20 includes an acquisition unit 61, a parameter change calculation unit 62, a primary difference calculation unit 63, an anomaly temporary determination unit 64, a secondary difference calculation unit 65, and an anomaly detection unit 66.

[0188] The acquisition unit 61 acquires the cell voltage V (state detection value) of each battery cell 13 detected by the voltage sensor 31 (state detection unit). The parameter change calculation unit 62 calculates the change in the cell voltage V acquired by the acquisition unit 51 over a specified period, i.e., the voltage change VA (parameter change).

[0189] The primary difference calculation unit 63 calculates the difference ΔV, i.e., the voltage difference VA (parameter change) between each battery cell group G1 to Gn and the target cell that is the object of anomaly detection, as the primary difference. The temporary anomaly determination unit 64 performs a temporary determination on each battery cell group G1 to Gn based on the voltage difference ΔV calculated by the primary difference calculation unit 63 to determine whether there is an anomaly in each battery cell 13.

[0190] The secondary difference calculation unit 65, in the battery cell groups G1~Gn among the multiple battery cell groups that are temporarily determined to have an anomaly by the temporary anomaly determination unit 64, designates the battery cell 13 adjacent to the negative electrode side of the target cell as the first adjacent cell, and the battery cell 13 adjacent to the positive electrode side of the target cell as the second adjacent cell. It calculates the difference in the amount of change between these adjacent cells and the target cell using subtraction operations in ascending or descending order of the cell series connection, respectively, as the first difference and the second difference, and calculates the difference between the first difference and the second difference as the secondary difference ΔΔV. Furthermore, the calculation order of the voltage difference ΔV and the secondary difference ΔΔV is as follows: Figure 8 As explained, the anomaly detection unit 66 determines that a predetermined deviation, which is a cell anomaly, has occurred in the target cell if the secondary difference ΔΔV is greater than a predetermined threshold.

[0191] Figure 22 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20. Figure 22 In the anomaly detection and handling, the process for temporary anomaly determination is as described above. Figure 5 The processing involves temporarily determining the anomalies of each battery cell 13 based on the voltage difference ΔV of the battery cell 13, and, as the formal anomaly determination process, the previously described procedures are executed. Figure 9 The processing involves formally determining the anomalies of each battery cell 13 based on the quadratic difference ΔΔV of the battery cell 13. The following shows... Figure 5 Anomaly detection and handling Figure 9 The corresponding anomaly detection and handling procedures are explained. Figure 22 The processing.

[0192] exist Figure 22 In steps S1101 and S1102, the cell voltage V(tm) of each battery cell 13 is obtained, and the change in cell voltage V(tm) from the initial value (initial voltage value) is calculated for each battery cell 13, i.e., the voltage change VA(tm) (VA(tm) = V(tm) - initial value). Furthermore, in steps S1103 and S1104, the absolute value of the voltage change VA(tm) is calculated, and for all battery cells 13, the difference in the absolute value of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as the voltage difference ΔVi+1,i(tm). Then, in step S1105, each battery cell 13 (cells C1 to Cn) is treated as a target cell, and for each target cell, the voltage difference ΔVi,i-1(tm) and ΔVi+1,i(tm) are compared with a predetermined threshold. The above steps S1101~S1105 are related to Figure 5 The same process applies to steps S101 to S105.

[0193] Furthermore, in the battery cell group being tested, if step S1105 is denied in all battery cells 13, the processing for this battery cell group is stopped at that point in time, and the anomaly detection process ends. Alternatively, in the battery cell group being tested, if step S1105 is affirmed in a particular battery cell 13, the process proceeds to step S1106. In step S1106, for the battery cell 13 where the condition of step S1105 is met, an anomaly is temporarily determined to have occurred.

[0194] Next, in step S1107, the difference between two adjacent voltage differences ΔVi+1,i(tm) and ΔVi,i-1(tm) in the series arrangement of the battery cells 13 is calculated as a secondary difference ΔΔVi+1,i,i-1(tm). In step S1108, it is determined whether the secondary difference ΔΔVi+1,i,i-1(tm) is below the predetermined negative-side threshold TH3. If step S1108 is affirmative, the process proceeds to step S1109.

[0195] In step S1109, the battery cell 13 for which the condition of step S1108 has been met is determined to have an anomaly. That is, it is determined that a specified deviation has occurred in the voltage change VA(tm) relative to other battery cells 13 within the target cell. The determination result of step S1109 is equivalent to the result of the formal determination. Then, in step S1110, the anomaly is notified. The aforementioned steps S1107~S1110 are related to... Figure 9The same processing steps S305, S306, S308, and S309 apply.

[0196] In the embodiment described above, anomaly detection of each battery cell 13 is performed in two stages. For each battery cell group, a temporary anomaly determination is made based on the voltage difference ΔV (first-order difference). Furthermore, for battery cell groups among multiple battery cell groups that are temporarily determined to have anomalies, a formal determination is made based on the second-order difference ΔΔV, indicating that a predetermined deviation has occurred in the battery cell 13. Here, comparing the anomaly determination based on the voltage difference ΔV (temporary determination) and the anomaly determination based on the second-order difference ΔΔV (formal determination), the computational load and detection accuracy required for anomaly detection are the same. In the anomaly determination based on the second-order difference ΔΔV, if... Figure 8 As explained, the computational load increases, but on the other hand, it enables a clear understanding of abnormal battery cells, thus allowing for high-precision detection of battery cell abnormalities.

[0197] Furthermore, in this embodiment, in the first stage (provisional determination), a low-load anomaly determination can be performed considering the large number of battery cell groups that are subject to anomaly determination, and in the second stage (formal determination), the accuracy of anomaly determination can be improved. In this case, not all battery cell groups (battery modules 11) in battery pack B may experience anomalies simultaneously, and appropriate anomaly detection can be performed sequentially starting from the battery cell groups including the battery cell 13 that has experienced anomalies.

[0198] Furthermore, in this embodiment, the processing for temporary anomaly determination and the processing for formal anomaly determination are combinations of processes with different processing loads, such as parameter calculation. The former is a relatively low-load process, while the latter is a relatively high-load process. Specifically, as the processing for temporary anomaly determination, the above-described... Figure 5 In the case of temporary anomaly determination of each battery cell 13 based on the voltage difference ΔV of the battery cell 13, the formal anomaly determination process can be any of the following: · Figure 6 The processing involves: detecting abnormalities in battery cell 13 based on the cumulative difference value QA obtained by accumulating the voltage difference ΔV of battery cell 13. · Figure 9 The processing is as follows: abnormal detection of battery cell 13 is performed based on the second difference ΔΔV of battery cell 13. · Figure 10 The processing involves detecting abnormalities in battery cell 13 by accumulating the cumulative value QB of the secondary difference ΔV obtained from accumulating the secondary difference of battery cell 13.

