JP2011127973A - State of charge estimation method and device for secondary battery - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately estimate the state of charge (SOC) of a secondary battery, prevent overcharge or overdischarge of the secondary battery, and prevent a battery life reduction by preventing damage to the battery. <P>SOLUTION: An estimated value of the SOC is calculated in a consequent part by a simple inference method by calculating a current grade and a voltage grade from a membership function relating to a battery current and a cell voltage in an antecedent part. The minimum value and maximum value are determined among a plurality of battery modules connected in series and displayed and output. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a technique for estimating the SOC (State of Charge) of a secondary battery, and more specifically, to a technique for estimating the remaining amount of electricity of the secondary battery.

  Examples of the rechargeable secondary battery include a lead storage battery, a nickel metal hydride battery, and a lithium ion battery. Applications of such chargeable / dischargeable secondary batteries include secondary batteries mounted on moving bodies such as trains and construction machines, and secondary batteries used in ground storage facilities such as substations and wind power plants. It has become like this.

  In a working machine such as an automobile or a construction machine, it is driven by a generator using an engine as a drive source, a secondary battery that stores electric power generated by the generator, and electric power supplied from the secondary battery. A so-called hybrid system including an electric motor is employed. In such a hybrid system, the remaining capacity of the secondary battery (charged state: SOC) is preferably maintained at about 50% in order to cope with the charge / discharge of the secondary battery according to the load state of the engine. Therefore, it is necessary to detect the SOC.

  In the case of trains and the like, in consideration of the environment, energy is saved by storing regenerative power generated during braking in a ground power storage facility installed at a substation and using it as a power source for the train. In wind power plants, secondary batteries are used to level the power. In order to effectively operate such a ground power storage facility, it is desirable that the SOC of the secondary battery used for the ground power storage facility is maintained within an appropriate range.

  Also, recently, from the viewpoint of harmony of the traffic system seen in LRV, etc., a secondary battery is mounted on a train traveling on a route having an overhead line-less section, and the overhead line-less section is operated using electricity charged in the overhead line section. A system has been proposed. The secondary battery used for such an application must be in a sufficiently charged state so that it can travel in an overhead line-free section, and requires maintenance and management of the SOC.

  In other words, for secondary batteries, the current remaining capacity is estimated by some method for the purpose of making the user recognize the charging time or battery state, etc., and the state of charge of the secondary battery as an energy source is within an appropriate range. Is required to be maintained within.

  As a technique for estimating the SOC of the secondary battery for this purpose, a technique has been proposed in which the charge / discharge current of the secondary battery is integrated and the SOC is estimated based on the integrated value (Patent Document 1). . It is also obtained by sampling current, voltage, and temperature over a predetermined period of time to verify whether the current distribution is appropriate, verifying whether the temperature distribution is appropriate, and integrating the current values. Test whether or not the fluctuation range of the amount of stored electricity is appropriate, linearly approximate the current and voltage sampled when all of them are appropriate, and calculate the electromotive force from the intercept and the internal resistance from the slope Thus, a technique for estimating the SOC of a secondary battery has been proposed (Patent Document 2).

  In addition, a method has been proposed in which the remaining capacity of a nickel metal hydride battery using a hydrogen storage alloy as a negative electrode is estimated from the volume change of the negative electrode accompanying charge / discharge by providing a strain gauge or a pressure sensor on the negative electrode (for example, a patent) Reference 3).

  By the way, since the secondary battery deteriorates in battery performance represented by input / output power and full charge capacity in the process of use, the performance changes in time series. In particular, when input / output (charging / discharging) of electric power exceeding the storage capacity of the battery is performed, the deterioration is rapidly accelerated and the life is significantly reduced. For this reason, it is important to always grasp the state of charge of the secondary battery accurately and detect the limit amount of the charge / discharge capacity of the secondary battery.

  A technique using fuzzy reasoning as a means for predicting battery life has been proposed. For example, Patent Document 4 discloses an apparatus for determining battery life using fuzzy reasoning for UPS. Fuzzy reasoning does not require a strict mathematical model in its logic, and by using expert know-how for reasoning, it is possible to obtain reasoning results that are close to those of experts. For this reason, control or estimation using inference has high robustness, and a sense of security and reliability for inference that cannot be obtained by control based on modern control theory can be expected.

