JP2017129402A - Battery state estimation method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately estimate an open-circuit voltage OCV used in estimating a battery state, without using an advanced algorithm and irrespective of an elapsed time from ignition-off.SOLUTION: In steps S101, S102, an elapsed time from previous IGN-off is referenced, and if the elapsed time is 6 hours or more, step 106 is entered in order to find the OCV on the basis of latest V, T. In step S106, a 25°C equivalent OCV (OCV 25) is found on the basis of the latest V, T and a preset dark current value. If the elapsed time is 2 hours and over to less than 6 hours, a step S103 is entered, and a third-order Yule-walker equation is determined on the basis of the time series of voltages V in order to estimate the convergent value of voltages V as OCV by an autoregression (AR) model, and its AR coefficient is found. In step S105, a 25°C equivalent OCV is found as the convergent value of voltages V on the basis of the calculation result of the AR coefficient.SELECTED DRAWING: Figure 12

Description

  The present invention relates to a battery state estimation method and apparatus, and more particularly to a battery state estimation method and apparatus suitable for estimating the state of an in-vehicle secondary battery.

  In recent years, a vehicle having an idling stop system (ISS) function and a vehicle having a function of charging regenerative energy during braking have attracted attention from the viewpoint of environmental protection through energy saving and exhaust gas reduction.

  Such a vehicle is called a μHEV or a micro-hybrid, and higher accuracy is required for the state estimation so that the mounted secondary battery can maintain a state corresponding to each function. As an index representative of a battery state, SOC (State of Charge) indicating a charge state and a charge rate and SOH (State of Health) indicating a health state and a degree of deterioration are known.

  Patent Document 1 discloses a technique for estimating SOC based on an open circuit voltage (OCV). Patent Document 2 discloses a technique for estimating SOH based on OCV and internal resistance. Patent Document 3 discloses a technique for obtaining the SOC from the differential internal resistance (the slope of the voltage-current straight line) and a technique for integrating the charge / discharge current.

JP-A-4-264371 Japanese Unexamined Patent Publication No. 2006-10601 JP-A-6-242193

  The SOC of the battery can be estimated from the terminal voltage, but it is necessary to estimate from the open circuit voltage OCV in order to estimate the SOC with high accuracy. As a method for obtaining the SOC while the vehicle is running, it is common to use a current integration method in which the charge / discharge current of the battery is continuously integrated and the change in SOC from the initial SOC is calculated. However, in the current integration method, it is necessary to know the full charge capacity of the battery. However, since the full charge capacity is reduced due to the deterioration of the battery, a calculation for deterioration compensation is separately required.

  In response to such technical problems, in recent years, a method for estimating the OCV by obtaining an overvoltage generated during operation is known. According to this estimation method, the OCV can be estimated without knowing the initial SOC and the full charge capacity.

  However, the overvoltage is caused by the internal resistance of the battery. If the internal resistance is to be determined accurately, the internal resistance is represented by an equivalent circuit model, and an advanced algorithm is required to sequentially estimate the circuit parameters. .

  An object of the present invention is to solve the above technical problem and to provide a battery state estimation method and apparatus capable of accurately estimating the open circuit voltage OCV used for battery state estimation without using an advanced algorithm.

  In order to achieve the above object, the present invention provides a battery state estimation device that estimates the state of an in-vehicle battery based on the open circuit voltage OCV of the battery, means for measuring the terminal voltage of the battery at a predetermined period, Means for measuring an elapsed time since the ignition switch is turned off, means for adopting the terminal voltage as an open-circuit voltage OCV if the elapsed time is equal to or longer than a predetermined first reference time, and an elapsed time less than the first reference time Then, there is provided means for estimating a convergence value of the terminal voltage predicted based on the time series of the terminal voltage of the battery as the open circuit voltage OCV.

  According to the present invention, the following effects are achieved.

  (1) Since the OCV estimation algorithm is made different according to the elapsed time from the previous ignition switch off, the OCV whose relationship with the terminal voltage V changes according to the elapsed time is changed regardless of the elapsed time. It becomes possible to estimate accurately.

  (2) Since the OCV is estimated as the convergence value of the terminal voltage V predicted based on the time series of the battery voltage V, the OCV can be accurately estimated.

  (3) Focusing on the fact that the battery terminal voltage V decreases exponentially with time and converges to OCV, and a multi-order yule-walker equation is determined based on the time series of voltage V, and its AR Since the coefficient is obtained and the OCV is predicted, OCV estimation suitable for the time variation characteristic of the terminal voltage V is possible.

