JP7306932B2 - Rechargeable battery status detection device and rechargeable battery status detection method - Google Patents

Description

The present invention relates to a rechargeable battery state detection device and a rechargeable battery state detection method.

Conventionally, there are technologies disclosed in Patent Document 1 and Patent Document 2 as technologies for detecting an abnormality occurring in the electrode plate of a rechargeable battery.

In the technology disclosed in Patent Document 1, disturbance due to repeated rise/fall compared to a preset voltage or current or a voltage or current determined by learning charging behavior in a state where no abnormality occurs. If it is large, it is determined that an electrode plate abnormality has occurred.

Further, in the technique disclosed in Patent Document 2, the open-circuit voltage of a rechargeable battery after completion of charging is obtained for a predetermined period at a predetermined sampling cycle, the obtained open-circuit voltage is approximated by a power approximation formula, and the reduction rate of the convergence voltage is calculated. exceeds a predetermined reference level, it is determined that an abnormality (cell short) has occurred in the electrode plate.

JP-A-07-165016 JP-A-2005-043339

By the way, in the technology disclosed in Patent Document 1, since detection is performed during current-limited charging, it is difficult to ensure current-limited charging in a rechargeable battery mounted on a vehicle, for example. There is a problem that it is difficult to detect the abnormality of the electrode plate.

In the technique disclosed in Patent Document 2, the voltage of the rechargeable battery may change due to current fluctuations, and in such cases, there is a problem that the electrode plate is erroneously determined to be abnormal.

SUMMARY OF THE INVENTION The present invention has been made in view of the circumstances described above. It is an object of the present invention to provide a state detection device and a rechargeable battery state detection method.

In order to solve the above problems, the present invention provides a rechargeable battery state detecting device for detecting the state of a rechargeable battery, wherein detecting means detects the voltage of the rechargeable battery based on a signal output from a voltage detecting section. and optimization means for optimizing by adjusting coefficients of a predetermined mathematical model representing the internal state of the rechargeable battery based on the voltage value detected by the detection means; determining means for determining whether or not there is an abnormality in the electrode plate existing in the electrolyte of the rechargeable battery based on the value calculated by the mathematical model with the adjusted coefficient or the value of the coefficient; and an output means for outputting the determination result.
According to such a configuration, it is possible to accurately detect an abnormality occurring in the electrodes of the rechargeable battery mounted on the vehicle regardless of the usage environment.

Further, the present invention is characterized in that the electrode plate abnormality is an abnormality in which the electrical characteristics of the rechargeable battery change discontinuously.
According to such a configuration, it is possible to detect the electrode plate abnormality of the rechargeable battery based on the electrical characteristics.

Also, the present invention is characterized in that the electrode plate abnormality is a short circuit in a partial region of the electrode plate.
According to such a configuration, it becomes possible to accurately detect a plate abnormality that causes a voltage drop.

Further, the present invention is characterized in that the mathematical model includes a term indicating voltage change of the rechargeable battery over time.
With such a configuration, it is possible to determine whether or not there is an abnormality in the electrode plate based on the voltage change over time.

Further, the present invention is characterized in that the mathematical model includes a term indicating voltage change accompanying change in current of the rechargeable battery.
According to such a configuration, it is possible to determine the presence or absence of electrode plate abnormality even when there is a change in current.

Also, the present invention is characterized in that the mathematical model represents a voltage change after charging/discharging of the rechargeable battery is stopped.
According to such a configuration, it is possible to accurately determine the presence or absence of electrode plate abnormality by utilizing the time when the voltage after charging/discharging is stopped is stabilized.

Further, according to the present invention, the mathematical model representing a voltage change after charging and discharging of the rechargeable battery is stopped is an exponential decay function or It is characterized by being an inversely proportional function.
According to such a configuration, it is possible to accurately detect voltage changes.

Also, the present invention is characterized in that the mathematical model represents voltage changes during charging and discharging of the rechargeable battery.
According to such a configuration, it is possible to determine the presence or absence of electrode plate abnormality even during charging and discharging.

Further, the present invention is characterized in that the mathematical model representing voltage changes during charging and discharging of the rechargeable battery is an equivalent circuit model of the rechargeable battery.
According to such a configuration, it is possible to determine the presence or absence of electrode plate abnormality even during charging and discharging based on the equivalent circuit model.

Further, in the present invention, the optimizing means includes a first mathematical model representing a voltage change after the rechargeable battery stops charging and discharging, and a second mathematical model representing a voltage change while the rechargeable battery is charging and discharging. a model is selected according to the state of charge of the rechargeable battery to select the coefficients, and the detecting means is based on the first mathematical model or the second mathematical model selected by the optimizing means; It is characterized in that the electrode plate abnormality is detected.
According to such a configuration, it is possible to determine the presence or absence of electrode plate abnormality regardless of the charging/discharging state.

Further, in the present invention, the determining means determines whether or not the electrode plate is abnormal based on the value calculated by the mathematical model and the actual measurement value of the voltage of the rechargeable battery at two or more different times. It is characterized by
According to such a configuration, it is possible to accurately determine the presence or absence of the electrode plate abnormality based on the values calculated at different times and the measured values.

Also, in the present invention, the determination means determines whether each of the difference values between the value calculated by the mathematical model and the actual measurement value of the voltage of the rechargeable battery is equal to or greater than a predetermined threshold at two or more different times. Alternatively, when the sum of the difference values or the sum of squares of the difference values is equal to or greater than a predetermined threshold value, the electrode plate abnormality is determined.
According to such a configuration, it is possible to accurately determine the presence or absence of electrode plate abnormality by a simple method.

Further, in the present invention, the determination means determines that the frequency or frequency rate at which the difference value between the value calculated by the mathematical expression model and the measured voltage value of the rechargeable battery is outside a predetermined range is greater than or equal to a predetermined threshold value. When it becomes so, it is determined that the electrode plate is abnormal.
According to such a configuration, it is possible to accurately determine the presence or absence of the electrode plate abnormality based on the frequency or the frequency rate.