JP 2005-269824 A JP 2000-323183 A JP-A-8-194037 JP-A-6-242192

The method of integrating current values during charging / discharging and estimating based on this integrated value (Patent Document 1) accumulates detection errors based on current values when operated for a long period of time. Due to the characteristic that the SOC history is different, there is a problem that the estimation accuracy of the SOC gradually decreases.

As another method for estimating the SOC, a method for estimating the SOC based on the correlation between the battery voltage and the SOC has also been proposed. However, in this estimation method, the voltage change with respect to the SOC is small as in the nickel-hydrogen secondary battery. When the region exists, it is difficult to accurately estimate the SOC. This tendency is particularly remarkable in the intermediate region (for example, about 20 to 80%) of the SOC generally used in the secondary battery used in the ground power storage facility.

  Furthermore, the method of estimating the SOC when using the battery from the correlation between the internal resistance value of the battery calculated from the current-voltage characteristics during discharge and the SOC prepared in advance (Patent Document 2) The internal resistance value is greatly affected by the charge / discharge history depending on the conditions. This is because the measured values of internal resistance of the battery include reaction overvoltage, diffusion overvoltage, and ohmic loss as resistance components, but the reaction rate-determining process changes depending on the battery charge / discharge history, and is not related to the battery reaction. This is considered to be due to the fact that no capacitor component is included. Therefore, normally, if accurate SOC estimation is performed for various charge / discharge patterns in actual use, the amount of correlation characteristic data to be prepared in advance becomes enormous, and the estimation system becomes complicated. There is a problem.

  In the SOC estimation method disclosed in Patent Document 3, since the shrinkage / expansion of the hydrogen storage alloy is affected by temperature, the relationship between the volume change of the negative electrode and the SOC also changes according to the temperature. Since the SOC is estimated based only on the volume change of the hydrogen storage alloy constituting the negative electrode, it is difficult to estimate the SOC with high accuracy.

  Here, in order to appropriately manage the SOC of the secondary battery in the hybrid system and the ground storage facility, and further to check the state of charge of the secondary battery mounted on the train traveling in the section without an overhead wire, the SOC is used. As described above, it is necessary to accurately grasp.

  If the secondary battery is near the fully charged state or the fully discharged state, further charging or discharging will cause the secondary battery to be overcharged or overdischarged, and this will significantly reduce the life of the secondary battery. In order to prevent overcharge and overdischarge, it is required to grasp the SOC of the secondary battery with high reliability.

  Accordingly, an object of the present invention is to provide a method and apparatus for estimating the state of charge of a battery with high reliability and accuracy by using fuzzy reasoning in order to solve the above-described problem, and to overcharge the secondary battery. The purpose is to prevent discharge and prevent a significant decrease in the life of the secondary battery.

  The problem to be solved by the present invention is as described above. Next, means for solving the problem will be described.

(Claim 1)
In order to achieve the above-described object, the secondary battery state of charge estimation method according to the present invention includes:
A method for estimating a state of charge of the secondary battery in a power storage facility having a secondary battery, the step of measuring the current value of the secondary battery, the step of measuring the voltage value of the secondary battery, A current membership calculation step for calculating a current grade from the current value as a condition part, a voltage membership calculation step for calculating a voltage grade from the voltage value, and the current grade and the current grade as a consequent part And a SOC membership calculation step for calculating SOC membership according to fuzzy inference rules, and a synthesis operation step for combining the calculation result of the antecedent part and the calculation result of the consequent part.

  According to this procedure, the current flowing through the secondary battery and the terminal voltage of the secondary battery are measured, the current grade is calculated from the separately prepared current membership function, and the voltage grade is calculated from the voltage membership function. Has a part.

  And it has the consequent part which calculates the estimated value of SOC from SOC membership function from current grade and voltage grade using the fuzzy reasoning rule prepared separately. A fuzzy inference rule may be defined in the form of if-then-, preferably in the form of a table. Since the fuzzy inference rule is composed of a plurality of rules, the estimated value of SOC can be plural.

  Then, the estimated values of the secondary batteries are calculated by combining the estimated values of the plurality of SOCs calculated in the consequent part and expressing them as profile figures on the membership function.