[0199] Alternatively, as a temporary exception handling measure, it can be implemented during the execution of the previously described steps. Figure 5 In the case of temporary anomaly determination of each battery cell 13 based on the voltage difference ΔV of the battery cell 13, the formal anomaly determination process can be any of the following: · Figure 11 The processing is as follows: taking the average value of the voltage change VA of the remaining battery cells 13 (excluding one battery cell 13 (the target cell)) as a benchmark, the deviation degree Ri of the voltage change VA of the target cell is calculated, and the abnormality detection of the battery cell 13 is performed based on the deviation degree Ri. · Figure 12 The processing involves: accumulating the deviation of the voltage change VA of the target cell whenever the deviation is calculated, and calculating the cumulative deviation value QCi; and performing abnormal detection of battery cell 13 based on the cumulative deviation value QCi. · Figure 13 The processing involves: accumulating the deviation Di of the voltage change VA of each battery cell 13 to calculate the cumulative deviation value QDi, and performing abnormal detection of the battery cell 13 based on the cumulative deviation value QDi. · Figure 14 , Figure 15 The processing involves: calculating the standard deviation σAi of the quadratic difference ΔV of each battery cell 13, calculating the degree of deviation RAi based on the standard deviation σAi, and performing abnormal detection of the battery cell 13 based on the degree of deviation RAi. · Figure 16 The processing involves accumulating the deviation degree RAi to calculate the cumulative deviation degree value QEi, and then performing abnormal detection of the battery cell 13 based on this cumulative deviation degree value QEi.

[0200] (Eighth Implementation Method) In this embodiment, the anomaly detection unit for detecting anomalies in each battery cell 13 includes: a first determination unit that determines whether a predetermined deviation has occurred in the target cell based on the primary difference of the voltage change VA, i.e., the voltage difference ΔV; and a second determination unit that determines whether a predetermined deviation has occurred in the target cell based on the secondary difference ΔV between the voltage differences ΔV of the two cells. Furthermore, these first and second determination units are used separately based on predetermined switching conditions. In this embodiment, the anomaly determination of each battery cell 13 based on the voltage difference ΔV is used as the first determination process, and the anomaly determination of each battery cell 13 based on the secondary difference ΔΔV is used as the second determination process.

[0201] The anomaly determination of each battery cell 13 based on the voltage difference ΔV is equivalent to performing anomaly detection of each battery cell 13 through low-load computational processing with relatively low computational load (first anomaly detection unit). Furthermore, the anomaly determination of each battery cell 13 based on the secondary difference ΔΔV is equivalent to performing anomaly detection of each battery cell 13 through high-load computational processing with relatively high computational load (second anomaly detection unit). Moreover, based on predetermined switching conditions, these anomaly detection processes are switched (switching execution unit).

[0202] Figure 23 This is a flowchart illustrating the steps of abnormal detection and processing of battery cell 13 in this embodiment, which is executed by battery ECU 20.

[0203] exist Figure 23 In step S1200, based on the specified switching conditions, it is determined whether to execute the first determination process or the second determination process. The determination in step S1200 can be performed as follows.

[0204] The battery ECU 20, regarding the battery module 11 being tested, determines whether an anomaly determination based on the voltage difference ΔV for each battery cell 13 has been performed and whether an anomaly has been determined through this determination. If no anomaly is determined based on the voltage difference ΔV, step S1200 is affirmed and the process proceeds to step S1201. Conversely, if an anomaly is determined based on the voltage difference ΔV, step S1200 is negated and the process proceeds to step S1202. In this case, determining an anomaly based on the voltage difference ΔV is equivalent to a switching condition for transitioning from the first determination process to the second determination process.

[0205] Alternatively, the battery ECU20 can be configured to determine that the battery module 11 is in a state where each battery cell 13 has deteriorated or the probability of deterioration is high (deterioration determination unit), and use this determination result as a switching condition. In this case, determining that the battery cell 13 has deteriorated or the probability of deterioration is high is equivalent to a switching condition for switching from the first determination process to the second determination process.

[0206] In cases where battery cells 13 have deteriorated or are highly likely to deteriorate, the state of health (SOH) of the battery cells 13 can be used as a criterion. The SOH of the battery cell 13 can be calculated as the ratio of the actual fully charged capacitance Cf [Ah] of the battery cell 13 to the reference fully charged capacitance Cf0 [Ah] of the battery cell 13 (SOH = Cf / Cf0). In this case, if the SOH is less than a specified value, an anomaly is determined through a first determination process; if the SOH is greater than a specified value, an anomaly is determined through a second determination process.

[0207] The condition of whether each battery cell 13 has deteriorated or is highly likely to deteriorate can also be determined based on the number of years since the vehicle was manufactured or the vehicle's mileage. For example, if the vehicle has been in use for more than 5 years or has traveled more than 100,000 km, it is determined that each battery cell 13 has deteriorated or is highly likely to deteriorate.

[0208] Furthermore, if the battery cell 13 is at a specified low temperature (e.g., below 10°C), lithium will deposit on the electrodes, and the battery cell 13 will deteriorate. Therefore, it is also possible to determine whether the battery cell 13 is in a state of deterioration or a state with a high probability of deterioration based on whether the battery cell 13 is at a low temperature.

[0209] In step S1201, anomaly determination is performed on each battery cell 13 based on the voltage difference ΔV. Furthermore, the processing in step S1201 is related to... Figure 5 The steps S101~S105 and S107 are the same, so the explanation is omitted here.

[0210] Additionally, in step S1202, anomaly determination is performed on each battery cell 13 based on the secondary difference ΔΔV of the battery cell 13. Furthermore, the processing in step S1202 is related to... Figure 9 The steps S301~S306 and S308 are the same, so the explanation is omitted here.

[0211] In the embodiment described above, two types of anomaly determinations can be performed during anomaly detection of each battery cell 13, and these anomaly determinations can be switched according to predetermined switching conditions. Here, in the case of long-term use of the battery cells 13, situations where cell anomalies are easily detected and situations where they are difficult to detect may occur during this period. In such cases, by switching the anomaly determination according to predetermined switching conditions, excessive computational load can be suppressed, and appropriate anomaly detection can be achieved.

[0212] Furthermore, in this embodiment, based on the degree of degradation of each battery cell 13, a first determination process based on the voltage difference ΔV (primary difference) and a second determination process based on the secondary difference ΔΔV are switched. In this case, by performing anomaly detection according to the degree of degradation of each battery cell 13, excessive computational load can be suppressed, and appropriate anomaly detection can be achieved.

[0213] exist Figure 23 In this context, the combination of the low-load computation processing performed in step S1201 and the high-load computation processing performed in step S1202 can also be... Figure 5 processing and Figure 9 Combinations other than those for processing. For example, it can also be configured to perform low-load computational processing. Figure 5 , Figure 6 , Figure 9 , Figure 10 The processing of either party, and, as a high-load computational process, execution. Figures 11-16 The handling of either party.

[0214] (Other implementation methods) For example, the above implementation method can be modified as follows.

[0215] • As an abnormality detection and handling function for battery cell 13, battery ECU20 can also perform Figure 24 The processing is shown.

[0216] exist Figure 24 In step S1301, the cell voltage V(tm) of each battery cell 13 at time tm is obtained. In step S1302, the change in cell voltage from the initial value, i.e., the voltage change VA(tm), is calculated for each battery cell 13. Furthermore, steps S1301 to S1302 are related to... Figure 5 The same process applies to steps S101 to S102.

[0217] Then, in step S1303, the correlation coefficient CR(tm) between adjacent battery cells 13 in the series path is calculated. The correlation coefficient CR(tm) is an indicator of the strength of the correlation (association) between two data points, and specifically, it can be calculated by the following equation (5).

[0218]

Mathematical Formula 1

[0219] In step S1303, the voltage changes VA of the target cell (cell Ci) over the past N times from the current time point are set as data x1, x2, ..., xN, and the voltage changes VA of the first adjacent cell (cell Ci-1) adjacent to the negative side of the target cell over the past N times from the current time point are set as data y1, y2, ..., yN, and the correlation coefficient CRi,i-1(tm) is calculated. Additionally, the voltage changes VA of the target cell (cell Ci) over the past N times from the current time point are set as data x1, x2, ..., xN, and the voltage changes VA of the second adjacent cell (cell Ci+1) adjacent to the positive side of the target cell over the past N times from the current time point are set as data y1, y2, ..., yN, and the correlation coefficient CRi+1,i(tm) is calculated. At this point, in the combination of the target cell and adjacent cells, if the target cell is an abnormal cell, the correlation between the two cells weakens, meaning the correlation coefficient CR(tm) decreases. Therefore, it can be determined that the cell is abnormal.