Further, according to the present invention, the determination means determines that, when the standard deviation of the difference value between the value calculated by the mathematical model and the measured value of the voltage of the rechargeable battery is equal to or greater than a predetermined threshold value, the electrode plate It is characterized by judging that it is abnormal.
According to such a configuration, it is possible to accurately determine the presence or absence of electrode plate abnormality based on the standard deviation.

Further, in the present invention, the determination means determines that the electrode plate is abnormal when there are a plurality of peaks in the distribution of the difference values between the value calculated by the mathematical model and the voltage measurement value of the rechargeable battery. It is characterized by judging.
With such a configuration, it is possible to accurately determine whether or not there is an abnormality in the electrode plate based on the peak of the distribution.

Also, the present invention is characterized in that the determining means determines that the electrode plate is abnormal based on at least one coefficient of the mathematical model optimized by the optimizing means.
According to such a configuration, it is possible to accurately determine the presence or absence of the electrode plate abnormality based on the coefficient.

Further, according to the present invention, the determining means determines that at least one coefficient of the mathematical model optimized by the optimizing means differs from a coefficient when optimized in the past by a predetermined threshold or more. (3) is characterized by determining that the electrode plate is abnormal.
According to such a configuration, it is possible to accurately determine whether or not there is an abnormality in the electrode plate by comparison with past coefficients.

Further, according to the present invention, the rechargeable battery is mounted on a vehicle, the output means notifies a higher-level control device mounted on the vehicle of the determination result by the determination means, and the control device receiving the notification , issuing a warning to the user based on the determination result and/or suspending execution of idling stop for stopping the engine during idling.
According to such a configuration, it is possible to notify the user of the occurrence of the electrode plate abnormality and prevent the engine from being unable to restart.

The present invention also provides a rechargeable battery state detection method for detecting a state of a rechargeable battery, comprising: a detection step of detecting the voltage of the rechargeable battery based on a signal output from a voltage detection section; an optimization step of optimizing by adjusting coefficients of a predetermined mathematical model representing the internal state of the rechargeable battery based on the detected voltage value; determining the presence or absence of an electrode plate abnormality in the electrode plate existing in the electrolyte of the rechargeable battery based on the value calculated by the mathematical model or the value of the coefficient; and outputting the determination result of the determining step. and an output step.
According to such a method, it is possible to accurately detect an abnormality occurring in the electrodes of the rechargeable battery mounted on the vehicle regardless of the usage environment.

According to the present invention, there is provided a rechargeable battery state detecting device and a rechargeable battery state detecting method capable of accurately detecting an abnormality occurring in the electrodes of a rechargeable battery mounted on a vehicle regardless of the usage environment. can do.

1 is a diagram showing a configuration example of a vehicle power supply system having a rechargeable battery state detection device according to an embodiment of the present invention; FIG. 2 is a block diagram showing a detailed configuration example of a control unit in FIG. 1; FIG. FIG. 2 is a diagram showing changes in voltage and current of a rechargeable battery after charging and discharging are stopped when the rechargeable battery is normal; FIG. 2 is a diagram showing changes in voltage and current of a rechargeable battery after charging and discharging is stopped when electrode plate abnormality occurs in the rechargeable battery; 4 is a flowchart for explaining an example of processing executed in an embodiment of the present invention; 7 is a flowchart for explaining an example of a process of optimizing coefficients of a mathematical model; 4 is a flowchart for explaining an example of a process for determining the presence or absence of electrode plate abnormality; FIG. 4 is a diagram showing the relationship between voltage measurement timing and the number of terms of an exponential decay function; FIG. 4 is a diagram showing the relationship between error and current when current flows through a rechargeable battery; FIG. 10 is a diagram showing the distribution of difference values with and without electrode plate abnormality; 4 is a flowchart for explaining an example of a process for determining the presence or absence of electrode plate abnormality; FIG. 4 shows the process of optimizing the modulus of a healthy rechargeable battery; FIG. 3 is a diagram showing the process of optimizing the coefficient of a rechargeable battery with plate anomalies; FIG. 14 is a diagram showing respective values of coefficients A1 to A6 optimized in FIG. 12 or FIG. 13; 4 is a flowchart for explaining an example of a process for determining the presence or absence of electrode plate abnormality; It is a figure for demonstrating an example of a time window. It is a figure which shows an example of an equivalent circuit model.

Next, embodiments of the present invention will be described.

(A) Description of Configuration of Embodiment of the Invention FIG. 1 is a diagram showing a power supply system of a vehicle having a rechargeable battery state detection device according to an embodiment of the invention. In this figure, the rechargeable battery state detection device 1 has a control unit 10 as a main component, and a voltage detection unit 11, a current detection unit 12, a temperature detection unit 13, and a discharge circuit 15 are connected to the outside. The state of the rechargeable battery 14 is detected. Note that the control unit 10, the voltage detection unit 11, the current detection unit 12, the temperature detection unit 13, and the discharge circuit 15 may not be configured separately, but may be configured by integrating some or all of them.

Here, the control unit 10 refers to outputs from the voltage detection unit 11, the current detection unit 12, and the temperature detection unit 13, detects the state of the rechargeable battery 14, and outputs information on the detection result to the outside. At the same time, the state of charge of the rechargeable battery 14 is controlled by controlling the voltage generated by the alternator 16 . The control unit 10 does not control the charging state of the rechargeable battery 14 by controlling the voltage generated by the alternator 16. For example, an ECU (Electric Control Unit) (external control unit) (not shown) controls the control unit 10. You may make it control a charge state based on the information from.

The voltage detection unit 11 detects the terminal voltage of the rechargeable battery 14 and supplies it to the control unit 10 as a voltage signal. The current detection unit 12 detects the current flowing through the rechargeable battery 14 and supplies it to the control unit 10 as a current signal. The temperature detection unit 13 detects the ambient temperature or internal temperature of the rechargeable battery 14 and supplies it to the control unit 10 as a temperature signal.

The discharge circuit 15 is composed of, for example, a semiconductor switch and a resistance element connected in series. can be made Note that the discharge circuit 15 can be excluded from the configuration.

Alternator 16 is driven by engine 17 to generate AC power which is converted to DC power by a rectifier circuit to charge rechargeable battery 14 . The alternator 16 is controlled by the controller 10 and is capable of adjusting the generated voltage.