(Claim 2)
In the secondary battery state of charge estimation method according to the present invention, the antecedent part includes a step of calculating the smallest value of the current grade and the voltage grade as the calculation result, and the SOC membership calculating step in the consequent part. May include the step of using the largest value as the calculation result, and the combining step may include a step of obtaining a center of gravity of the SOC membership.

  According to this procedure, the smaller value of the current grade and the voltage grade in the antecedent part is selected as the grade of the antecedent part (so-called Min method), and the SOC value obtained in the SOC membership calculation step in the antecedent part is selected. When the values are the same, the larger value is selected as the grade of the SOC value (so-called Max method). The figure centroid of the SOC profile figure calculated in the consequent part is calculated as a charge / discharge state estimation value of the secondary battery (so-called centroid method).

(Claim 3)
In the secondary battery charge state estimation method according to the present invention, a step of performing a filter operation on the estimated value of the SOC calculated by the above-described method may be provided.

  According to this procedure, the SOC value indicating the charge / discharge state of the secondary battery is processed by the digital filter, so that unnecessary noise components can be removed. The filter operation is preferably calculated by FIR, but may be an IIR filter. In general, since the filter operation is accompanied by a time delay, a rapid change in SOC is mitigated.

(Claim 4)
In the secondary battery charge state estimation method according to the present invention, the filter calculation performed on the estimated value of the SOC may be a primary delay.

  According to this procedure, if the first-order lag calculation is performed as the filter calculation, the calculation formula is relatively simple, so that the processing time is short. Since the degree of time delay can be determined by the first-order delay time constant, it can be adjusted.

(Claim 5) (Claim 6)
In the secondary battery charge state estimation method according to the present invention, as a consequent part, based on the charge / discharge characteristics of a plurality of secondary batteries prepared in advance, the current value and the voltage value for each charge / discharge characteristic. The method may include a step of calculating a provisional SOC value of the secondary battery and a step of calculating the SOC membership from the provisional SOC value and the current grade value or the voltage grade value. The SOC membership may be a single numerical value.

  According to this procedure, since a singleton of SOC is obtained by the simplified inference method, the calculation is simplified, the calculation processing time can be shortened, and the program development time can be expected to be shortened.

  The charge / discharge characteristics mentioned here may be a so-called SOC map in which the battery voltage is taken on the vertical axis and the SOC is taken on the horizontal axis, and is a map of the charge state, the map of the discharge state, and the low state of charge / discharge. Also good. A more appropriate SOC can be obtained by switching the map for calculating the provisional SOC value depending on whether the secondary battery is in a charged state or in a discharged state. The SOC value obtained here is not a final SOC estimated value, but is provisionally obtained for convenience of calculation processing.

(Claim 7)
In the secondary battery charge state estimation method according to the present invention, the provisional SOC value calculation step may calculate the provisional SOC value by linear interpolation of the current value.

  According to this procedure, generally, when obtaining the SOC characteristics, charging is performed at a constant current, discharging is performed at a constant current, and the time elapsed at that time, and therefore, the integral value of the current is obtained. Therefore, the SOC map is expressed using the charging current or discharging current as a parameter. From the SOC map prepared in advance, an SOC map is created by performing linear interpolation on the measured current value, and the provisional SOC is obtained from the map.

(Claim 8)
The secondary battery charge state estimation method according to the present invention includes a secondary battery charge state estimation step for estimating a charge state of a secondary battery for each of a plurality of battery modules constituting the secondary battery, and the secondary battery Selecting a minimum value of the secondary battery charge state estimation step when the battery is in a charged state, and selecting a maximum value of the secondary battery charge state estimation step when the secondary battery is in a discharged state. A delay calculation step of performing a delay calculation on the calculation result of the selection step.

  According to this procedure, when the secondary battery constitutes a battery system composed of a plurality of battery modules, the charge / discharge state of the battery system can be estimated. That is, when the battery module is in the charged state, the minimum value among the values indicating the charge / discharge state (that is, the estimated SOC value) is set as the estimated SOC value of the battery system, and when the battery module is in the discharged state, the maximum value of the estimated SOC value. Is the estimated SOC of the battery system. By doing in this way, it becomes possible to discover the battery module which may be overcharged or overdischarged at an early stage.

  If the estimated SOC value is calculated for the unit battery constituting the battery module, and the maximum or minimum SOC estimated value can be selected according to the charge / discharge state, the charge / discharge state of the battery system can be estimated for each unit battery. It becomes.