It is the figure which showed the structure of (micro | micron | mu) HEV by which the battery controller (BC) for lead secondary batteries which concerns on one Embodiment of this invention is mounted. It is the block diagram which showed the structure of one Embodiment of the battery controller BC. It is a transition diagram regarding the operation mode of the battery controller BC. It is the figure which showed the typical change of the battery voltage V and an IGN signal in the time of parking and driving | running | working of (mu) HEV. It is the main flow which showed the main operation | movement of the battery controller BC. It is a functional block diagram of CPU and BCIA that realize the main operation of the battery controller BC. It is the flowchart which showed the content of the measurement command during sleep. It is the flowchart which showed the procedure of the process during IGN ON. It is the figure which showed an example of the 25 degreeC conversion table of internal resistance R. FIG. It is the flowchart which showed the procedure of 10ms timer interruption processing. It is the flowchart which showed the procedure of 2ms timer interruption processing. It is the flowchart which showed the procedure of OCV estimation processing. It is the figure which showed an example of the 25 degreeC conversion table of the open circuit voltage OCV. It is the figure which showed the procedure which calculates the convergence value of OCV25 as time infinity by AR. It is the flowchart which showed the procedure of SOC initial value and SOH calculation. It is a functional block diagram of CPU and BCIA which implement | achieve the procedure of SOC initial value and SOH calculation. It is the figure which showed the relationship between SOH, SOC, and OCV25 which measured the lead secondary battery of the same specification beforehand in various deterioration states. It is the figure which showed the relationship of R25, SOC index x, and SOH which measured the lead secondary battery of the same specification beforehand in various deterioration states. It is the flowchart which showed the procedure of ISS propriety determination. It is the flowchart which showed the procedure of battery replacement recommendation determination. It is a functional block diagram of CPU and BCIA which realize battery replacement recommendation judgment. It is the flowchart which showed the determination procedure of battery replacement determination threshold value Vst_th.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a configuration of a μHEV in which a battery controller (BC) for a lead secondary battery according to an embodiment of the present invention is mounted. Here, a configuration unnecessary for explaining the present invention is illustrated. It is omitted.

  Drive wheels W are connected to the output shaft of the engine ENG 1 via a transmission 2, and a starter / alternator ISG (Integrated Starter Generator) 4 is connected via an alternator belt 3. The electric power consumed by the ISG 4 as a starter is supplied from the battery 5, and the electric power generated as an alternator is consumed by each auxiliary machine and charged to the battery 5. The battery controller BC6 calculates a state of charge (SOC) and a state of health (SOH) of the battery 5, and notifies the engine control unit ECU7 of the state.

  FIG. 2 is a block diagram showing a configuration of an embodiment of the battery controller BC6, a battery controller IC (BCIA) 9 as a measurement IC, a CPU 8 as a calculation IC, and a CAN (Controller Area Network) for communication. ) The driver 10 is a main component, and these are connected to each other via a serial bus 11. The CAN driver 10 is a type of communication protocol for control equipment, and communicates with the ECU 7 to provide various battery information.

  The BC6 of the present invention measures the voltage V, current I, and temperature T of the lead secondary battery 5 for μHEV having an idling stop function and a regenerative charging function, and based on these measurement results upon input of an ignition signal. The SOC and SOH of the battery 5 are calculated and output to the CAN driver 10. The μHEV ECU 7 executes various controls including idling stop, regenerative charging or battery diagnosis based on the information acquired from the CAN driver 10 of BC6.

  The BCIA 9 is connected to a temperature sensor 12 that detects the temperature T of the battery 5, a shunt resistor 13 that measures the charge / discharge current I of the battery 5, and a positive electrode and a negative electrode for measuring the terminal voltage V of the battery 5. Are connected. An ignition (IGN) switch 14 is connected to the interrupt terminal INT of the CPU 8.

  In this embodiment, during parking, the CPU 8 stands by in the sleep state and the BCIA 9 stands by in order to reduce the current consumption of the battery 5. Meanwhile, the BCIA 9 is activated by a timer interrupt every minute, measures the V, I, and T of the battery 5 and accumulates their time series data. The CPU 8 returns from the sleep state in response to the IGN ON signal, and collects time series data of V, I, and T measured and accumulated by the BCIA 9 during the sleep.

  FIG. 3 is a transition diagram regarding the operation mode of BC6, which is roughly divided into an operation mode and a maintenance mode, and the operation mode has a parking mode and a traveling mode. In the operation mode, when the IGN on is detected in the parking mode, the mode transits to the running mode, and when the IGN off is detected in the running mode, the mode transits to the parking mode. In addition, when a CAN packet requesting the transition to the maintenance mode is received in the running mode, the mode transits to the maintenance mode, and when an operation for ending the maintenance mode is detected in the maintenance mode, the mode transits to the running mode.