The engine 17 is composed of, for example, a reciprocating engine such as a gasoline engine and a diesel engine, a rotary engine, or the like, and is started by a starter motor 18 to drive drive wheels through a transmission to provide propulsion to the vehicle. to generate electric power. The starter motor 18 is composed of, for example, a direct-current motor, and generates rotational force by electric power supplied from the rechargeable battery 14 to start the engine 17 . An electric motor may be used instead of the engine 17 .

The load 19 includes, for example, an electric steering motor, a defogger, a seat heater, an ignition coil, a car audio system, a car navigation system, and the like, and operates with power supplied from the rechargeable battery 14 . In the example of FIG. 1, only the engine 17 is configured to output driving force, but a hybrid vehicle equipped with an electric motor that assists the engine 17 may be used. In the case of a hybrid vehicle, the rechargeable battery 14 activates a high voltage system (system for driving an electric motor) composed of a lithium battery or the like, and the high voltage system starts the engine 17 .

FIG. 2 is a diagram showing a detailed configuration example of the control unit 10 shown in FIG. As shown in this figure, the control unit 10 includes a CPU (Central Processing Unit) 10a as a processor, a ROM (Read Only Memory) 10b, a RAM (Random Access Memory) 10c, a communication unit 10d, and an I/F (Interface) 10e. , and a bus 10f. Here, the CPU 10a controls each section based on a program 10ba stored in the ROM 10b. The ROM 10b is composed of a semiconductor memory or the like, and stores programs 10ba and the like. The RAM 10c is configured by a semiconductor memory or the like, and stores data generated when executing the program 10ba and data 10ca such as tables. The communication unit 10d communicates with an ECU or the like, which is a host device, and notifies the host device of detected information or control information. The I/F 10e converts the voltage signal and the current signal supplied from the voltage detection unit 11, the current detection unit 12, and the temperature detection unit 13 into digital signals and takes them in, and the discharge circuit 15, the alternator 16, and the , the starter motor 18 and the like to control them. The bus 10f is a group of signal lines for interconnecting the CPU 10a, ROM 10b, RAM 10c, communication section 10d, and I/F 10e and enabling information exchange therebetween. The control of the alternator 16, the starter motor 18, etc. may be executed by the ECU. It is not limited to the configuration of FIG.

In addition, although one CPU 10a is provided in the example of FIG. 2, distributed processing may be executed by a plurality of CPUs. Alternatively, the CPU 10a may be replaced by a DSP (Digital Signal Processor), FPGA (Field Programmable Gate Array), ASIC (Application Specific Integrated Circuit), or the like. Alternatively, the processing may be performed by a computer on a server using a general-purpose processor or cloud computing that executes the functions by loading a software program. In addition, although the ROM 10b and the RAM 10c are provided in FIG. 2, for example, a storage device other than these (for example, an HDD (Hard Disk Drive) which is a magnetic storage device) may be used.

(B) Description of operation of embodiment of the present invention Next, operation of the embodiment of the present invention will be described. In the following, after describing the operation of the embodiment of the present invention, the processing of the flowchart for realizing such an operation will be described.

First, the outline of the operation of the embodiment of the present invention will be described. The rechargeable battery 14, for example, in the case of a lead-acid battery, has an anode made of lead dioxide and a cathode made of lead. Such an electrode plate may have an abnormality (hereinafter simply referred to as an "electrode plate abnormality") accompanied by a discontinuous change in electrical characteristics (for example, terminal voltage or internal impedance). For example, if lead sulfate generated by discharge is bridged between the anode and cathode and charged, the lead sulfate is reduced and changed to lead, which causes a short circuit in a part of the electrode plate. There is In addition, vibrations of a vehicle or the like may cause the electrode plates to peel off, and the separated electrode plates may cause the anode and the cathode to be electrically connected, thereby causing a short circuit in a part of the electrode plate. A short circuit that occurs in a partial region of the electrode plate is hereinafter referred to as a "partial short circuit."

In the embodiment of the present invention, the voltage of the rechargeable battery 14 is represented by a mathematical model, which will be described later, and the coefficients of the mathematical model are optimized so that the values calculated by the mathematical model and the measured voltage values match. When the voltage value calculated by the mathematical model thus obtained deviates from the actually measured voltage value, it is determined that an electrode plate abnormality accompanied by a discontinuous change in electrical characteristics has occurred. Then, the determination result is notified to the ECU, which is a higher control device. The notified ECU or control device may issue a warning to the user based on the determination result. Further, when the electrode plate abnormality occurs, the engine 17 may be prevented from being unable to be restarted by, for example, suspending the execution of the idling stop.

The operation of the present invention will now be described in detail. In this embodiment, when the engine 17 of the vehicle is stopped and the rechargeable battery 14 is in a stable state, the CPU 10a acquires the second-order exponential decay function shown in the following equation (1) from the ROM 10b.

・・・（１）

... (1)

Here, VFn (n, In) is a second-order exponential decay function, n indicates the number of sampled data, In is the n-th current sampling value, and Δt is the data sampling period. , and exp indicates an exponential function. In addition, “A1 exp (A3 Δt n) + A2 exp (A4 Δt n) + A5” is a term that indicates the change in voltage over time, and “A6 In” is due to the influence of dark current. This is a term indicating a voltage change.

The CPU 10a initializes the coefficients of equation (1) obtained from the ROM 10b. More specifically, the CPU 10a sets the value of Δt (eg, 10 seconds).

Next, the CPU 10a acquires a voltage value and a current value from the voltage detection section 11 and the current detection section 12, and sets them as Vn and In, respectively.

Next, the CPU 10a acquires the initial values of the coefficients A1 to A6 of Equation (1) from the ROM 10b and sets them in Equation (1). These coefficients A1 to A6 are used to derive the optimum solution by the method of least squares, and the values are sequentially updated in the process of calculation as described later. As the initial values of the coefficients A1 to A6, predetermined values obtained in advance through experiments may be used.

Next, the CPU 10a substitutes n and the current measurement value In into the equation (1) to calculate VFn(n, In). As a result, the value of VFn(n, In), which is the result of calculation by the mathematical model, is obtained.