(Claim 9)
In order to achieve the above object, a state of charge estimation device for a secondary battery according to the present invention is a device for estimating a state of charge of the secondary battery in a power storage facility having a secondary battery, the current of the secondary battery A current sensor for detecting a value, a voltage sensor for detecting a voltage value of the secondary battery, a current membership calculation unit for calculating a current grade from the current value, and a voltage grade from the voltage value as an antecedent part A voltage membership calculation unit that calculates the SOC, and an SOC membership calculation unit that calculates an SOC membership from the current grade and a fuzzy inference rule from the current grade as a consequent part, and an output of the antecedent part A fuzzy inference operation unit is provided for performing a composition operation on the output of the consequent unit.

  As a specific example, the secondary battery charge state estimation device according to the present invention is configured by a computer system including a current sensor, a voltage sensor, a computer, and software for performing a series of data processing.

  According to this configuration, the current membership function and the voltage membership function may be, for example, control data stored in a storage device connected to a computer, and similarly store fuzzy inference rules and SOC membership functions. Control data stored in the apparatus may be used. If control data is not stored in a program but stored in a storage device, it is convenient for tuning in fuzzy inference.

  Each calculation unit is configured by a program module, and the estimated SOC value is displayed and output on an external display device.

(Claim 10)
The secondary battery state of charge estimation device according to the present invention comprises a minimum value selection circuit for selecting the smallest value of the current grade and the voltage grade in the antecedent part, and the SOC membership in the consequent part. The calculation unit may include a maximum value selection circuit, and the synthesis calculation unit may include a centroid calculation unit that calculates the centroid of the SOC membership.
According to this configuration, the maximum value selection circuit and the minimum value selection circuit may be configured by an electronic circuit, and preferably by a program module.

(Claim 11)
In the secondary battery charge state estimation device according to the present invention, a delay filter calculation unit may be connected to an output of the fuzzy inference calculation unit.

(Claim 12)
In the secondary battery charge state estimation device according to the present invention, as a consequent part, based on the charge / discharge characteristics of a plurality of secondary batteries prepared in advance, the current value and the voltage value for each charge / discharge characteristic. A provisional SOC calculation unit for calculating the SOC of the secondary battery; and a SOC estimation value calculation unit for calculating a single SOC membership from the output of the provisional SOC calculation unit and the current grade value or the voltage grade value. You may have.

  According to this configuration, the SOC map representing the charge / discharge characteristics prepared in advance may be stored as a file in the storage device connected to the computer. The SOC map for calculating the provisional SOC value may be obtained by linear interpolation with respect to the current value from the SOC map stored as a file in the storage device. In this case, the SOC map obtained by linear interpolation is stored in the computer. It may be a temporary memory.

(Claim 13)
In the secondary battery charge state estimation device according to the present invention, the secondary battery is composed of a plurality of battery modules, and the secondary battery charge state estimation device provided for each battery module includes a maximum value selection circuit and a minimum value selection. A value of a minimum value selection circuit is selected when the secondary battery is in a charged state, and a maximum value selection circuit is selected when the secondary battery is in a discharged state. A delay calculation unit may be connected to the delay unit.

The present invention has the above-described procedure and makes it possible to estimate the SOC with high reliability by performing fuzzy inference reflecting the expert's knowledge database, thereby preventing overcharge and overdischarge. This can prevent a decrease in battery life.
In addition, by accurately estimating the SOC, the state of charge of the secondary battery can be accurately grasped, and an appropriate SOC can be maintained, and the battery can be appropriately operated.

It is a system configuration figure showing the outline of the charge condition estimating device of a rechargeable battery concerning one embodiment of the present invention. It is a block diagram which shows the outline of the battery module measurement part which concerns on one Embodiment of this invention. It is a partially broken side view which shows the battery module which concerns on one Embodiment of this invention. It is the system block diagram showing the charge condition estimation method of the secondary battery which concerns on one Embodiment of this invention using the block diagram. This is a membership function for obtaining the current grade in the antecedent part. This is a membership function for obtaining a voltage grade in the antecedent part. It is explanatory drawing which shows a fuzzy inference rule. It is a membership function which calculates | requires SOC in the consequent part which concerns on other embodiment. A charge map is shown among the charge-discharge characteristic diagrams which show SOC with respect to battery voltage. A discharge map is shown among the charge-discharge characteristic diagrams which show SOC with respect to battery voltage. The charging condition estimation method which concerns on one Embodiment of this invention is represented with the flowchart. It is a figure which shows the test result which concerns on the charge condition estimation method which concerns on one Embodiment of this invention.