  FIG. 4 is a diagram illustrating typical changes in the battery voltage V and the IGN signal during parking and running of the μHEV. At time t1, IGN is turned on and power supply to each auxiliary machine is started.At time t2, the battery voltage V is reduced at a stroke due to the cranking load.When the engine start is completed at time t3, constant voltage charging is started, and time t4 The idling is stopped by the ISS function, cranking is performed by the ISS function at time t5, and the IGN is turned off at time t6.

  FIG. 5 is a main flow showing the main operation of the BC 6, and FIG. 6 is a functional block diagram of the CPU 8 and the BCIA 9 for realizing the main operation. In the present embodiment, power supply from the battery 5 is started by being mounted on the vehicle, or processing is started after restart by CPU reset, and thereafter, each processing is repeated at a predetermined cycle.

  First, referring to FIG. 6, the BCIA 9 receives a measurement command receiving unit 91 that receives a measurement command from the CPU 8, a measurement unit 92 that measures V, I, and T of the battery 5 in response to the measurement command, and a measurement result. A memory 93 for temporarily storing a time series of the above, a measurement cycle setting unit 94 for setting the measurement cycle of the measurement unit 92 to a cycle according to the measurement command, and a measurement result response unit for responding the time series of measurement results to the CPU 8 95.

  The CPU 8 has a measurement command unit 81 during sleep for issuing a measurement command during sleep for requesting the BCIA 9 to perform measurement in the first cycle during its sleep, and BCIA 9 performs measurement in the second cycle during travel after returning from sleep. In-travel measurement command unit 82 that issues a measurement command during travel for requesting to, a measurement result acquisition unit 83 that acquires each measurement result from BCIA 9, and a battery state estimation that estimates the battery state based on each acquired measurement result And a running measurement command canceling unit 85 that cancels the running measurement command before the sleep transition.

  Referring to FIG. 5, in step S1, initialization processing is performed in CPU 8, and various reference constants stored in flash memory (not shown) are developed on RAM (not shown). In addition, the power-on reset flag Fpor is set. In step S2, the presence / absence of an external interrupt to the CPU 8 is determined. In this embodiment, the IGN switch 14 is connected to the INT terminal of the CPU 8, and the process proceeds to step S9 while maintaining the sleep state until an interrupt request to the CPU 8 is generated due to the IGN being turned on.

  In step S9, it is determined whether a timer interrupt with a period of 60 seconds has occurred in BCIA 9, and if no timer interrupt has occurred, the process returns to step S2. When the timer interrupt occurs, the process proceeds to step S10, and the measurement unit 92 of the BCIA 9 sets the V, I, and T of the battery 5 in response to the measurement command notified from the measurement command unit 81 during sleep to the BCIA 9 in step S7 described later. The time series data is stored in the memory 93.

  Thereafter, an interrupt request to the CPU 8 is generated by turning on the IGN. When this is detected in step S2, the process proceeds to step S3, and the sleep mode of the CPU 8 is canceled. In step S4, the activation time variable TM indicating the elapsed time after the IGN is turned on is reset and new measurement is started. In step S5, an “IGN on process” described later is executed. In step S6, various data calculated in the IGN on-process are written as log information in the flash memory.

  In step S7, a sleep measurement command for causing the BCIA 9 to periodically measure the V, I, and T of the battery 5 is notified from the sleep measurement command unit 81 to the BCIA 9 by a timer interrupt during the sleep of the CPU 8. In step S8, the CPU 8 shifts to the sleep mode, returns to step S2, and waits for the next IGN-on interrupt request.

  FIG. 7 is a flowchart showing the contents of the measurement command during sleep notified to the BCIA 9 in the step S7 of the main flow. The BCIA 9 performs the measurement process in response to the measurement command with a 60-second timer interrupt. Execute in S10.

  In FIG. 7, in step S21, a count value corresponding to the timer interrupt cycle is added to the BCIA timer counter TMbcia. In this embodiment, since the timer interruption period is set to 60 seconds, a count value corresponding to 60 seconds is added.

  In step S22, whether or not the remainder of TMbcia divided by 15 minutes is zero (TMbcia mod 15 = 0) in order to measure the battery V, I, and T in a cycle of 15 minutes during CPU sleep after IGN is turned off. Is determined. If the remainder is zero, the process proceeds to step S23 and subsequent steps, and V, I, and T of the battery 5 are measured 10 times every 20 milliseconds by the measuring unit 92 of the BCIA 9, and the average value is temporarily stored.