[0220] Next, in step S1304, it is determined whether the correlation coefficients CRi+1,i(tm) and CRi,i-1(tm) are both below the specified threshold TH9. The threshold TH9 is a value greater than 0 and less than 1.

[0221] Furthermore, among all battery cells 13, if there is no battery cell that meets the condition that both correlation coefficients CRi+1,i(tm) and CRi,i-1(tm) are below the threshold TH9, then step S1304 is negated, and in step S1305, m is increased by 1, and the process returns to step S1301. Alternatively, among all battery cells 13, if any battery cell 13 meets the condition that both correlation coefficients CRi+1,i(tm) and CRi,i-1(tm) are below the threshold TH9, then step S1304 is affirmed, and the process proceeds to step S1306. In step S1306, the battery cell 13 for which the condition of step S1304 is met is determined to have generated an anomaly. Then, in step S1307, the anomaly is notified.

[0222] If any of the battery cells 13 in the battery module 11 contains an abnormal cell, the correlation coefficient CR(tm) calculated based on the voltage change VA(tm) of two normal cells will differ from the correlation coefficient CR(tm) calculated based on the voltage change VA(tm) of the two cells containing the abnormal cell. Therefore, it is possible to appropriately detect abnormalities in the target cell.

[0223] In the above embodiment, the change in cell voltage V(tm) from the initial value is calculated as the voltage change VA(tm) of each battery cell 13, but this can be modified. For example, the change in cell voltage V(tm) over a specified time period, i.e., the time rate of change of voltage change VA(tm), can be calculated as voltage change VA(tm). In this case, the change in cell voltage V(tm) per unit time, i.e., the slope of cell voltage V(tm) on the time axis, can be calculated as voltage change VA(tm).

[0224] In the above embodiment, the cell voltage V(tm) detected by the voltage sensor 31 is used as the state parameter of the battery cell 13, but it can be changed. For example, it can be configured such that a temperature sensor is provided for each battery cell 13 of the battery module 11, and the cell temperature detected by each temperature sensor is used as the state parameter. In this case, the change in the cell temperature of each battery cell 13 over a specified period can be calculated as the temperature change (parameter change), and an anomaly of the target cell can be detected based on whether a specified deviation has occurred in the temperature change relative to other battery cells 13 within the target cell. The specific anomaly detection method can be the same as the methods in the above embodiments. Alternatively, a strain sensor for detecting the strain generated by the battery cell 13 and a pressure sensor for detecting the internal pressure of the battery cell 13 can be provided for each battery cell 13 of the battery module 11, and the detection values ​​(state detection values) of these sensors can be used as state parameters.

[0225] Furthermore, as a state parameter, a state calculation value can also be used to represent the state of the battery cell 13 based on state detection values ​​such as cell voltage and cell temperature. Specifically, the impedance of each battery cell 13 is calculated based on its cell voltage and current. Furthermore, the change in impedance of each battery cell 13 over a specified period can be used as the impedance change (parameter change), and an anomaly in the target cell can be detected based on whether a specified deviation has occurred in the impedance change relative to other battery cells 13. The specific anomaly detection method can be the same as the methods described in the above embodiments.

[0226] Alternatively, the configuration can be such that the State of Charge (SOC) is calculated as a state parameter based on the cell voltage and current of each battery cell 13, and anomalies in the battery cell 13 are detected based on the change in SOC over a specified period. Furthermore, the impedance can be calculated based on the temperature detection value of the temperature sensor of each battery cell 13, and this impedance can be used as a state parameter. Alternatively, in the battery module 11, a current sensor is provided for each cell to detect the current flowing through each battery cell 13, and the cell temperature is calculated as a state parameter based on the charge carried by each battery cell 13.

[0227] In the third embodiment described above, the deviation Ri(tm) of the voltage change VA(tm) in each battery cell 13 is calculated by the ratio of the absolute value of the difference between the voltage change VAi(tm) and the average value μi(tm) to the standard deviation σi(tm). However, this can be modified. For example, the deviation Ri(tm) can be defined as the difference between the voltage change VAi(tm) of the target cell and the average value μi(tm) of the voltage change VAi(tm) in each battery cell 13.

[0228] In this case, the battery ECU20 may calculate the average value μi(tm) of the voltage change VAi(tm) of the remaining battery cells 13 excluding the target cell, and calculate the deviation of the average value μi(tm) from the voltage change VAi(tm) of the target cell as the degree of deviation Ri(tm).

[0229] • In the internal short circuit process, which is the initial stage of the thermal runaway process of battery cell 13, the cell voltage gradually changes as the internal short circuit resistance of battery cell 13 changes during the initial period of the internal short circuit process, and the cell temperature (internal temperature) gradually rises in the next period after the initial period. Figure 25 The text indicates the changes in cell voltage and cell temperature during the initial period T1 and the next period T2 in the internal short-circuit process. It can be that, given these changes in cell voltage and cell temperature, anomaly detection of battery cell 13 is performed based on the amount of change in cell voltage during the initial period T1, and based on the amount of change in cell temperature during the next period T2.

[0230] Figure 26 An example of the battery system structure is shown. Each battery cell 13 of the battery module 11 is connected to a voltage sensor 31. In addition, a temperature sensor 34 is provided for each battery cell 13, and the detection signal of each temperature sensor 34 is input to the monitoring device 12. The voltage sensor 31 is equivalent to the "first sensor", and the temperature sensor 34 is equivalent to the "second sensor".

[0231] The battery ECU 20 calculates the voltage change VA (first parameter change) of the cell voltage detected by the voltage sensor 31 over a specified period, and calculates the temperature change TA (second parameter change) of the cell temperature detected by the temperature sensor 34 over a specified period. Furthermore, the battery ECU 20 can perform a first anomaly detection process based on the voltage change VA of each battery cell 13 to detect anomalies in the target cell, and a second anomaly detection process based on the temperature change TA of each battery cell 13 to detect anomalies in the target cell. The second anomaly detection process is performed if an anomaly is detected in the target cell through the first anomaly detection process.

[0232] Figure 27 This is a flowchart illustrating the steps involved in the abnormal detection and handling of battery cell 13, which is executed by battery ECU 20.

[0233] exist Figure 27 In step S1401, the cell voltage of each battery cell 13 detected by voltage sensor 31 and the cell temperature of each battery cell 13 detected by temperature sensor 34 are obtained as state detection values. The cell voltage and cell temperature are obtained at a predetermined period. In step S1402, the change in cell voltage of each battery cell 13 from the initial value is calculated as voltage change VA(tm), and the change in cell temperature of each battery cell 13 from the initial value is calculated as temperature change TA(tm). The initial value can be any predetermined timing.

[0234] In step S1403, it is determined whether the abnormality flag F is 0. The abnormality flag F indicates whether a certain battery cell 13 is determined to be abnormal in the first abnormality detection process. If there is no abnormality, the operation is F=0; if there is an abnormality, the operation is F=1.