Hereinafter, embodiments according to the present invention will be described with reference to the drawings. However, the present invention is not limited to the embodiments.

  As shown in FIG. 1, the secondary battery charge state estimation device 1 according to this embodiment is connected to a secondary battery, and the secondary battery is composed of a plurality of battery modules B. You may comprise. In order to distinguish between the battery module and the secondary battery according to the present invention, the secondary battery according to the present invention constituted by the battery module is hereinafter referred to as a battery system A.

  In the present embodiment, the secondary battery is a nickel hydrogen battery, but may be another type of secondary battery such as a lead storage battery or a lithium ion battery. The battery module is configured by stacking a plurality of unit batteries. In FIG. 3, the module main body 2 which is a main part of the battery module B is shown partially broken.

  The battery module B includes a battery stack 3 in which a plurality of unit batteries C, which are sealed rectangular batteries, are stacked in the thickness direction of the unit battery C, and a side plate for fastening and fixing the battery stack 3 in the stacking direction X. 3 and the compression plate 5 are provided as main components.

  Here, as shown in FIG. 4, the battery stack 3 of the battery module B is obtained by stacking the unit batteries C and the heat radiating plates 6, and the unit battery C is one surface of the adjacent unit battery C. And the other surface are laminated in a direction opposite to each other, and each surface serves also as a positive electrode side terminal and a negative electrode side terminal of the unit battery C.

  A wiring (not shown) is connected to each of the positive electrode side terminal and the negative electrode side terminal of the unit battery C, and is connected to the voltmeter 13 via the wiring (not shown).

  The power of the battery module B can be taken out from the battery terminals 7 arranged in front of and behind the compression plate 5.

  In FIG. 1, a plurality of battery modules B are connected in series, and their outputs are connected to a load (not shown). A current sensor 12 is attached to the power wiring 11 that connects the battery module B and the load, and the current value flowing from the battery system A to the load (not shown) can be measured. The number of battery modules B connected in series is determined by the voltage specifications required from the load side. In this embodiment, 18 series batteries are connected to supply a voltage with a rating of 1.2 V × 10 × 18 = 216 V to the load. .

  A system voltage sensor 10 for measuring the voltage of the entire battery system is attached in the middle of the power wiring 11 connected to the load of the battery system A.

  The current sensor 12 and the system voltage sensor 10 are connected to the estimation calculation processing device 21 via the wirings 12a and 10a, and the current value and the voltage value of the battery system A are incorporated in the estimation calculation processing device 21. It is possible to measure.

  On the other hand, in FIG. 2 showing the details of the battery module part of FIG. 1, the pressure sensor 15 and the temperature sensor 14 attached to the battery module B are connected to the module measuring instrument 18 via the wirings 15a and 14a, respectively. The pressure and temperature of the battery module B can be measured. The voltage sensor 13 attached to the unit battery C is connected to the module measuring instrument 18 via the wiring 13a, and the voltage of the unit battery C can be measured by the module measuring instrument 18. In addition, since the voltage sensor 13 is prepared by the number of unit batteries, in this embodiment, 30 voltage sensors 13 are attached to one battery module B.

  The module measuring instrument 18 attached to each battery module B is connected to the communication control unit 24 of the measuring unit 22 built in the estimation calculation processing device 21 in accordance with an equation based on the communication line 18a. The pressure and temperature of the battery module B and the voltage signal for each unit battery C sent from the are received, sent to the measuring unit and converted into process data (numerical values).

  In the present embodiment, CAN (Control Area Network) is adopted as a communication line system. CAN is a method of serial communication line, but it may be a communication line of another method, or may be parallel communication or process interface.

  The process data measured by the measurement unit 22 is sent to the calculation unit 23 built in the estimation calculation processing device 21 via an internal bus (not shown), where the estimation calculation of the SOC for each battery module B is performed. Executed.