  That is, in step S23, the measurement number counter N is reset. In step S24, V, I, and T of the battery 5 are measured. In step S25, the measurement number counter N is incremented. In step S26, the process returns to step S24 until each measurement number counter N = 10, and each measurement of V, I, and T is repeated every 20 milliseconds. In step S <b> 27, an average value of 10 measurement values for V, I, and T is temporarily stored in the memory 93.

  Thus, in this embodiment, during the CPU sleep until the external interrupt request due to the IGN being turned on is detected, the battery V, I, and T are measured 10 times each by the BCIA 9 in a cycle of 15 minutes, and the average value is obtained. Calculated and temporarily stored as time series data of V, I, and T.

  FIG. 8 is a flowchart showing the procedure of the IGN-on process executed in step S5 of the main flow as an interrupt process triggered by IGN being turned on, and is executed until IGN is turned off. The IGN on-process is composed of a main routine for calculating SOC and SOH and two timer interrupt processes (10 ms timer interrupt process and 2 ms timer interrupt process). Each timer interrupt process will be described in detail later with reference to FIGS.

  In the main routine of the IGN ON process, the OCV (open voltage) of the battery 5 is not stable immediately after the IGN is turned off and is a function of SOH and SOC, while the internal resistance R of the battery 5 is stable and the SOH Focusing on the characteristic peculiar to lead-acid batteries that are a function of SOC and SOC, immediately after the first cranking when OCV and R are determined, solve the simultaneous equations of OCV and R to obtain SOH. Obtain the initial SOC value. Thereafter, SOH is fixed and SOC is corrected periodically.

  In FIG. 8, in step S31, the CPU 8 reads provisional values of SOC and SOH. Further, the time series data of V, I, and T measured and accumulated by the BCIA 9 in step S10 during the CPU sleep with the IGN off are taken into the CPU 8 and developed on the RAM.

  In step S32, measurement commands for V, I, and T are output from the running measurement command unit 82 of the CPU 8 to the BCIA 9. In the present embodiment, a measurement command with a 1 millisecond cycle is output for V and I, and a measurement command with a 1 second cycle is output for temperature T. This measurement command is received by the measurement command receiving unit 91 of the BCIA 9, and thereafter, in the BCIA 9, the measurement is repeated at a 1 millisecond cycle for V and I and at a 1 second cycle for T, and the time series data is accumulated. The

  In step S33, the engine start count Nstart is reset. In step S34, 10 ms timer interrupt processing and 2 ms timer interrupt processing are set, and each processing is started. In step S35, OCV estimation processing for estimating the open circuit voltage OCV based on the V, I, and T of the battery 5 measured while the IGN is off is performed, and the 25 ° C. converted value OCV25 is obtained. The OCV estimation process will be described in detail later with reference to FIGS.

  In step S36, V, I, and T measured while the IGN is off are stored in the flash memory in a nonvolatile manner as log data. In step S37, it is determined whether the IGN is still on. If the IGN is on, the process proceeds to step S38, and it is determined based on the voltage / current waveform whether or not the engine has been started. If it is determined that the engine has been started, the process proceeds to step S39, and the engine start count Nstart is incremented. In step S40, the cranking flag Fcrank is set.

  In step S41, it is determined whether the current engine start is the first engine start after the IGN is turned on or the restart by ISS. If it is the first engine start-up, the process proceeds to step S42, the internal resistance R of the battery 5 is calculated based on the current voltage change before cranking, and its 25 ° C. converted value (R25) is obtained.

  In this embodiment, it is a period before cranking after IGN is turned on, and V, I measured at the timing (time t1 to t2 in FIG. 4) when the power supply to the auxiliary equipment is started and the voltage drop occurs. , T, the internal resistance R is calculated, and the 25 ° C. converted value R25 is calculated. R25 is obtained by applying the measurement results of the internal resistance R and the temperature T to the conversion table shown in FIG.

  The calculation timing and calculation method of the internal resistance R are not limited to the above. Each current I measured in the range of -1A to -100A (minus sign means discharge current) at the time of cranking. (t), the voltage V (t) at that time, and the time t may be extracted to obtain a current-voltage correlation (IV characteristic) during cranking, and the internal resistance R may be calculated based on this.