[0235] If step S1403 is affirmed, then proceed to step S1404. In step S1404, a first anomaly detection process is performed. In the first anomaly detection process, anomaly detection of the target cell is performed based on the voltage change VA(tm) of each battery cell 13. In the first anomaly detection process, processing based on any of the anomaly detection processes in the above embodiments can be implemented. For example, using... Figure 5 In the case described above, for all battery cells 13, the absolute difference of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as a voltage difference ΔVi+1,i(tm). Based on this voltage difference ΔVi+1,i(tm), it is determined whether each battery cell 13 contains an abnormal cell. Furthermore, if it is determined that each battery cell 13 contains an abnormal cell, a notification of the abnormality is issued.

[0236] In step S1405, it is determined whether an abnormal battery cell has been detected through the first abnormality detection process. If no abnormal battery cell is detected, the process ends directly. In this case, the first abnormality detection process is repeatedly executed at a predetermined cycle. Alternatively, if an abnormal battery cell is detected, the process proceeds to step S1406, where the abnormality flag F is set to 1.

[0237] After the anomaly flag F is set to 1, step S1403 is negated, and the process proceeds to step S1407. In step S1407, a second anomaly detection process is performed. In the second anomaly detection process, anomaly detection of the target cell is performed based on the temperature change TA(tm) of each battery cell 13. In the second anomaly detection process, processing based on any of the anomaly detection processes described in the above embodiments can be implemented. For example, using... Figure 5 In the case described above, the voltage change VA(tm) can be replaced with the temperature change TA(tm) for execution. In this case, for all battery cells 13, the difference in the absolute value of the temperature change TA(tm) between two adjacent battery cells 13 in the series path is calculated as the temperature difference ΔTi+1,i(tm). Based on this temperature difference ΔTi+1,i(tm), it is determined whether each battery cell 13 contains an abnormal cell. Furthermore, if it is determined that each battery cell 13 contains an abnormal cell, a notification of the abnormality is issued.

[0238] Based on the above structure, it is possible to appropriately detect signs of thermal runaway in the battery cell 13 in stages while considering the state changes of the battery cell 13. A strain sensor or a pressure sensor can also be used instead of the temperature sensor 34. A strain sensor is installed in each battery cell 13, for example, to detect the strain of the casing of the battery cell 13. In this case, the strain detection value detected by the strain sensor is used as a state detection value, and anomalies in each battery cell 13 are detected based on the change in the strain detection value. Alternatively, a pressure sensor is installed in each battery cell 13, for example, to detect the pressure inside the casing of the battery cell 13. In this case, the pressure detection value detected by the pressure sensor is used as a state detection value, and anomalies in each battery cell 13 are detected based on the change in the pressure detection value.

[0239] · Figure 28 This is a graph showing the voltage-capacitance characteristic (OCV) of battery cell 13. This voltage-capacitance characteristic has a flat, stable region A1 and a steep region A2 where the voltage change relative to the capacitance is steeper than in the stable region A1. The steep region A2 includes areas near the lower SOC side and the higher SOC side compared to the stable region A1. For example, in the fully charged state of battery cell 13, the SOC of battery cell 13 enters the steep region A2 near the higher SOC side.

[0240] Here, we compare the case where the State of Charge (SOC) enters a stable region A1 with the case where the SOC enters a steep region A2. In the steep region A2, the voltage change relative to the SOC change is larger compared to the stable region A1. Therefore, it is considered that when the battery module 11 contains abnormal cells, the difference in voltage change VA between normal and abnormal cells further increases when the SOC enters the steep region A2. For example, if each battery cell 13 of the battery module 11 (battery pack B) is fully charged, and the battery module 11 contains abnormal cells, the difference in voltage change VA between the abnormal cells and normal cells becomes significant, improving the accuracy of anomaly detection. In view of this, it is possible to implement anomaly detection based on the voltage change VA (parameter change) as a condition that the SOC of each battery cell 13 becomes a steep region.

[0241] Figure 29 This is a flowchart illustrating the steps involved in the abnormal detection and handling of battery cell 13, which is executed by the battery ECU 20. Figure 29 In step S1501, it is determined whether the SOC of the battery cell 13 has entered the steep region A2. At this time, for example, if the SOC of all the battery cells 13 in the battery module 11 has entered the steep region A2, then step S1501 is affirmed.

[0242] Furthermore, if step S1501 is affirmed, the process proceeds to step S1502, where anomaly detection of each battery cell 13 is performed. In this anomaly detection, processing based on any of the anomaly detection processes described in the above embodiments can be implemented. For example, when using… Figure 5 In the case described above, for all battery cells 13, the absolute difference of the voltage change VA(tm) between two adjacent battery cells 13 in the series path is calculated as a voltage difference ΔVi+1,i(tm). Based on this voltage difference ΔVi+1,i(tm), it is determined whether each battery cell 13 contains an abnormal cell. Furthermore, if it is determined that each battery cell 13 contains an abnormal cell, a notification of the abnormality is issued.

[0243] Furthermore, anomaly detection for each battery cell 13 can be performed separately when the SOC of the battery cell 13 enters the stable region A1 and when the SOC of the battery cell 13 enters the steep region A2. In this case, the configuration can be such that, for each battery cell 13, considering that the amount of voltage change corresponding to the SOC change differs depending on whether the SOC enters the stable region A1 or the steep region A2, the anomaly detection threshold (TH11, TH12, etc.) for the voltage difference ΔV can be changed according to whether the SOC enters the stable region A1 or the steep region A2. For example, when the SOC is in the steep region A2, compared to the case where it is in the stable region A1, the positive threshold TH11 can be increased to the positive side, and the negative threshold TH12 can be increased to the negative side.

[0244] In the above embodiments, for the target cell, the difference between the voltage change VA(tm) and that of two different battery cells 13 is used as the voltage difference ΔV(tm) for calculation. However, this can be modified so that the difference between the voltage change VA(tm) and that of three or four different battery cells 13 is used as the voltage difference ΔV(tm) for calculation. In this case, it can be determined whether a predetermined deviation has occurred in the target cell based on the voltage difference ΔV(tm) calculated with other cells using three or four combinations. In this case, by increasing the combinations of calculating the voltage difference ΔV(tm), the reliability of anomaly detection can be improved.

[0245] In the above embodiment, the battery ECU20 is used as a battery status detection device, but this structure can be modified. For example, the monitoring device 12 of each battery cell 10 can be used as a battery status detection device, or an external device of the vehicle can be used as a battery status detection device.

[0246] The battery unit 10 in the above embodiments can be applied to stationary battery systems in addition to mobile bodies such as vehicles and airplanes.

[0247] The control device and method described in this disclosure can be implemented using a dedicated computer, which is provided by comprising a processor programmed to perform one or more functions embodied in a computer program and a memory. Alternatively, the control device and method described in this disclosure can be implemented using a dedicated computer provided by a processor composed of one or more dedicated hardware logic circuits. Alternatively, the control device and method described in this disclosure can be implemented using one or more dedicated computers, which are configured by a combination of a processor programmed to perform one or more functions and a memory, and a processor composed of one or more hardware logic circuits. Furthermore, the computer program can be stored as instructions to be executed by the computer on a computer-readable, non-temporary tangible recording medium.

[0248] The following describes the technical ideas extracted from the above-described embodiments.

[0249] [Structure 1] A battery state detection device (20) is applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series. The state of each battery cell is detected by a state detection unit. The battery state detection device includes: an acquisition unit that acquires the state detection value of each battery cell detected by the state detection unit; a parameter change calculation unit that takes one of the state detection value acquired by the acquisition unit and a state calculation value representing the state of the battery cell calculated based on the state detection value as a state parameter of the battery cell, and calculates the change of the state parameter over a predetermined period as a parameter change; and an anomaly detection unit that takes each of the multiple battery cells as a target cell for anomaly detection, and detects the anomaly of the target cell based on whether a predetermined deviation has occurred in the parameter change relative to other battery cells.