  The calculation unit 23 receives the process data from the measurement unit 22 and processes the antecedent part of the fuzzy inference, and calculates the SOC membership from the calculation result of the antecedent part processing part and the fuzzy inference rule. A consequent part processing unit 26 for calculating, and a composition operation part 27 for calculating the center of gravity of the SOC membership from the calculation result of the consequent part processing unit 26 are provided.

  Further, the arithmetic processing unit 21 includes an SOC calculation unit that calculates a provisional SOC from process data (current value, voltage value) from the measurement unit 22 and various SOC maps stored in the basic characteristic data storage unit 30. 28, the provisional SOC value calculated by the SOC calculation processing unit is sent to the consequent processing unit 26 and used for the calculation of the SOC membership.

  The estimated value of the SOC for each battery module B calculated by the combining calculation unit 27 is displayed and output on the estimation result display 31 together with the identification number of the battery module B. Moreover, the estimated value of SOC of the battery system A is displayed and output as a reference value.

  The functional block diagram of FIG. 4 showing the charging state estimation method of the secondary battery according to the embodiment of the present invention in the functional block diagram corresponds to the system configuration diagram of 1, but does not necessarily correspond to the one-to-one. It is not a thing.

  In the fuzzy rule antecedent section 41 of FIG. 4, the antecedent section membership calculation is executed from the battery voltage E and the battery current J.

  The battery current J is a value obtained by sampling the current value measured by the current sensor 12 at regular intervals and then performing anti-aliasing filtering. The voltage of the unit battery C is referred to as a cell voltage in order to distinguish it from the voltage of the battery module B. The battery voltage E is obtained by sampling 30 cell voltages at regular intervals and then anti-aliasing filtering. It is a numerical value obtained by adding up the values and dividing by 30 and represents the average value of the unit battery (cell) voltage.

  For this reason, the battery current J is a value common to the 18 battery modules B, but the battery voltage E can take different values for each battery module. The current value is not expressed in amperes but in a multiple of C, which is a characteristic value of the battery.

  5 and 6 show fuzzy functions of the battery current J and the cell voltage E, respectively. In both FIGS. 5 and 6, the current J and cell voltage E, which are process data, are scaled on the horizontal axis, and the current grade and voltage grade are normalized and scaled between 0 and 1 on the vertical axis.

  When calculating the fuzzy function, two current grades are calculated when the battery current J is between -1 and 1, and two voltage grades are calculated when the cell voltage E is between 1.21V and 1.37V. If it is in the other range, one current grade or voltage grade is calculated. Then, the smaller value (minimum value) of the current grade and voltage grade is adopted as the grade value of the antecedent part (Min method). This series of calculation processing is referred to as antecedent part membership calculation processing. These processes are calculated by the antecedent part processing unit 25 in the calculation unit 23 built in the estimation calculation processing device 21 shown in FIG.

  The block of the fuzzy rule consequent part 42 calculates the SOC membership by applying a fuzzy inference rule to the processing result of the fuzzy rule antecedent part 41 and the provisional SOC value calculated separately. At this time, when the same SOC has different grades, the largest grade is set as the grade of the SOC (Max method). This series of calculation processing is referred to as consequent part membership calculation processing, and is calculated in the consequent part processing unit 26 shown in FIG.

  FIG. 7 shows a fuzzy inference rule according to this embodiment as an example.

  The provisional SOC value is determined by using the charge map 51 when the battery module B is in a charged state, using the discharge map 52 when the battery module B is in a discharged state, and stopping charging / discharging when the battery current J is zero or almost zero. Using the map 53, the SOC calculation blocks 56, 57 and 58 calculate the battery current J and the cell voltage E.

  FIG. 10 shows an example of the charge map and charge / discharge stop map, FIG. 9 shows an example of the discharge map, the vertical axis indicates the cell voltage E (V), the horizontal axis indicates SOC (%), and the battery current J is a parameter. Shown in These maps 51, 52 and 53 are stored in the basic characteristic data storage unit 30 of FIG.

  In the charge map 51 and the discharge map 52, the provisional SOC value is calculated using characteristics obtained by performing linear interpolation using the battery current J from a plurality of current curves.

  In the present embodiment, the consequent processing is executed by calculating a singleton by a so-called simplified inference method. Accordingly, the SOC membership calculated here is a singleton having a single numerical value.