  In step S43, SOH and SOC initial values are obtained based on 25 ° C. converted values R25 and OCV25 of the internal resistance R and the open circuit voltage OCV, as will be described in detail later. In the present embodiment, the SOC initial value and SOH are obtained by solving the simultaneous equations of R25 and OCV25 immediately after the initial cranking in which OCV25 and R25 are determined.

  In step S44, “battery replacement recommendation determination” is performed to determine whether or not the battery 5 should be replaced based on the SOH, the lowest voltage Vst during cranking, and the OCV25, and the result is registered in the replacement recommendation flag Frmd. . This “battery replacement recommendation determination” will be described in detail later with reference to FIG.

  In step S45, a battery replacement determination is performed to determine whether or not the battery has been replaced. In this embodiment, in the cranking immediately after the initialization process, the SOH update value is compared with the previous SOH, and when the SOH is rejuvenated, it is determined that the battery has been replaced, and the battery replacement flag Fchanged is set. The Further, the power-on reset flag Fpor is reset regardless of the determination result.

  In step S46, as a measure against CPU 8 memory shortage and sudden power-off, the data measured and calculated so far is written to the flash memory as an intermediate log and stored in a nonvolatile manner. In step S47, the cranking flag Fcrank is canceled. In step S48, the cranking number counter Ncrank is incremented.

  On the other hand, if it is determined in step S37 that ING is not on, the process proceeds to step S49, where the timer interrupt permission is released. In step S50, the running measurement command notified to BCIA 9 in step S32 is canceled.

  FIG. 10 is a flowchart showing a procedure of a 10 ms timer interrupt process activated by a 10-millisecond timer interrupt in the IGN on process, and a timer interrupt is permitted in step S34 of the ING on process. And is repeatedly executed at a cycle of 10 milliseconds until the permission is canceled in step S49.

  In step S61, the V, I, and T of the battery 5 measured by the BCIA 9 in response to the measurement command during traveling are acquired. In step S62, SOC calculation processing for updating the SOC with the acquired integrated value of the current I is performed. In this embodiment, the current charge amount is obtained by multiplying the average value of the five most recent current values measured at a cycle of 2 milliseconds by the period length (10 milliseconds), and this is added to the previous SOC. By doing so, the SOC is updated to the latest value.

  In step S 63, V, I, and T measured by the BCIA 9 are transmitted to the CAN driver 3. In this embodiment, the voltage V and current I are transmitted to the CAN driver 3 every 10 milliseconds, and the temperature T is transmitted to the CAN driver 3 every 8 seconds. The CAN driver 3 transmits the received V, I, and T to the ECU 7.

  In step S64, it is determined whether cranking is in progress by referring to the cranking flag Fcrank. If cranking is not in progress, the process proceeds to step S65 to attempt to receive a CAN packet. In step S66, it is determined whether or not the shift to the maintenance mode is requested by the received CAN packet.

  If the transition to the maintenance mode is requested, the process proceeds to step S67, the timer interrupt is prohibited, and the BCIA 9 transitions to the sleep mode. In step S68, the maintenance mode is executed. When the maintenance mode ends, in step S69, the prohibition of the timer interrupt is canceled and the BCIA 9 returns from the sleep mode.

  On the other hand, when it is determined in step S66 that the transition to the maintenance mode is not requested, the process proceeds to step S70, and it is determined whether or not it is a failure determination timing. In the present embodiment, the activation time variable TM mod τ2 is calculated so as to make a failure determination at a predetermined period τ2, and if this is zero, the procedure goes to step S71 to make a failure determination. In the present embodiment, the failure determination is made when the measured values of V, I, and T indicate impossible values.

  In step S72, it is determined whether or not it is a full charge determination timing. In the present embodiment, TM mod τ5 is calculated to perform full charge determination at a predetermined period τ5. If this is zero, the process proceeds to step S73 and full charge determination is performed. As a result, if a state in which the voltage V is equal to or higher than a predetermined full charge voltage (for example, 13.5 V) and the charging current I is low (for example, less than 1 A) is continuously observed for a predetermined time or more, the battery is determined to be fully charged. And the SOC is set to 100%.

  In step S74, it is determined whether or not it is the timing for determining whether or not to stop idling. In the present embodiment, TM mod τ4 is calculated so as to make a determination at a predetermined period τ4. If this is zero, the routine proceeds to step S75 where idling stop (ISS) availability determination is performed. This “ISS availability determination” will be described in detail later with reference to FIG.

  In step S76, SOC, SOH, battery replacement recommendation determination result, failure determination result, and ISS availability determination result are sent to CAN. In step S77, 10 milliseconds is added to the activation time variable TM.