Next, the CPU 10a applies VFn(n, In) obtained by the equation (1) and the voltage value Vn as the measured value obtained from the voltage detection unit 11 to the following equation (2) to obtain the value of Rn get

・・・（２）

... (2)

Next, the CPU 10a calculates partial differential terms corresponding to the coefficients A1 to A6 based on the following equation (3) in order to apply the least squares method.

・・・（３）

... (3)

Next, the CPU 10a calculates a matrix B to be applied to the simultaneous equations of the least-squares method using each partial differential term obtained by the equation (3), based on the following equation (4).

・・・（４）

... (4)

Note that the matrix B in the formula (4) is a 6×6 square matrix and a symmetric matrix satisfying B(x, y)=B(y, x).

Next, the CPU 10a calculates dR represented by the following formula (5) using Rn obtained by the formula (2) and the partial differential term obtained by the formula (4).

・・・（５）

... (5)

Next, the CPU 10a calculates the difference dd given by the following formula (6) using the matrix B obtained by the formula (4) and the dR obtained by the formula (5).

・・・（６）

... (6)

Since the six differences dd1 to dd6 corresponding to the coefficients A1 to A6 can be obtained in the above manner, the CPU 10a satisfies the following equation (7) for these six differences dd1 to dd6. Determine whether or not

・・・（７）

... (7)

Note that Th on the right side of Equation (7) can be set to, for example, 10 ⁻¹² . Note that 10 ⁻¹² is just an example, and other values that can be judged as values close to zero may be used. As a result, when it is determined that the formula (7) is not satisfied, the coefficients A1 to A6 are updated based on the following formula (8), and the same processing as described above is repeated.

・・・（８）

... (8)

On the other hand, if it is determined that the formula (7) is satisfied, the differences dd1 to dd6 are sufficiently close to zero, so it is determined that the optimum solution of the least-squares method has been obtained at that point.

Through the above processing, a mathematical model having optimized coefficients A1 to A6 can be obtained.

In the mathematical model obtained as described above, when the electrode plate is normal, the actually measured voltage and the voltage estimated by the mathematical model agree well. FIG. 3 shows changes in the measured voltage of the rechargeable battery 14 after stopping the engine 17 (indicated by circles), changes in voltage calculated by the mathematical model (indicated by diamonds), and changes in current (indicated by triangles). ). As shown in FIG. 3, when the electrodes of the rechargeable battery 14 are normal, the circles indicating the measured voltage and the diamonds indicating the voltage calculated by the mathematical model match well.

On the other hand, when an electrode plate abnormality occurs in the rechargeable battery 14, a voltage fluctuation occurs due to the electrode plate abnormality, and a deviation occurs between the voltage calculated by the mathematical model and the actually measured voltage. FIG. 4 shows changes in the voltage measured by the rechargeable battery 14 when a cell short (partial short), which is an electrode abnormality, occurs around 1000 seconds (indicated by circles), and changes in the voltage calculated by the mathematical model ( diamonds) and changes in current (triangles). In FIG. 4, a deviation occurs between the actually measured voltage of the rechargeable battery 14 and the calculated voltage of the mathematical model after 1000 seconds when the cell short occurs. By detecting such deviation, it is possible to detect the electrode plate abnormality of the rechargeable battery 14 .

More specifically, the CPU 10a obtains the model voltage VFn(n, In) by substituting the measured value of the current In into the equation (1) obtained by obtaining the optimum solution for the coefficients A1 to A6 through the above processing. . Next, the CPU 10a obtains the absolute value of the difference between VFn(n, In) and the voltage measurement value Vn. Otherwise, it is determined that there is no electrode plate abnormality. It should be noted that there is a possibility that an erroneous determination will occur if only one determination is made. Therefore, it may be determined that an electrode plate abnormality has occurred when an abnormality is determined by multiple determinations.

Next, details of the processing performed in the embodiment of the present invention will be described with reference to FIGS. 5 to 7. FIG.

FIG. 5 is a diagram for explaining the main processing executed in this embodiment. The processing shown in FIG. 5 is executed with a predetermined time window (for example, a one-hour window) as a repeating cycle, and the mathematical model shown in Equation (1) is optimized within the time window. When the process of the flowchart shown in FIG. 5 is started, the following steps are executed.

At step S1, the CPU 10a executes a process of optimizing the coefficients A1 to A6 of the mathematical model shown in Equation (1). Details of the processing in step S1 will be described later with reference to FIG.

In step S2, the CPU 10a calculates VFn(n, In) (n=1, 2, . Execute the process to calculate As a result, Ns (for example, 60) pieces of data can be obtained.

In step S3, the CPU 10a refers to the mathematical model calculated in step S2 and executes an electrode plate abnormality determination process for determining whether or not an electrode plate abnormality has occurred. Details of the processing in step S3 will be described later with reference to FIG.

In step S4, the CPU 10a proceeds to step S5 if it is determined that an electrode plate abnormality has occurred in the electrode plate abnormality determination process shown in step S3 (step S4: Y); :N), the process proceeds to step S7. For example, if it is determined in step S4 that an electrode plate abnormality has occurred, it is determined as Y and the process proceeds to step S5.

In step S5, the CPU 10a determines that an electrode plate abnormality has occurred, and notifies, for example, an ECU (not shown) of the occurrence of the electrode plate abnormality.

In step S6, the ECU executes a process when an electrode plate abnormality occurs. More specifically, the ECU can suspend execution of a so-called idling stop that temporarily stops the engine 17 during idling in order to reduce fuel consumption. As a result, it is possible to prevent the engine 17 from being unable to restart when recovering from idling stop. In addition, the ECU can notify the user by displaying a message prompting replacement of the rechargeable battery 14 promptly on a display unit (not shown) because the electrode plate abnormality has occurred. Of course, processing other than these may be executed.

In step S7, the CPU 10a determines whether or not to repeat the process. If it is determined to repeat the process (step S7: Y), the process returns to step S1 to repeat the same process as described above. If so (step S7: N), the process is terminated. It should be noted that the repetition in step S7 is executed according to the time window described above.