  The SOC estimation block 43 receives the processing of the fuzzy rule consequent unit 42, performs the SOC membership synthesis operation, and outputs the result as the SOC estimated value. In the present embodiment, the synthesis operation is preferably executed by the centroid method (centroid method). This processing is performed by the synthesis operation unit 27 shown in FIG. The center of gravity when the membership is a singleton is obtained by Equation 1.

_Here, y _1, y 2, the · · _{y n singleton, h 1, h 2, ··} h n is increment the horizontal axis, a center of gravity which is _{y 0} determined.

The estimated SOC value calculated in the SOC estimation block 43 is time-delayed in the next-stage delay processing block 44 and output as the SOC estimation result. In the present embodiment, the time delay process is a primary time delay process using an FIR filter, but the present invention is not limited to this.
Since the series of SOC inference processes is executed for each battery module B, the estimated SOC value is displayed and output together with the identification number for each of the 18 battery modules B.

  Further, when the battery system A is in a charged state, the identification number of the battery module B that has the smallest SOC estimated value among the 18 battery modules B and the estimated SOC value are displayed outside as the estimated value of the battery system A. Is output. Further, when the battery system A is in a discharged state, the identification number of the battery module B having the maximum estimated SOC value and the estimated SOC value are displayed and output to the outside as the estimated value of the battery system A.

Due to the nature of secondary batteries, it is difficult to avoid product variations, and such variations may be found at the use stage rather than at the manufacturing stage. Due to such product variations, the state of charge of the battery may vary from one battery module to another. When a secondary battery is used as a battery system, such as a power storage facility, detecting the battery module that has been most overcharged or overdischarged means that the secondary battery is significantly damaged by overcharge and overdischarge, and the battery life It is important to detect the battery module that exhibits the maximum or minimum SOC estimation value in the battery system A, and is significant.

  FIG. 11 shows a flowchart of processing according to the present embodiment.

  FIG. 12 shows a test result regarding the SOC estimation according to the present embodiment. In FIG. 12, the vertical axis shows the SOC in percent and the horizontal axis shows the time (minute: second). This is a test result in a charged state from when the SOC is 0% for a certain battery module, and the estimated SOC value is indicated by a bold line. The thin line is the estimation result of the conventional method and is shown for reference. The wavy line is the calculated SOC value obtained by integrating the battery current, and is calculated by a method that is said to show a relatively correct value during charging in the initial stage of the product.

  FIG. 8 shows the SOC membership function of the consequent part according to another embodiment. In the consequent part calculation process according to another embodiment shown here, the consequent part calculation process is performed using the membership function shown in FIG. 8 instead of the simplified inference method. This calculation process is executed by the consequent part processing unit 26 in FIG.

In the calculation of the antecedent part, the minimum value of the current grade and the voltage grade is adopted as the grade value of the antecedent part (Min method), and the maximum value (Max) among the SOC membership functions corresponding to the minimum grade value of the antecedent part. Law) is the calculation result of the consequent part processing unit 26.
In the composition calculation unit 27, the following formula 2 is used instead of the formula 1.

Here, _{y 0} is the center of gravity for obtaining, M (y) is the membership function of the SOC obtained by the consequent processing section 26, y corresponds to SOC.

  The secondary battery charge / discharge state estimation method according to the present invention can be used in a power storage and supply device using a secondary battery as a power storage means.

A battery system
B Battery module
C unit battery
1 Charge state estimation device
2 Module body
3 Battery stack
4 Side plate
5 Compression plate
6 Heat sink
7 Battery terminal
10 System voltage sensor
11 Power wiring
12 Current sensor
13 Voltage sensor
14 Temperature sensor
15 Pressure sensor
18 module measuring instruments
21 Estimator
22 Measuring unit
23 Calculation unit
24 Communication control unit
25 Antecedent Processing Unit
26 Consecutive Department Processing Department
27 Composite operation section
28 SOC calculator
30 Basic characteristic data storage
31 Estimation result display
41 Fuzzy rule antecedent block
42 Fuzzy rule consequent block
43 SOC estimation block
44 time delay processing block
45 SOC estimated value block
51 Charging map
52 Discharge map
53 Charge / discharge stop map
56, 57, 58 SOC calculation block

Claims (13)