[0250] [Structure 2] According to the battery state detection device of structure 1, there is a difference calculation unit. The difference calculation unit calculates the difference between the parameter change amount of the target battery cell and two or more different battery cells as the change amount difference value. The anomaly detection unit determines whether the specified deviation has occurred in the target battery cell based on the change amount difference value calculated by the difference calculation unit, and detects the anomaly of the target battery cell based on the determination result.

[0251] [Structure 3] According to the battery state detection device described in Structure 2, the difference calculation unit calculates the difference in change between the first non-target cell and the target cell as a first difference, and calculates the difference in change between the second non-target cell and the target cell as a second difference. The first non-target cell and the second non-target cell are battery cells in the battery module that are different from the target cell. The anomaly detection unit determines that the specified deviation has occurred in the target cell when the absolute values ​​of the first difference and the second difference are both greater than a predetermined threshold.

[0252] [Structure 4] According to the battery state detection device of structure 2, the difference calculation unit calculates the difference in change between the first non-target cell and the target cell as a first difference at a predetermined period, and calculates the difference in change between the second non-target cell and the target cell as a second difference. The first non-target cell and the second non-target cell are battery cells in the battery module that are different from the target cell. The difference calculation unit accumulates the absolute values ​​of the first difference and the second difference each time the difference is calculated to calculate a first difference accumulation value and a second difference accumulation value. When both the first difference accumulation value and the second difference accumulation value are greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target cell.

[0253] [Structure 5] According to the battery state detection device described in Structure 2, the difference calculation unit takes the battery cell adjacent to the negative electrode side of the target battery cell as the first adjacent battery cell and the battery cell adjacent to the positive electrode side of the target battery cell as the second adjacent battery cell. The difference in the change between these adjacent battery cells and the target battery cell is calculated by subtraction operation in ascending or descending order of the battery cell series sequence, and set as the first difference and the second difference. The difference calculation unit calculates the difference between the first difference and the second difference as the secondary difference. If the secondary difference is greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target battery cell.

[0254] [Structure 6] According to the battery state detection device described in Structure 2, the difference calculation unit takes the battery cell adjacent to the negative electrode side of the target battery cell as the first adjacent battery cell and the battery cell adjacent to the positive electrode side of the target battery cell as the second adjacent battery cell. The difference in the change between these adjacent battery cells and the target battery cell is calculated by subtraction operation in ascending or descending order of the battery cell series sequence, and set as the first difference and the second difference. The difference calculation unit calculates the difference between the first difference and the second difference as the secondary difference. Then, whenever the difference calculation unit calculates the secondary difference, it accumulates the secondary difference to calculate the secondary difference accumulation value. When the secondary difference accumulation value is greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target battery cell.

[0255] [Structure 7] According to the battery state detection device of structure 2, in the battery system, multiple battery cell groups are formed by three or more battery cells connected in series with each other. The battery state detection device includes: a primary difference calculation unit, which calculates, for each battery cell group, the difference between the parameter change amount and that of the target cell from two or more different battery cells as a primary difference; an anomaly temporary determination unit, which, for each battery cell group, performs a temporary determination of whether there is an anomaly in the multiple battery cells based on the primary difference calculated by the primary difference calculation unit; and a secondary difference calculation unit, which calculates the secondary difference in the multiple battery cell groups for the anomaly... In the battery cell group that the temporary determination unit temporarily determines to be abnormal, the battery cell adjacent to the negative electrode side of the target cell is designated as the first adjacent cell, and the battery cell adjacent to the positive electrode side of the target cell is designated as the second adjacent cell. The difference in the amount of change between these adjacent cells and the target cell is calculated by subtraction operation in ascending or descending order of the cell series sequence, and set as the first difference and the second difference. The difference between the first difference and the second difference is set as the second difference for calculation. Based on the second difference calculated by the second difference calculation unit, the abnormality detection unit determines whether the specified deviation has occurred in the target cell.

[0256] [Structure 8] According to the battery state detection device of structure 2, it includes: a primary difference calculation unit, which calculates the difference in parameter change between the target cell and two or more different battery cells as primary differences for the target cell; and a secondary difference calculation unit, which takes the battery cells adjacent to the negative electrode side of the target cell as first adjacent cells, and the battery cells adjacent to the positive electrode side of the target cell as second adjacent cells, calculates the difference in the change between each of these adjacent cells and the target cell by subtraction operation in ascending or descending order of the cell series sequence, and sets them as the first difference and the second difference, and sets the first difference as the second difference. The difference between the first difference and the second difference is calculated as a secondary difference. The anomaly detection unit includes: a first determination unit that determines whether the specified deviation has occurred in the target cell based on the primary difference calculated by the primary difference calculation unit; a second determination unit that determines whether the specified deviation has occurred in the target cell based on the secondary difference calculated by the secondary difference calculation unit; and a switching execution unit that switches between detecting the anomaly of the target cell based on the determination result of the first determination unit or based on the determination result of the second determination unit, according to specified switching conditions.

[0257] [Structure 9] According to the battery state detection device of structure 8, a degradation determination unit is provided. The degradation determination unit determines the following situation: the battery module is in a state where each of the battery cells has deteriorated or the probability of degradation is high. The switching execution unit uses the determination result of the degradation determination unit as the switching condition. When it is determined that the battery cells have deteriorated or the probability of degradation is high, the abnormality of the target cell is detected according to the determination result of the second determination unit.

[0258] [Structure 10] According to the battery state detection device of structure 1, there is a deviation degree calculation unit. The deviation degree calculation unit calculates the degree to which the parameter change of the target cell deviates from the average value of the parameter change of each battery cell in the battery module as the deviation degree. The anomaly detection unit determines whether the predetermined deviation has occurred in the target cell based on the deviation degree calculated by the deviation degree calculation unit, and detects the anomaly of the target cell based on the determination result. The deviation degree calculation unit calculates the degree to which the parameter change of the target cell deviates from the average value of the parameter change of the remaining battery cells (excluding the target cell) among the plurality of battery cells as the deviation degree.

[0259] [Structure 11] According to the battery state detection device of structure 1, there is a deviation degree calculation unit. This deviation degree calculation unit calculates the deviation degree of the parameter change of the target cell relative to the average value of the parameter change of each of the battery cells in the battery module at a predetermined period. Each time the deviation degree is calculated, the deviation degree is accumulated to calculate a cumulative deviation degree value. The anomaly detection unit determines whether the predetermined deviation has occurred in the target cell based on the cumulative deviation degree value calculated by the deviation degree calculation unit, and detects anomalies of the target cell based on the determination result. The deviation degree calculation unit calculates the degree to which the parameter change of the target cell deviates relative to the average value of the parameter change of the remaining battery cells (excluding the target cell) among the plurality of battery cells as the deviation degree.

[0260] [Structure 12] The battery state detection apparatus according to Structure 1 includes: a deviation calculation unit that calculates a deviation, which is the difference between the parameter change of the target cell and the average value of the parameter change of each of the battery cells; a standard deviation calculation unit that calculates a standard deviation by taking the square root of the square of the difference between the parameter change of each of the battery cells and its average value; and a cumulative value calculation unit that accumulates the absolute value of the deviation calculated by the deviation calculation unit each time the deviation is calculated to calculate a cumulative deviation value. The anomaly detection unit determines whether the predetermined deviation has occurred in the target cell by comparing the cumulative deviation value calculated by the cumulative value calculation unit with a threshold set based on the standard deviation calculated by the standard deviation calculation unit. The deviation calculation unit calculates the deviation using the average value of the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells, and the standard deviation calculation unit calculates the standard deviation using the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells.