FIG. 6 is a flowchart for explaining the details of the processing in step S1 of FIG. More specifically, FIG. 6 is a flowchart for explaining the flow of processing executed to optimize the coefficients A1 to A6 of the mathematical model shown in Equation (1). The process shown in FIG. 6 is performed when the engine 17 is stopped and the voltage of the rechargeable battery 14 is stabilized (for example, when a predetermined time (for example, several hours) has passed since the engine 17 was stopped). executed. When the process of the flowchart shown in FIG. 6 is started, the following steps are executed.

In step S10, the CPU 10a executes a process of initializing the coefficients of equation (1). More specifically, the CPU 10a sets the value of Δt (eg, 10 seconds). In addition to the values described above, the values to be initialized include, for example, the number of sample acquisitions Ns and the stabilization time Tx required for the voltage of the rechargeable battery 14 to stabilize. For example, initial setting values such as ΔTs=10 (seconds), Ns=60 (pieces), and Tx=100000 (seconds) may be used. In this case, although it is possible to predetermine an appropriate fixed initial setting value according to the characteristics of the rechargeable battery 14, the initial setting value may be appropriately changed according to the operating conditions or the like.

In step S11, CPU 10a executes a process of measuring voltage Vn and current In of rechargeable battery . More specifically, the CPU 10a sequentially reads the output signals of the voltage detection section 11 and the current detection section 12 to obtain a plurality of voltage values and current values of the rechargeable battery 14 on the time axis. As a result, Ns voltage and current values measured at the sampling period ΔTs are sequentially obtained. The CPU 10a sequentially stores the acquired voltage values and current values in the RAM 10c as data 10ca, and reads them out as necessary. In the following description, the n-th (n=1, 2, 3, .

At step S12, the CPU 10a executes a process of initializing the coefficients A1 to A6 of Equation (1). More specifically, the CPU 10a reads out the initial values obtained in advance by experiments and stored in the ROM 10b, and sets them to the coefficients A1 to A6 of the equation (1).

In step S13, the CPU 10a calculates VFn(n, In) by substituting n and In into Equation (1). As a result, the value of VFn(n, In), which is the result of calculation by the mathematical model, is obtained.

In step S14, the CPU 10a applies VFn(n, In) obtained by the formula (1) in step S13 and the voltage value Vn obtained from the voltage detection unit 11 in step S11 to the aforementioned formula (2), Get the value of Rn.

In step S15, the CPU 10a calculates the partial differential term corresponding to each of the coefficients A1 to A6 based on the above equation (3).

In step S16, the CPU 10a calculates the matrix B based on the formula (4) described above.

In step S17, the CPU 10a calculates dR represented by equation (5) using Rn determined by equation (2) in step S15 and the partial differential term determined by equation (3) in step S16. .

In step S18, the CPU 10a applies the matrix B determined by equation (4) in step S16 and dR determined by equation (5) in step S17 to equation (6) to calculate the difference dd.

In step S19, the CPU 10a determines whether the six differences dd1 to dd6 corresponding to the coefficients A1 to A6 obtained in step S18 satisfy equation (7), and satisfies equation (7). If so (step S19: Y), the process returns to the process of FIG. 5 assuming that the optimum solution is obtained. Otherwise (step S19: N), the process proceeds to step S20. Note that Th on the right side of Equation (7) can be set to, for example, 10 ⁻¹² . Of course, other values may be used.

In step S20, the CPU 10a updates the coefficients A1 to A6 based on the above-described equation (8), returns to step S13, and repeats the same processing as described above.

Through the above processing, the coefficients A1 to A6 of Equation (1) can be optimized.

Next, details of the electrode plate abnormality determination process shown in step S3 of FIG. 5 will be described with reference to FIG. When the process of the flowchart shown in FIG. 7 is started, the following steps are executed.

In step S30, the CPU 10a calculates the absolute value ΔVn of the difference between the value VFn(n, In) of the mathematical model shown in Equation (1) and the voltage measurement value Vn based on Equation (9) below. . As a result, for example, Ns (eg, 60) absolute values ΔVn of differences can be obtained.

・・・（９）

... (9)

In step S31, the CPU 10a sets the largest ΔVn among the ΔVn obtained in step S30 as ΔVn1.

In step S32, the CPU 10a sets the second largest ΔVn among the ΔVn obtained in step S30 as ΔVn2.

In step S33, the CPU 10a determines whether or not ΔVn1 is greater than a predetermined threshold value Th2. If ΔVn1>threshold value Th2 is satisfied, the process proceeds to step S34. )do. Note that the threshold value Th2 is a value that cannot be obtained when the rechargeable battery 14 is normal, but a value that can be obtained when the electrode plate is abnormal can be obtained, for example, by actual measurement.

In step S34, the CPU 10a determines whether or not ΔVn2 is greater than a predetermined threshold value Th2. If ΔVn2>threshold value Th2 is satisfied, the process proceeds to step S35. return). Note that the threshold used in step S34 may be a threshold smaller than the threshold used in step S33.

In step S35, the CPU 10a returns to the process of FIG. 5 after determining that the electrode plate abnormality has occurred.

According to the above process, it is possible to detect the occurrence of electrode plate abnormality of the rechargeable battery 14 using the mathematical model in which the coefficients A1 to A6 are optimized by the process of FIG.

In addition, in the processing of FIG. 7, the determination is made based on the two values ΔVn1 and ΔVn2, but the determination may be made based on three or more values. According to such a method, it is possible to reduce the occurrence of erroneous determinations.

(C) Description of Modified Embodiment The above-described embodiment is merely an example, and needless to say, the present invention is not limited to the case described above. For example, in the above embodiments, a quadratic function is used as the exponential decay function, but an exponential decay function of a first order or less or a third order or more may be used. For example, the following equation (10) can be used as the fourth-order exponential decay function.