  FIG. 11 is a flowchart showing a procedure of a 2 ms timer interrupt process activated by a timer interrupt with a 2-millisecond period in the IGN on process. When the interrupt is permitted in step S34, the process proceeds to step S49. It is executed every 2 milliseconds until the permission is released.

  In step S81, it is determined whether cranking is in progress. In this embodiment, the average duration of cranking is 1 second, in which case a discharge current of several hundreds A is generated, whereas a normal auxiliary current is about 20 A, so 90 A When the above discharge current continues for 2 milliseconds or more, it is determined that cranking is started. As a result, if it is determined that cranking is in progress, the routine proceeds to step S82, where the cranking flag Fcrank is set.

  In step S83, it is determined whether it is within 8 milliseconds after the start of cranking. If it is within 8 milliseconds, the process proceeds to step S84, where the minimum voltage is obtained from the voltages V acquired from BCIA 9, and this is set as the lowest voltage Vst at the time of one cranking.

  On the other hand, if it is determined in step S81 that cranking is not in progress, the process proceeds to step S86, and it is determined whether or not it is before the first cranking after the ignition is turned on. If it is before the first cranking, the process proceeds to step S87, where V and I measured by BCIA 9 are cut out and stored as time series Vt and It associated with the current time t.

  FIG. 12 is a flowchart showing the procedure of the OCV estimation process performed in step S35 of the ignition ON process. In this embodiment, the OCV estimation algorithm is changed according to the elapsed time after the IGN is turned off. Yes.

  In steps S101 and S102, the elapsed time from the previous IGN off is referenced, and if the elapsed time is 6 hours or more, the OCV is calculated based on the latest V and T measured in the 2 ms timer interrupt process. Therefore, the process proceeds to step S106. In step S106, the latest V and T and the preset dark current value are applied to the OCV 25 ° C. conversion table shown in FIG. 13 as an example, so that the OCV (OCV 25) converted to 25 ° C. is The constant term p and the SOH term q are calculated based on (1).

  On the other hand, if the elapsed time is 2 hours or more and less than 6 hours, the process proceeds to step S103, and the OCV is estimated based on V, I, and T measured during parking. In the present embodiment, attention is paid to characteristics inherent to the lead secondary battery in which the battery voltage V decreases exponentially with time and converges to OCV during parking. Then, in order to estimate the convergence value of the voltage V as an OCV by AR (autoregressive model), a third-order yule-walker equation is determined based on the time series of the voltage V as described in detail below, and the AR The coefficient was calculated.

  In the present embodiment, it is assumed that a recurrence formula is created from past time series data V (1), V (2). The recurrence formula is determined by the least square method based on past time series data, and is expressed by the following formula (2) as the sum of m exponential functions.

  Here, n is the time at which the voltage V is to be predicted, a1, a2... Am are intermediate outputs, m is the order of AR, c is an error term, and the input is a past time series V (1), V ( 2) ... V (k) and the final output is V (∞). In step S104, it is determined whether or not the AR coefficient has been successfully calculated. If successful, the process proceeds to step S105.

  In step S105, an OCV converted to 25 ° C. is obtained as the convergence value of the voltage V based on the calculation result of the AR coefficient. FIG. 14 is a diagram showing a procedure for calculating the convergence value V (∞) of the voltage V as time infinite by AR in the steps S103 and S105.

  On the other hand, if it is determined that the calculation of the AR coefficient has failed, or if it is determined in step S102 that the elapsed time from the previous IGN off is less than 2 hours, the process proceeds to step S107, and the OCV calculation is performed. Set the failure flag Focv-fail and end the current process.

  FIG. 15 is a flowchart showing a procedure of “SOC initial value and SOH calculation” performed in step S43 of the ignition ON process. FIG. 16 is a functional block diagram of the BCIA 9 and the CPU 8 that realize the procedure. is there.

  The BCIA 9 includes an internal resistance measuring unit 96 that measures R25 of the battery 5 and an open-circuit voltage measuring unit 97 that measures the open-circuit voltage OCV25. The CPU 8 stores a first equation representing the relationship between OCV25 and SOH and SOC (the following equation (3)), and a second equation representing the relationship between R25 and SOH and SOC (the following equation (4)). The equation storage unit 86 and the measurement results of the R25 and OCV25 are applied to the respective equations, and a solution finding unit 87 for obtaining SOH and SOC as solutions of the simultaneous equations is provided.

  Referring to FIG. 15, in step S201, it is determined whether OCV25 calculation by open-circuit voltage measurement unit 101 is successful based on OCV failure flag Focv-fail, and if calculation of OCV25 is successful, the process proceeds to step S202. In step S202, whether or not the R25 calculation by the internal resistance measuring unit 102 is successful is determined. If the R25 calculation is successful, the process proceeds to step S203.