[0261] [Structure 13] According to the battery state detection device of structure 1, it includes: a difference calculation unit that, in the case of determining a specified cell one by one among the plurality of battery cells, designating the battery cell adjacent to the negative electrode side of the specified cell as a first adjacent cell, and the battery cell adjacent to the positive electrode side of the specified cell as a second adjacent cell, calculates the difference in the change of parameter change between the specified cell and the first adjacent cell and the difference in the change of parameter change between the specified cell and the second adjacent cell respectively by subtraction operation in ascending or descending order of cell series sequence, and sets them as a first difference and a second difference, and calculates the difference between the first difference and the second difference as a second difference; and a deviation degree calculation unit that, assuming the target cell is the specified cell, calculates the difference in parameter change between the specified cell and the second adjacent cell by subtraction operation in ascending or descending order of cell series sequence sequence. The quadratic difference calculated for a specified battery cell is a specific quadratic difference. The degree to which the specific quadratic difference deviates from the average of the quadratic differences calculated by setting each of the battery cells as the specified battery cell is used as the degree of deviation. Based on the degree of deviation calculated by the degree of deviation calculation unit, the anomaly detection unit determines whether the specified deviation has occurred in the target battery cell. Based on the determination result, the anomaly of the target battery cell is detected. The degree of deviation calculation unit uses the degree to which the specific quadratic difference deviates from the average of the remaining quadratic differences calculated by setting each of the battery cells as the specified battery cell, excluding the specific quadratic difference, as the degree of deviation.

[0262] [Structure 14] The battery state detection device according to Structure 1 includes: a difference calculation unit that, in the case of determining a specified cell one by one among the plurality of battery cells, designating the battery cell adjacent to the negative electrode side of the specified cell as a first adjacent cell, and the battery cell adjacent to the positive electrode side of the specified cell as a second adjacent cell, calculates the difference in the change of parameter change between the specified cell and the first adjacent cell and the difference in the change of parameter change between the specified cell and the second adjacent cell respectively by subtraction operation in ascending or descending order of cell series sequence, and sets them as a first difference and a second difference, and calculates the difference between the first difference and the second difference as a second difference; and a deviation degree calculation unit that sets the second difference calculated when the target cell is designated as the specified cell as a specific second difference. The degree of deviation is calculated as the extent to which a specific secondary difference deviates from the average of the secondary differences calculated by individually designating each of the battery cells as the designated cell. Based on the degree of deviation calculated by the degree of deviation calculation unit, the anomaly detection unit determines whether the specified deviation has occurred in the target cell, and detects anomalies in the target cell based on the determination result. The degree of deviation calculation unit calculates the degree of deviation as the extent to which a specific secondary difference deviates from the average of the secondary differences calculated by individually designating each of the battery cells as the designated cell, excluding the specific secondary difference and the secondary differences calculated by designating the first adjacent cell and the second adjacent cell as the designated cell.

[0263] [Structure 15] The battery state detection device according to any one of structures 1 to 14 includes an abnormal temporary determination unit, which temporarily determines that a battery cell in the plurality of battery cells whose difference from the average value of the parameter change is greater than or equal to a predetermined value is an abnormal cell, and the abnormal detection unit takes each of the abnormal cells temporarily determined by the abnormal temporary determination unit as the target cell and determines whether the predetermined deviation has occurred in the target cell.

[0264] [Structure 16] According to the battery state detection device of structure 1, there is a correlation coefficient calculation unit. The correlation coefficient calculation unit calculates a correlation coefficient for the target cell, which represents the strength of the correlation between the parameter change and two or more different battery cells. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the correlation coefficient calculated by the correlation coefficient calculation unit, and detects the anomaly of the target cell based on the determination result.

[0265] [Structure 17] According to the battery state detection device of structure 1, the state detection unit includes a first sensor (31) for detecting the inter-terminal voltage, i.e., cell voltage, of each battery cell and a second sensor (34) for detecting the temperature, strain, or internal pressure of each battery cell. The acquisition unit acquires the cell voltage of each battery cell detected by the first sensor and the temperature, strain, or internal pressure of each battery cell detected by the second sensor as the state detection value. The parameter change calculation unit calculates a first parameter change based on the cell voltage of each battery cell detected by the first sensor and calculates a second parameter change based on the temperature or strain of each battery cell detected by the second sensor as the parameter change. The anomaly detection unit can perform a first anomaly detection process based on the first parameter change and a second anomaly detection process based on the second parameter change, wherein the second anomaly detection process is performed if the anomaly of the target battery cell is detected by the first anomaly detection process.

[0266] [Structure 18] A battery state detection device is applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series. The state detection unit detects the state of each battery cell. The battery state detection device includes: a first anomaly detection unit that performs anomaly detection on each battery cell based on the state detection values ​​of each battery cell detected by the state detection unit through computational processing with relatively low computational load; a second anomaly detection unit that performs anomaly detection on each battery cell based on the state detection values ​​of each battery cell detected by the state detection unit through computational processing with relatively high computational load; and a switching execution unit that switches between the first anomaly detection unit and the second anomaly detection unit detecting anomalies on each battery cell according to predetermined switching conditions.

[0267] [Structure 19] According to the battery state detection device described in structure 18, the switching execution unit uses the detection of an abnormality in a certain battery cell in the battery module by the first abnormality detection unit as the switching condition, and with the fulfillment of the switching condition, the second abnormality detection unit performs abnormality detection.

[0268] [Structure 20] According to the battery state detection device of structure 18, there is a degradation determination unit. The degradation determination unit determines the following situation: the battery module is in a state where each of the battery cells has deteriorated or the probability of degradation is high. The switching execution unit uses the determination result of the degradation determination unit as the switching condition. When it is determined that the battery module is in a state where each of the battery cells has deteriorated or the probability of degradation is high, the second abnormality detection unit performs abnormality detection.

[0269] [Structure 21] According to any one of the structures 1 to 20, in the voltage-capacitance characteristic representing the relationship between voltage and capacitance of each battery cell, there is a stable region and a steep region where the voltage change relative to capacitance is steep compared to the stable region. The anomaly detection unit determines whether the specified deviation has occurred based on the parameter change amount, assuming that the capacitance of each battery cell is in the steep region.

[0270] [Structure 22] According to any one of structures 1 to 21, the battery state detection device is applied to a battery system having a voltage adjustment unit, the voltage adjustment unit adjusts the terminal voltage of each battery cell in the battery module to the same voltage, the acquisition unit acquires the terminal voltage of each battery cell as the state detection value, and the anomaly detection unit determines whether the predetermined deviation has occurred based on the parameter change amount, assuming that the voltage adjustment of the voltage adjustment unit has been performed.

[0271] This disclosure has been described based on embodiments, but it should be understood that this disclosure is not limited to these embodiments and structures. This disclosure also includes various modifications and variations within the same range. Furthermore, various combinations, forms, and other combinations and forms including only one element, more than one element, or less than one element also fall within the scope and spirit of this disclosure.