・・・（１０）

(10)

Also, the order of the exponential decay function may be changed according to the timing of starting the measurement. Specifically, as shown in FIG. 8, when the measurement start timing is 0 to 10 seconds, an exponential decay function including all of the first to fourth terms is applied. In such an initial stage, each term of the exponential decay function has a relatively large effect. This is because it is necessary to

On the other hand, at the timing 10 seconds after the start of the arithmetic processing, the first term of the exponential decay function decays to a negligible extent, so the exponential decay function with the second to fourth terms except for the first term apply. In addition, at the timing 60 seconds after the start of the arithmetic processing, the first and second terms of the exponential decay function are attenuated to a negligible level. Apply an exponential decay function of inclusive form. Furthermore, at the timing 600 seconds after the start of the arithmetic processing, the 3rd term in addition to the 1st and 2nd terms of the exponential decay function decays to a negligible level, so only the 4th term except for the 1st to 3rd terms Apply an exponential decay function of the form containing .

Also, in the above embodiment, an exponential decay function is used, but other than this, for example, an inverse proportional curve or the like may be used.

Further, in the flowchart shown in FIG. 7, when both the first largest ΔVn1 and the second largest ΔVn2 exceed the predetermined threshold value Th2, it is determined that the electrode plate is abnormal. However, it is also possible to calculate the absolute value of ΔVn1 which is the largest and the absolute value of ΔVn2 which is the second largest, add the obtained values, and compare the value obtained by the addition with the threshold. Also, the sum of squares may be added instead of the absolute value. Alternatively, values having the same sign of the difference between VFn (n, In) and Vn are selected and added, and the obtained value is compared with a threshold value. may be compared with a threshold value. In short, the electrode plate abnormality can be detected by detecting the case where the calculated value obtained by the mathematical model changes discontinuously compared to the normal state.

Further, in the flowchart shown in FIG. 7, the electrode plate abnormality cannot be detected until the voltage disturbance due to the electrode plate abnormality can be clearly distinguished from the influence of the current. Otherwise, it may not be possible, or detection may be missed. Therefore, in order to avoid such a situation, for example, a first threshold may be set for the divergence or error, and the frequency or the frequency rate of points exceeding this first threshold may be used for determination. .

This is because, as shown in FIG. 9, the error caused by the influence of current tends to occur in the short term at the instant when the current changes suddenly. In the example of FIG. 9, the horizontal axis represents time, and the vertical axis represents the error as the difference value between VFn (n, In) and Vn. , and the error increases accordingly. On the other hand, in the case of electrode plate anomaly, the state of deviation from the mathematical model tends to continue for a certain period of time. The frequency of those exceeding .

FIG. 10 is a diagram showing the difference value between VFn(n, In) and Vn measured as data over a predetermined period, and showing the measured data stratified according to the voltage. In FIG. 10, the hatched rectangles indicate the case where the electrode plate is abnormal, and the non-hatched rectangles indicate the case where the electrode plate is not abnormal. As shown in FIG. 10, when there is no electrode plate abnormality, most of the data falls within the range of -2 V and +2 V, which are the first threshold values, and there is one peak. Most of the data do not fall within the range of the first threshold and have multiple peaks.

Therefore, if the frequency of the data falling within the range of the first threshold is equal to or greater than the predetermined threshold, it is determined that there is no electrode plate abnormality, and otherwise, it is determined that there is an electrode plate abnormality. good. Alternatively, when the number of data peaks is one, it may be determined that there is no electrode plate abnormality, and when the number of data peaks is plural, it may be determined that there is an electrode plate abnormality. In addition, instead of the frequency indicating the number of data, a frequency rate indicating the ratio of data existing outside a predetermined range (range of -2 to +2) among all data may be used for determination. .

FIG. 11 is a flow chart showing an example of abnormality determination processing using a frequency rate f that indicates the ratio of data existing within a predetermined range (range of -Th3 to +Th3) among all data. When the process of the flowchart shown in FIG. 11 is started, the following steps are executed.

In step S50, the CPU 10a calculates ΔVn=VFn(n, In)−Vn, which is the difference between VFn(n, In) and Vn.

In step S51, the CPU 10a stratifies ΔVn calculated in step S50 for each value. As a result, for example, as shown in FIG. 10, data stratified by ΔVn values are obtained.

In step S52, the CPU 10a calculates the frequency f at which ΔVn stratified in step S51 is within a predetermined range, ie, −Th3<ΔVn<+Th3. For example, as shown in FIG. 10, the frequency f of ΔVn within the range of −2<ΔVn<+2 is calculated.

In step S53, the CPU 10a determines whether or not the frequency f calculated in step S52 is less than a predetermined threshold value Th4. If f<Th4 is satisfied (step S53: Y), the process proceeds to step S54. At (step S53: N), the process of FIG. 5 is returned (returned). Note that the determination is made based on f1/f2, which is the ratio of the frequency f1 "within" the range -Th3<ΔVn<+Th3 and the frequency f2 "outside" the range -Th3<ΔVn<+Th3. good too.

In step S54, the CPU 10a returns to the process of FIG. 5 after determining that the electrode plate abnormality has occurred. For example, when an electrode plate abnormality occurs, as shown in FIG. 10, the frequency f as the number of data falling within a predetermined range (the range of −2 to +2 in FIG. 10) is less than a predetermined threshold value Th4. Therefore, in such a case, it is determined that the electrode plate abnormality has occurred, and the processing of FIG. 5 is resumed.

Further, as another embodiment for improving detection delays and detection omissions, a method of detecting an electrode plate abnormality based on whether or not the standard deviation of the difference value between VFn(n, In) and Vn is greater than or equal to a predetermined threshold. There is More specifically, when the difference value between VFn(n, In) and Vn is _xi , and the average value of _xi is _xav , the standard deviation s can be obtained by the following equation (11).

・・・（１１）

(11)

If the standard deviation s obtained in this manner is equal to or greater than a predetermined threshold value, it may be determined that an electrode plate abnormality has occurred.

In addition, as another embodiment for improving detection delays and detection omissions, the electrode plate abnormality is detected depending on whether the coefficient of the optimized mathematical model exceeds the range of values that the healthy rechargeable battery 14 can take. There is a way. For example, taking the mathematical model of formula (1) as an example, the coefficients A1 to A5 are the coefficients of the mathematical model expressing voltage that changes over time, and A6 is the coefficient of the mathematical model expressing the effect of current. is the coefficient. More specifically, A1 and A2 correspond to the amount of voltage change over time, A3 and A4 correspond to the rate of change, and A5 corresponds to the base absolute voltage. Also, the factor for current influence of A6 corresponds to the internal resistance of the rechargeable battery 14 .