A method for estimating a charging state of the secondary battery in a power storage facility having a secondary battery,
Measuring a current value of the secondary battery;
Measuring a voltage value of the secondary battery;
As an antecedent part, a current membership calculation step for calculating a current grade from the current value, and a voltage membership calculation step for calculating a voltage grade from the voltage value;
And a SOC membership calculation step of calculating an SOC membership from the current grade and a fuzzy inference rule from the current grade as a consequent part,
A method for estimating a charged state of a secondary battery, comprising a synthesis operation step for combining the calculation result of the antecedent part and the calculation result of the consequent part. In the antecedent part, including the step of setting the smallest value of the current grade and the voltage grade as the calculation result,
In the consequent part, the SOC membership calculation step comprises the step of setting the largest value as the calculation result,
The secondary battery charged state estimation method according to claim 1, wherein the combining calculation step includes a step of obtaining a center of gravity of the SOC membership.   The secondary battery charged state estimation method according to claim 1, further comprising a step of performing a filter operation on the estimated SOC value calculated in claim 1.   The secondary battery charge state estimation method according to claim 2 or claim 3, wherein the filter operation in claim 2 is a first-order lag. As the consequent part,
Calculating a provisional SOC value of the secondary battery for each charge / discharge characteristic from the current value and the voltage value based on the charge / discharge characteristics of the plurality of secondary batteries prepared in advance;
The secondary battery charged state estimation according to any one of claims 1 to 4, further comprising: calculating the SOC membership from the provisional SOC value and the current grade value or the voltage grade value. Method.   The secondary battery charge state estimation method according to claim 5, wherein the SOC membership is a single numerical value.   The secondary battery charge state estimation method according to claim 5 or 6, wherein, in the temporary SOC value calculation step, the temporary SOC value is calculated by linear interpolation of the current value. The secondary battery charging state estimation step according to claim 1 to 7, wherein the charging state of the secondary battery is estimated for each of a plurality of battery modules constituting the secondary battery.
When the secondary battery is in a charged state, the minimum value of the secondary battery charged state estimating step is selected, and when the secondary battery is in a discharged state, the maximum value of the secondary battery charged state estimating step is selected. With a selection step to
The secondary battery charged state estimation method according to any one of claims 1 to 7, further comprising a delay calculation step of performing a delay calculation on the calculation result of the selection step. An apparatus for estimating a charge state of the secondary battery in a power storage facility having a secondary battery,
A current sensor for detecting a current value of the secondary battery;
A voltage sensor for detecting a voltage value of the secondary battery;
As the antecedent part, a current membership calculation unit for calculating a current grade from the current value, a voltage membership calculation unit for calculating a voltage grade from the voltage value,
An SOC membership calculation unit for calculating an SOC membership from the current grade and the current grade by fuzzy inference rules as a consequent part;
A secondary battery charged state estimation device provided with a fuzzy inference calculation unit for combining and calculating the output of the antecedent part and the output of the consequent part. A minimum value selection circuit for selecting the smallest value of the current grade and the voltage grade in the antecedent part and the SOC membership calculation part in the antecedent part comprises a maximum value selection circuit;
The secondary battery charged state estimation device according to claim 9, wherein the composite calculation unit includes a centroid calculation unit that calculates a centroid of the SOC membership.   The secondary battery charge state estimation device according to claim 9 or 10, wherein a delay filter calculation unit is connected to an output of the fuzzy inference calculation unit. As the consequent part,
Based on the charge / discharge characteristics of the plurality of secondary batteries prepared in advance, a provisional SOC calculation unit that calculates the SOC of the secondary battery for each charge / discharge characteristic from the current value and the voltage value;
The SOC estimated value calculation part which calculates the said single SOC membership from the output of the said provisional SOC calculation part, and the said current grade value or the said voltage grade value, The any one of Claims 9-11 provided with. Secondary battery charge state estimation device. The secondary battery is composed of a plurality of battery modules,
The secondary battery charge state estimation device provided for each battery module is connected to a maximum value selection circuit and a minimum value selection circuit,
When the secondary battery is in a charged state, a value of a minimum value selection circuit is selected, and when the secondary battery is in a discharged state, a maximum value selection circuit is selected,
The secondary battery charge state estimation apparatus according to any one of claims 9 to 12, wherein a delay calculation unit is connected to an output of the selection circuit.