Claims

1. A battery state detection device (20) is applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series, wherein the state of each battery cell is detected by a state detection unit, the battery state detection device being characterized in that... have: The acquisition unit acquires the state detection values ​​of each battery cell detected by the state detection unit; The parameter change calculation unit takes one of the state detection value obtained by the acquisition unit and a state calculation value representing the state of the battery cell calculated based on the state detection value as a state parameter of the battery cell, and calculates the change in the state parameter over a predetermined period as a parameter change; and The anomaly detection unit identifies each of the plurality of battery cells as a target cell for anomaly detection and detects anomalies in the target cell based on whether a predetermined deviation has occurred in the parameter change amount relative to other battery cells.

2. The battery state detection device according to claim 1, characterized in that, The device includes a difference calculation unit that, for the target battery cell, calculates the difference in parameter changes between the target cell and two or more different battery cells as the difference in changes. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the change difference value calculated by the difference calculation unit, and detects the anomaly of the target cell based on the determination result.

3. The battery state detection device according to claim 2, characterized in that, The difference calculation unit calculates a first difference by using the difference in change between the first non-target cell and the target cell as the first difference, and calculates a second difference by using the difference in change between the second non-target cell and the target cell as the second difference. The first non-target cell and the second non-target cell are battery cells in the battery module that are different from the target cell. The anomaly detection unit determines that the specified deviation has occurred in the target cell when both the absolute value of the first difference and the absolute value of the second difference are greater than a specified threshold.

4. The battery state detection device according to claim 2, characterized in that, The difference calculation unit calculates a first difference at a predetermined period based on the difference in change between the first non-target cell and the target cell, and a second difference based on the difference in change between the second non-target cell and the target cell. The first non-target cell and the second non-target cell are battery cells in the battery module that are different from the target cell. The difference calculation unit accumulates the absolute values ​​of the first difference and the second difference each time the difference is calculated, to calculate the first cumulative difference value and the second cumulative difference value. If both the first cumulative difference value and the second cumulative difference value are greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target battery cell.

5. The battery state detection device according to claim 2, characterized in that, The difference calculation unit designates the battery cells adjacent to the negative electrode side of the target battery cell as first adjacent cells and the battery cells adjacent to the positive electrode side of the target battery cell as second adjacent cells. It calculates the difference in change between these adjacent cells and the target battery cell using subtraction operations in ascending or descending order of the cell series connection sequence, and sets these values ​​as the first difference and the second difference. The difference calculation unit calculates the difference between the first difference and the second difference as a secondary difference. If the difference between the two values ​​is greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target battery cell.

6. The battery state detection device according to claim 2, characterized in that, The difference calculation unit designates the battery cells adjacent to the negative electrode side of the target battery cell as first adjacent cells and the battery cells adjacent to the positive electrode side of the target battery cell as second adjacent cells. It calculates the difference in change between these adjacent cells and the target battery cell using subtraction operations in ascending or descending order of the cell series connection sequence, and sets these values ​​as the first difference and the second difference. The difference calculation unit calculates the difference between the first difference and the second difference as a secondary difference. Furthermore, whenever the difference calculation unit calculates the second difference, it accumulates the second difference to calculate the cumulative value of the second difference. If the cumulative value of the secondary difference is greater than a predetermined threshold, the anomaly detection unit determines that the predetermined deviation has occurred in the target battery cell.

7. The battery state detection device according to claim 2, characterized in that, In the battery system, multiple battery cell groups are formed by three or more battery cells connected in series with each other. The battery state detection device includes: The primary difference calculation unit calculates the primary difference for each battery cell group and for the target cell, taking the difference in parameter variation with two or more different battery cells as primary differences. The temporary anomaly determination unit performs a temporary determination on whether there is an anomaly in the plurality of battery cells for each battery cell group based on the primary difference calculated by the primary difference calculation unit. as well as The secondary difference calculation unit, among the battery cell groups that are temporarily determined to be abnormal by the temporary abnormality determination unit, designates the battery cells adjacent to the negative electrode side of the target cell as first adjacent cells and the battery cells adjacent to the positive electrode side of the target cell as second adjacent cells. It calculates the difference in change between these adjacent cells and the target cell using subtraction operations in ascending or descending order of the cell series connection sequence, and sets these as a first difference and a second difference. Furthermore, it calculates the difference between the first difference and the second difference as a secondary difference. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the secondary difference calculated by the secondary difference calculation unit.

8. The battery state detection device according to claim 2, characterized in that, have: The primary difference calculation unit calculates, for the target battery cell, the differences in parameter variations with two or more different battery cells as primary differences; and The secondary difference calculation unit designates the battery cells adjacent to the negative electrode side of the target battery cell as first adjacent cells and the battery cells adjacent to the positive electrode side of the target battery cell as second adjacent cells. It calculates the difference in change between these adjacent cells and the target battery cell using subtraction operations in ascending or descending order of the cell series connection sequence, and sets these as a first difference and a second difference. Furthermore, it calculates the difference between the first difference and the second difference as a secondary difference. The anomaly detection unit has: The first determination unit determines, based on the primary difference calculated by the primary difference calculation unit, whether the specified deviation has occurred in the target cell; The second determination unit determines, based on the second difference calculated by the second difference calculation unit, whether the specified deviation has occurred in the target cell; as well as The switching execution unit switches between detecting the abnormality of the target battery cell based on the determination result of the first determination unit or based on the determination result of the second determination unit, according to the prescribed switching conditions.

9. The battery state detection device according to claim 8, characterized in that, The battery module is equipped with a degradation determination unit that determines whether the battery modules are in a state where degradation of each battery cell has occurred or where the probability of degradation is high. The switching execution unit uses the determination result of the degradation determination unit as the switching condition. When it is determined that each of the battery cells has deteriorated or that the possibility of degradation is high, the unit detects the abnormality of the target cell based on the determination result of the second determination unit.

10. The battery state detection device according to claim 1, characterized in that, The battery module includes a deviation degree calculation unit, which calculates the degree to which the parameter change of the target cell deviates from the average value of the parameter change of each of the individual battery cells in the battery module. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the deviation degree calculated by the deviation degree calculation unit, and detects the anomaly of the target cell based on the determination result. The deviation degree calculation unit calculates the degree to which the parameter change of the target cell deviates from the average of the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells.

11. The battery state detection device according to claim 1, characterized in that, The device includes a deviation degree calculation unit that calculates, at predetermined intervals, the deviation degree of the parameter change of the target cell relative to the average value of the parameter change of each of the individual battery cells in the battery module, and accumulates the deviation degree each time it is calculated to calculate a cumulative deviation degree value. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the cumulative value of the deviation degree calculated by the deviation degree calculation unit, and detects the anomaly of the target cell based on the determination result. The deviation degree calculation unit calculates the degree to which the parameter change of the target cell deviates from the average of the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells.

12. The battery state detection device according to claim 1, characterized in that, have: The deviation calculation unit calculates the deviation, which is the difference between the parameter change of the target cell and the average value of the parameter change of each of the battery cells. The standard deviation calculation unit calculates the standard deviation by taking the square root of the square of the difference between the parameter variation of each battery cell and its average value. as well as The cumulative value calculation unit accumulates the absolute values ​​of the deviations calculated by the deviation calculation unit each time the deviation is calculated to calculate a cumulative deviation value. The anomaly detection unit determines whether the specified deviation has occurred in the target battery cell by comparing the cumulative deviation value calculated by the cumulative value calculation unit with a threshold set based on the standard deviation calculated by the standard deviation calculation unit. The deviation calculation unit calculates the deviation using the average of the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells. The standard deviation calculation unit calculates the standard deviation using the parameter changes of the remaining battery cells (excluding the target cell) among the plurality of battery cells.