  In step S203, SOH is obtained by solving the following two simultaneous equations (3) and (4) for OCV25 and R25 in the solution finding unit 87. Each equation is stored in the equation storage unit 86. The SOC is defined by the following equation (5) and takes values from 100-SOH to 100. The SOC index x is current charge amount × 100 / current full charge capacity, and takes a value of 0-100.

  In this embodiment, focusing on the fact that the relationship of the above formula (3) is established for each OCV25 between SOC and SOH, the lead secondary battery of the same specification is measured in advance in various deterioration states, SOH and SOC. And coefficients a, b, and c in the above equation (3) are calculated and registered in advance based on the relationship between OCV 25 and OCV 25 (FIG. 17). The coefficients p and q are the constant term value and the SOH term coefficient value of the OCV25 calculation formula obtained in the OCV estimation process.

  Further, in the present embodiment, focusing on the fact that R25 is a function of x and SOH, which is an index of SOC, lead secondary batteries of the same specification are preliminarily measured in various deterioration states, R25, SOC index x and SOH. Based on the relationship (FIG. 18), the above equation (4) is obtained in advance.

  In step S204, the power-on-reset flag Fpor is referred to. If the flag Fpor = 1, the process proceeds to step S205 to set “100%” in SOH, and then proceeds to step S206. On the other hand, if the flag Fpor is not 1, the process immediately proceeds to step S206, and the SOC is calculated based on the following equations (6) and (7).

  If it is determined in step S201 that the OCV25 calculation has failed, the process proceeds to step S207, and the previous SOH is registered as the current SOH. However, SOH = 100% is registered immediately after power-on reset.

  FIG. 19 is a flowchart showing a procedure for determining whether or not the ISS is performed in step S75 of the 10 ms timer interrupt process.

  In step S161, it is determined whether the SOC and SOH satisfy the idling stop prohibition requirement. In the present embodiment, it is determined whether the SOC is equal to or less than the determination threshold SOCiss for determining whether or not to stop idling, or whether the current SOH is less than 50%. If any of the requirements is satisfied, the process proceeds to step S165, the ISS prohibition flag Fiss_ng is set, and the process ends.

  On the other hand, if neither requirement is satisfied, the process proceeds to step S162, and it is determined based on the ISS prohibition flag Fiss_ng whether or not idling stop is prohibited. If it is not prohibited, the process proceeds to step S163, and it is determined whether or not the minimum voltage Vst during cranking satisfies the requirement for idling stop prohibition. In this embodiment, if the minimum voltage Vst is equal to or less than the determination threshold Vst_iss for determining whether or not to stop idling, the process proceeds to step S165, the ISS prohibition flag Fiss_ng is set, and the process ends.

  On the other hand, if the minimum voltage Vst exceeds the determination threshold value Vst_iss, the process proceeds to step S164, and it is determined whether or not the voltage V satisfies the idling stop prohibition requirement. In the present embodiment, if the voltage V is equal to or less than the determination threshold Viss (for example, 11.5 V) for whether or not idling stop is possible, the process proceeds to step S165, the ISS prohibition flag Fiss_ng is set, and the process ends.

  On the other hand, if it is determined in step S162 that idling stop is prohibited, the process proceeds to step S166, and it is determined whether continuous charging is in progress. If the battery is continuously charged, the process proceeds to step S167, where the ISS prohibition flag Fiss_ng is reset (idling stop is permitted), and the process ends.

  FIG. 20 is a flowchart showing the procedure for battery replacement recommendation determination executed in step S44 of the IGN-on process, and FIG. 21 is a functional block diagram of the CPU 8 and BCIA 9 for realizing battery replacement recommendation determination. In the present embodiment, the battery replacement recommendation is determined based on the SOH and the lowest cranking voltage Vst only during the first cranking.

  First, referring to FIG. 21, in BCIA 9, the lowest voltage detecting unit 98 detects the lowest voltage Vst from the time series of the battery voltage V measured by the measuring unit 92 during cranking of the engine. In the CPU 8, the usage year accumulation unit 88 calculates the cumulative usage year of the battery 5. The replacement determination threshold value determination unit 89 determines the replacement determination threshold value Vst_th based on the minimum voltage Vst and the usage period. In this embodiment, as will be described in detail later, the replacement determination threshold value Vst_th is set higher as the use period becomes longer. When the minimum voltage Vst falls below the replacement determination threshold value Vst_th, the replacement time determination unit 90 determines the replacement time of the battery.