These are, of course, mathematical models assuming a healthy state (a state in which no electrode plate abnormality has occurred). It is a value within the normal range that reflects the 14 conditions. On the other hand, when the electrode plate abnormality occurs or the voltage fluctuation caused by the electrode plate abnormality occurs, if the coefficients are optimized for the observed voltage and current values, Affected by voltage fluctuations, the coefficient is outside the normal range. By utilizing this, it is possible to detect the occurrence of electrode plate abnormality of the rechargeable battery 14 .

FIG. 12 shows the process of optimizing the coefficients A1-A6 for a healthy rechargeable battery 14, and FIG. 14(A) shows the respective values of the optimized coefficients A1-A6. In FIG. 12, the circles indicate the actually measured voltage of the rechargeable battery 14, and the diamonds indicate the calculated values of the mathematical model.

Also, FIG. 13 shows the process of optimizing the coefficients A1 to A6 of the rechargeable battery 14 with the abnormal plate, and FIG. 14(B) shows the values of the optimized coefficients A1 to A6. ing. In FIG. 13, the circles indicate the measured voltage of the rechargeable battery 14, and the rhombuses indicate the calculated values of the mathematical model. From the comparison of FIGS. 14A and 14B, the hatched coefficients (A1 to A3, A5) in FIG. 14B are significantly different. Therefore, after optimizing the coefficients A1 to A6 by the optimization processing of FIG. may determine that an electrode plate abnormality has occurred.

FIG. 15 is a flowchart showing an example of abnormality determination processing based on coefficients A1 to A6. When the process of the flowchart shown in FIG. 15 is started, the following steps are executed.

In step S70, the CPU 10a executes a process of obtaining the coefficients A1 to A6 obtained by the previous optimization process (optimization process in the normal state). More specifically, the CPU 10a obtains the coefficients A1 to A6 obtained by the previous optimization process and stored in the RAM 10c.

In step S71, the CPU 10a executes a process of obtaining the coefficients A1 to A6 obtained by the optimization process of this time. More specifically, the CPU 10a obtains the coefficients A1 to A6 obtained by the current optimization process and stored in the RAM 10c.

In step S72, the CPU 10a compares the coefficients A1 to A6 obtained by the previous optimization process with the coefficients A1 to A6 obtained by the current optimization process.

In step S73, the CPU 10a determines whether there are, for example, a plurality of coefficients that differ by a predetermined value or more based on the comparison result in step S72. Otherwise (step S73: N), the process returns to the process of FIG. Note that Y may be determined when there is one dissociation, not a plurality, and the dissociation is equal to or greater than a predetermined threshold.

In step S74, the CPU 10a returns to the process of FIG. 5 after determining that the electrode plate abnormality has occurred. For example, when an electrode plate abnormality occurs, as shown in FIG. 14, if a plurality of coefficients differ by a predetermined value or more, it is determined that an electrode plate abnormality has occurred, and the process returns to FIG. .

In the process of FIG. 15, the presence or absence of an abnormality is determined by comparing the actual measured values of the coefficients A1 to A6 in the normal state with the actual measured values of the coefficients A1 to A6 this time. Typical values of coefficients A1 to A6 that are measured in advance and stored in the ROM 10b may be used instead of the actually measured values of A6. Further, if the value obtained by adding the difference values of the coefficients A1 to A6 with predetermined weighting is equal to or greater than a predetermined threshold value, it may be determined that the electrode plate is abnormal.

Further, in the above embodiment, as shown in FIG. 16A, the determination of the electrode plate abnormality is performed with a predetermined time window (for example, a one-hour window) as a delimiter. However, for example, as shown in FIG. 16B, the time windows may be set so that part of the time windows overlap each other, and the electrode plate abnormality may be determined in each time window. Note that, in FIG. 16, adjacent time windows are indicated by a solid line and a broken line in order to improve visibility. Further, as shown in FIG. 16C, time windows that do not overlap each other are set, and, for example, based on the results obtained by optimization processing in a plurality of (for example, two adjacent) time windows, the maximum A board abnormality may be determined. More specifically, the optimized coefficients A1 to A6 are compared in each of a plurality of time windows to determine the plate abnormality, or the difference value between VFn (n, In) and Vn is temporally determined. An abnormality of the electrode plate may be determined based on the change. Furthermore, although a plurality of time windows are provided in the example shown in FIG. 16A, one time window may be provided.

The above is an embodiment using a mathematical model assuming a state in which substantially no current flows through the rechargeable battery 14 after the engine 17 has stopped. good too. For example, the mathematical model disclosed in Japanese Patent No. 4532416 can be used. As shown in FIG. 17A, this mathematical model includes nine element constants RΩ, Ra1, Ra2, Ra3, Ra4, Ra5, Rb1, Rb2, and Rb3 representing resistors, and eight element constants representing capacitors. 1 is an equivalent circuit model of rechargeable battery 14 having Ca1, Ca2, Ca3, Ca4, Ca5, Cb1, Cb2, Cb3. As a method of optimizing such an equivalent circuit model, the rechargeable battery 14 is, for example, pulse discharged by the discharge circuit 15 shown in FIG. Element constants are optimized by using an extended Kalman filter, a neural network, or the like. Then, by comparing the element constants of the rechargeable battery 14 in a healthy state, it is possible to detect an abnormal electrode plate. As a comparison method, as in the case described above, the value of the voltage or current during discharge by the discharge circuit 15 is compared with the value of the voltage or current obtained by the equivalent circuit model. By comparing with a constant, occurrence of electrode plate abnormality can be detected. In addition, instead of causing pulse discharge by the discharge circuit 15, when current is flowing through the load 19 or the starter motor 18, the voltage and current are measured, and the elements constituting the equivalent circuit shown in FIG. may be optimized.

The equivalent circuit shown in FIG. 17A is an example, and other equivalent circuits may be used. For example, as shown in FIG. 17B, an equivalent circuit composed of RΩ, Ra1, and Ca1 may be used. Alternatively, as shown in FIG. 17(C), an equivalent circuit consisting only of RΩ and Ra1 may be used. Alternatively, at least one element of the equivalent circuits shown in FIGS. 17A to 17C may be used. Furthermore, it is desirable to change the mathematical model according to the equivalent circuit.