13. The battery state detection device according to claim 1, characterized in that, have: The difference calculation unit determines a specific cell from the plurality of battery cells one by one. If the battery cell adjacent to the negative electrode side of the specified cell is designated as the first adjacent cell, and the battery cell adjacent to the positive electrode side of the specified cell is designated as the second adjacent cell, the difference in the change of parameter between the specified cell and the first adjacent cell and the difference in the change of parameter between the specified cell and the second adjacent cell are calculated by subtraction operation in ascending or descending order of cell series sequence, and set as the first difference and the second difference. The difference between the first difference and the second difference is calculated as the second difference. as well as The deviation degree calculation unit defines a specific quadratic difference value as the quadratic difference value calculated by treating the target cell as the designated cell. It then calculates the degree of deviation of this specific quadratic difference value from the average of the quadratic differences calculated by treating each individual cell as the designated cell, and defines it as the deviation degree. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the deviation degree calculated by the deviation degree calculation unit, and detects the anomaly of the target cell based on the determination result. The deviation degree calculation unit calculates the degree of deviation of the specific secondary difference relative to the average of the remaining secondary differences calculated by treating each of the battery cells as the designated cell, excluding the specific secondary difference.

14. The battery state detection device according to claim 1, characterized in that, have: The difference calculation unit determines a specific cell from the plurality of battery cells one by one. If the battery cell adjacent to the negative electrode side of the specified cell is designated as the first adjacent cell, and the battery cell adjacent to the positive electrode side of the specified cell is designated as the second adjacent cell, the difference in the change of parameter between the specified cell and the first adjacent cell and the difference in the change of parameter between the specified cell and the second adjacent cell are calculated by subtraction operation in ascending or descending order of cell series sequence, and set as the first difference and the second difference. The difference between the first difference and the second difference is calculated as the second difference. as well as The deviation degree calculation unit defines a specific quadratic difference value as the quadratic difference value calculated by treating the target cell as the designated cell. It then calculates the degree of deviation of this specific quadratic difference value from the average of the quadratic differences calculated by treating each individual cell as the designated cell, and defines it as the deviation degree. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the deviation degree calculated by the deviation degree calculation unit, and detects the anomaly of the target cell based on the determination result. The deviation degree calculation unit calculates the degree of deviation of the specific secondary difference relative to the average of the secondary differences calculated by treating each of the battery cells as the designated cell, excluding the specific secondary difference and the secondary differences calculated by treating the first adjacent cell and the second adjacent cell as the designated cell, respectively.

15. The battery state detection device according to any one of claims 1 to 14, characterized in that, The system includes a temporary anomaly determination unit that temporarily determines a battery cell among the plurality of battery cells whose difference from the average value of the parameter variation is greater than or equal to a predetermined value as an abnormal cell. The anomaly detection unit takes each of the abnormal cells temporarily identified by the anomaly temporary determination unit as the target cell and determines whether the specified deviation has occurred in the target cell.

16. The battery state detection device according to claim 1, characterized in that, The device includes a correlation coefficient calculation unit that calculates, for each target battery cell, a correlation coefficient representing the strength of the correlation between the parameter variations and those of two or more different battery cells. The anomaly detection unit determines whether the specified deviation has occurred in the target cell based on the correlation coefficient calculated by the correlation coefficient calculation unit, and detects the anomaly of the target cell based on the determination result.

17. The battery state detection device according to claim 1, characterized in that, The state detection unit includes a first sensor (31) for detecting the inter-terminal voltage, i.e., the cell voltage, of each of the battery cells, and a second sensor (34) for detecting the temperature, strain, or internal pressure of each of the battery cells. The acquisition unit acquires the cell voltage of each battery cell detected by the first sensor, and the temperature, strain, or internal pressure of each battery cell detected by the second sensor, as the state detection value. The parameter change calculation unit calculates a first parameter change based on the cell voltage of each battery cell detected by the first sensor, and calculates a second parameter change based on the temperature or strain of each battery cell detected by the second sensor, and uses these as the parameter change amounts. The anomaly detection unit is capable of performing a first anomaly detection process based on the change in the first parameter to detect anomalies in the target battery cell and a second anomaly detection process based on the change in the second parameter to detect anomalies in the target battery cell. The second anomaly detection process is performed if an anomaly in the target battery cell is detected by the first anomaly detection process.

18. A battery state detection device, applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series, wherein the state of each battery cell is detected by a state detection unit, the battery state detection device being characterized in that... have: The first anomaly detection unit performs anomaly detection on each battery cell based on the state detection values ​​of each battery cell detected by the state detection unit, through computational processing with a relatively low computational load. The second anomaly detection unit, based on the state detection values ​​of each battery cell detected by the state detection unit, performs anomaly detection on each battery cell through computationally intensive processing; and The switching execution unit switches between detecting abnormalities in each battery cell by the first abnormality detection unit and detecting abnormalities in each battery cell by the second abnormality detection unit, according to the prescribed switching conditions.

19. The battery state detection device according to claim 18, characterized in that, The switching execution unit uses the detection of an abnormality in a certain battery cell in the battery module by the first abnormality detection unit as the switching condition. With the fulfillment of the switching condition, the second abnormality detection unit performs abnormality detection.

20. The battery state detection device according to claim 18, characterized in that, The battery module is equipped with a degradation determination unit that determines whether the battery modules are in a state where degradation of each battery cell has occurred or where the probability of degradation is high. The switching execution unit uses the determination result of the degradation determination unit as the switching condition. When it is determined that the degradation of each of the battery cells has occurred or that the possibility of degradation is high, the second abnormality detection unit performs abnormality detection.

21. The battery state detection device according to claim 1, characterized in that, The voltage-capacitance characteristics of each battery cell, which represent the relationship between voltage and capacitance, have a stable region and a steep region where the voltage change relative to the capacitance is steep compared to the stable region. The anomaly detection unit determines whether the specified deviation has occurred based on the parameter change, assuming that the capacitance of each battery cell is in the steep region.

22. The battery state detection device according to claim 1, characterized in that, The battery state detection device is applied to a battery system equipped with a voltage adjustment unit, which adjusts the terminal voltage of each battery cell in the battery module to the same voltage. The acquisition unit acquires the terminal voltage of each of the battery cells as the state detection value. The anomaly detection unit determines whether the specified deviation has occurred based on the parameter change, assuming that the voltage adjustment unit has performed the voltage adjustment.

23. A program executed by a computer (20), applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series, wherein the state of each battery cell is detected by a state detection unit, the program being characterized in that... have: The process involves obtaining the state detection values ​​of each battery cell detected by the state detection unit. The parameter change calculation process involves taking one of the state detection values ​​obtained from the acquisition process and a state calculation value representing the state of the battery cell calculated based on the state detection values ​​as the state parameter of the battery cell, and calculating the change in the state parameter over a specified period as the parameter change; and The anomaly detection process involves treating each of the multiple battery cells as a target cell for anomaly detection. The anomaly of the target cell is detected based on whether it exhibits a predetermined deviation in the parameter changes relative to other battery cells.

24. A program executed by a computer (20), applied to a battery system having a battery module (11) formed by connecting multiple battery cells (13) in series, wherein the state of each battery cell is detected by a state detection unit, the program being characterized in that... This will cause the following processing to be performed: The first anomaly detection process involves detecting anomalies in each battery cell based on the state detection values ​​detected by the state detection unit, using computational processing with a relatively low computational load. The second anomaly detection process involves performing anomaly detection on each battery cell based on the state detection values ​​detected by the state detection unit, using computationally intensive processing. The switching process involves switching, based on specified switching conditions, between the first anomaly detection process and the second anomaly detection process for detecting anomalies in each battery cell.

Images