  Referring to FIG. 20, in step S171, battery replacement recommendation flag Fchange is reset. In step S172, SOH is compared with a predetermined replacement determination threshold value SOH_th. If SOH <SOH_th, the routine proceeds to step S173, where the battery replacement recommendation flag Fchange is set. In step S174, the 25 ° C. converted value Vst25 of the lowest voltage Vst during cranking is calculated as a function of Vst, T and OCV25, and is temporarily stored as a result of the current cycle.

  In step S175, the average value Vst0 is calculated based on the most recent Vst25. In step S176, as will be described in detail later with reference to FIG. 22, the replacement determination threshold value Vst_th for Vst is calculated by the replacement determination threshold value determination unit 89 as a function of Vst0 and OCV25. In step S177, the replacement time determination unit 90 compares the Vst and Vst_th. If Vst <Vst_th, the process proceeds to step S178 where the battery replacement recommendation flag Fchange is set.

  FIG. 22 is a flowchart showing a procedure for determining the replacement determination threshold value Vst_th. In this embodiment, although the lowest voltage Vst during cranking is one representative index for estimating the battery replacement timing, it has been newly found that the influence of aging deterioration is difficult to be reflected, and aging deterioration is progressing. For batteries with high possibility, replacement judgment is made at a higher threshold.

  In step S181, the standard value of Vst_th is calculated as a function of Vst0 and OCV25. In steps S182 and S183, the service life of the battery 5 determined by the service year type part 88 is determined. If the service life is 7 years or more, the process proceeds to step S184, and Vst_th is set as shown in the following equation (8). It is calibrated to the value with 0.2V added. Such threshold calibration is based on the consideration that the influence of aging deterioration is almost constant regardless of the years of use in a battery having an age of 7 years or more.

  If the number of years of use is 3 years or more and less than 7 years, the process proceeds to step S185, and Vst_th adds a value obtained by multiplying the number of years used by 3 years by 0.05V as shown in the following equation (9). It is calibrated to the value. Such threshold calibration is based on the consideration that the influence of aging deterioration is proportional to the age of use in a battery having a service life of less than 7 years.

  Thus, in this embodiment, the aging of the battery is represented by the years of use, and the replacement determination threshold Vst_th is dynamically changed according to the years of use of the battery, so that the replacement determination threshold is set to a more appropriate value. Battery replacement determination can be made more accurately.

  DESCRIPTION OF SYMBOLS 1 ... Engine ENG, 2 ... Transmission, 3 ... Alternator belt, 4 ... Starter and alternator ISG, 5 ... Battery, 6 ... Battery controller BC, 7 ... Engine control unit ECU, 8 ... CPU, 9 ... Battery controller IC (BCIA), 10 ... CAN driver, 11 ... serial bus, 12 ... temperature sensor, 13 ... shunt resistor, 14 ... IGN switch

Claims (6)

In the battery state estimation device that estimates the state of the in-vehicle battery based on the open voltage of the battery,
Means for measuring the terminal voltage of the battery at a predetermined period;
Means for measuring the elapsed time since the last ignition switch off;
Means for adopting the terminal voltage as an open voltage if the elapsed time is equal to or greater than a predetermined first reference time;
And a means for estimating a convergence value of the terminal voltage predicted based on a time series of the terminal voltage of the battery as an open voltage if the elapsed time is less than the first reference time. State estimation device.   The battery state estimating apparatus according to claim 1, wherein the estimating unit predicts an open-circuit voltage using an autoregressive model.   3. The battery state according to claim 2, wherein the estimating means determines a yule-walker equation based on a time series of a terminal voltage of the battery, obtains an AR coefficient thereof, and applies it to the autoregressive model. Estimating device.   The battery state estimation device according to any one of claims 1 to 3, wherein the open-circuit voltage is not estimated if the elapsed time is less than a second reference time shorter than the first reference time.   The battery state estimation apparatus according to claim 1, wherein the battery is a lead secondary battery. In the battery state estimation method for estimating the state of the in-vehicle battery based on the open voltage of the battery,
A procedure for measuring the terminal voltage of the battery at a predetermined cycle;
Procedure to measure the elapsed time since the last ignition switch off,
A procedure for adopting the terminal voltage as an open voltage if the elapsed time is equal to or longer than a predetermined first reference time;
If the elapsed time is less than the first reference time, the battery state estimation method includes an open-circuit voltage and a procedure for a convergence value of the terminal voltage predicted based on a time series of the terminal voltage of the battery. .