Alternatively, different mathematical models may be used when the engine 17 is stopped and when it is in operation, and the electrode plate abnormality may be determined based on each mathematical model. For example, when the engine 17 is stopped (when the rechargeable battery 14 is not charged), the determination is based on the above-described formula (1), and when the engine 17 is operating (when the rechargeable battery 14 is charged), The electrode plate abnormality may be determined based on the mathematical model of Japanese Patent No. 4532416 mentioned above.

In the example of FIG. 1, the control unit 10 controls charging and discharging, but an ECU (not shown) receives data such as SOC or OCV from the control unit 10, and the ECU controls based on these values. You may make it

Further, although the above embodiment does not describe the influence of temperature, the temperature detected by the temperature detection unit 13 may also be taken into consideration when measuring the voltage. For example, the measured voltage may be corrected to the voltage at the standard temperature, and the voltage may be calculated based on the voltage obtained as a result of the correction.

In the above embodiment, the rechargeable battery state detection device 1 has the discharge circuit 15 and the temperature detection section 13, but the discharge circuit 15 and the temperature detection section 13 may be omitted.

Also, the flowcharts shown in FIGS. 5, 6, 7, 11, and 15 are examples, and the present invention is not limited to the processing of these flowcharts.

1 Rechargeable Battery State Detector 10 Control Unit 10a CPU
10b ROM
10c RAM
10d communication unit 10e I/F
REFERENCE SIGNS LIST 11 voltage detector 12 current detector 13 temperature detector 14 rechargeable battery 15 discharge circuit 16 alternator 17 engine 18 starter motor 19 load

Claims (14)

In a rechargeable battery state detection device that detects the state of a rechargeable battery,
detection means for detecting the voltage of the rechargeable battery based on the signal output from the voltage detection section;
optimization means for optimizing by adjusting coefficients of a predetermined mathematical model representing the internal state of the rechargeable battery based on the voltage value detected by the detection means;
Judgment for determining whether or not an electrode plate abnormality exists in the electrolyte of the rechargeable battery based on the value calculated by the mathematical model in which the coefficient is adjusted by the optimization means or the value of the coefficient means and
an output means for outputting a determination result by the determination means;
has
Detecting the state of a rechargeable battery, wherein the mathematical expression model includes a term indicating a voltage change over time of the rechargeable battery and indicates a voltage change after charging/discharging of the rechargeable battery is stopped. Device. 2. The rechargeable battery state detection device according to claim 1, wherein the electrode plate abnormality is an abnormality in which electrical characteristics of the rechargeable battery change discontinuously. 3. The rechargeable battery condition detecting device according to claim 2, wherein the plate abnormality is a short circuit in a partial area of the plate. 4. The rechargeable battery state detection device according to claim 1, wherein the mathematical model includes a term indicating a voltage change accompanying a change in current of the rechargeable battery. The mathematical model representing a voltage change after charging and discharging of the rechargeable battery is stopped,
5. The rechargeable battery state detection device according to any one of claims 1 to 4, characterized by being an exponential decay function or an inverse proportional function having one or more terms for calculating the voltage value of the rechargeable battery. . The determination means determines the presence or absence of the electrode plate abnormality based on the value calculated by the mathematical model and the actual measurement value of the voltage of the rechargeable battery at two or more different times. Item 2. The rechargeable battery state detection device according to item 1. The determination means determines whether each of the difference values between the value calculated by the mathematical model and the actual measurement value of the voltage of the rechargeable battery is equal to or greater than a predetermined threshold at two or more different times, or the difference value or the sum of squares of the difference values is greater than or equal to a predetermined threshold value, it is determined that the electrode plate is abnormal. When the frequency or the frequency rate of the difference between the value calculated by the mathematical model and the measured voltage of the rechargeable battery is outside a predetermined range is equal to or greater than a predetermined threshold, the determination means 7. The rechargeable battery state detection device according to claim 6 , wherein the state detection device determines that the electrode plate is abnormal. The determining means determines that the electrode plate is abnormal when the standard deviation of the difference between the value calculated by the mathematical model and the measured voltage of the rechargeable battery is equal to or greater than a predetermined threshold. 7. The rechargeable battery condition detection device according to claim 6 . The determining means determines that the electrode plate is abnormal when a plurality of peaks exist in the distribution of difference values between the value calculated by the mathematical model and the measured voltage value of the rechargeable battery. 7. The rechargeable battery condition detection device according to claim 6 . 2. The rechargeable battery state detection device according to claim 1, wherein said determination means determines said plate abnormality based on at least one coefficient of said mathematical model optimized by said optimization means. If at least one coefficient of the mathematical model optimized by the optimizing means differs from a coefficient optimized in the past by a predetermined threshold or more, the determination means determines that the electrode plate 12. The rechargeable battery state detection device according to claim 11 , characterized in that it determines that there is an abnormality. The rechargeable battery is mounted on a vehicle,
The output means notifies a higher control device mounted on the vehicle of the determination result by the determination means,
13. Any one of claims 1 to 12 , wherein the control device that has received the notification issues a warning to the user based on the determination result and/or suspends execution of an idling stop that stops the engine during idling. 1. A rechargeable battery state detection device according to claim 1. In a rechargeable battery state detection method for detecting a state of a rechargeable battery,
a detection step of detecting the voltage of the rechargeable battery based on a signal output from a voltage detection unit;
an optimization step of optimizing by adjusting coefficients of a predetermined mathematical model representing the internal state of the rechargeable battery based on the voltage value detected in the detection step;
Determining whether or not there is an electrode plate abnormality in the electrolyte of the rechargeable battery based on the value calculated by the mathematical model in which the coefficient is adjusted in the optimization step or the value of the coefficient a step;
an output step of outputting the determination result in the determination step;
has
Detecting the state of a rechargeable battery, wherein the mathematical expression model includes a term indicating a voltage change over time of the rechargeable battery and indicates a voltage change after charging/discharging of the rechargeable battery is stopped. Method.

Images